Let's get acquainted with the Intel H55 Express using the example of the ASRock H55M Pro motherboard. "Integrated" chipsets

H55 and H57 Express are two "integrated" chipsets from Intel.

Integrated is usually what we call solutions with integrated video, but now the GPU has left the chipset and moved into the CPU, as have the memory controller and PCI Express controller for graphics, so these chipsets are "integrated" in parentheses.

H55 and H57 are very close in functionality, but H57 is the senior, and H55 is the junior ICH PCH in the family, with reduced functionality.

If we compare the capabilities of these chipsets with the chipset for processors of the Socket 1156 - P55 socket, it turns out that the H57 is most similar to it, having only two differences in the implementation of the video system.

H57 Key Features:

. up to 8 PCIEx1 ports (PCI-E 2.0, but with PCI-E 1.1 data transfer speed);
. up to 4 PCI slots;

. the ability to organize a RAID array of levels 0, 1, 0+1 (10) and 5 with the Matrix RAID function (one set of disks can be used in several RAID modes at once - for example, on two disks you can organize RAID 0 and RAID 1, for each array its own part of the disk will be allocated);
. 14 USB 2.0 devices (on two EHCI host controllers) with the ability to individually disable;

H55 Features:

Supports all processors with the Socket 1156 socket (including the corresponding Core i7, Core i5, Core i3 and Pentium families) based on the Nehalem microarchitecture, when connected to these processors via the DMI bus (with a bandwidth of ~2 GB/s);
. FDI interface for receiving a fully rendered screen image from the processor and a unit for outputting this image to the display device(s);
. up to 6 PCIEx1 ports (PCI-E 2.0, but with PCI-E 1.1 data transfer speed);
. up to 4 PCI slots;
. 6 Serial ATA II ports for 6 SATA300 devices (SATA-II, second generation of the standard), with support for AHCI mode and functions like NCQ, with the ability to be individually disabled, with support for eSATA and port splitters;
. 12 USB 2.0 devices (on two EHCI host controllers) with the ability to individually disable;
. Gigabit Ethernet MAC controller and a special interface (LCI/GLCI) for connecting a PHY controller (i82567 for Gigabit Ethernet implementation, i82562 for Fast Ethernet implementation);
. High Definition Audio (7.1);
. harness for low-speed and outdated peripherals, etc.

The architecture is one chip, without division into north and south bridges (de facto it is just a south bridge).

The H57 has a specialized FDI interface, through which the processor sends the generated screen image (be it a Windows desktop with application windows, a full-screen movie demonstration or 3D games), and the chipset’s task is to pre-configure display devices to ensure timely output of this image to the desired screen ( Intel HD Graphics supports up to two monitors).

Any of the processors with the Socket 1156 socket will work in a board on any of these chipsets, the only question is whether its owner will lose integrated graphics, for which they have already paid anyway.
If you want to use Clarkdale's built-in graphics, get the H57.
If you want to create a normal (2 x16) SLI/CrossFire, take the P55.

When you plan to use one external video card for video, there is no difference at all between the P55 and H57.

Introduction

In this course work I will consider the “Integrated” Intel H55 and H57 chipsets. At the very beginning of January 2010, Intel practically ended the glorious era of processors based on the Core microarchitecture. Now, ironically, Core will (for some time) only produce ultra-budget models under the Celeron brand for Socket 775. As you already know from the presentation of processors based on the Clarkdale core, the updated platform implies the inclusion of new chipsets - H55 and H57 - among the possible applications. However, it cannot be said that the use of new chipsets is an indispensable condition or allows the potential of new processors to be fully revealed: in some places the potential will be revealed more fully, and in others it will be completely hidden. Well, let's get acquainted with the first "integrated" chipsets for Nehalem (more precisely, Clarkdale).

1. History of the creation of INTEL

It all started with the fact that in 1955, the inventor of the transistor, William Shockley, opened his own company, ShockleySemiconductorLabs in Palo Alto (which, among other things, served as the beginning of the creation of Silicon Valley), where he recruited quite a lot of young researchers. In 1959, for a number of reasons, a group of eight engineers left him, who were not satisfied with working “for their uncle” and wanted to try to implement their own ideas. The “eight traitors,” as Shockley called them, including Moore and Noyce, founded the company Fairchild Semiconductor.

Bob Noyce took on the position of Director of Research and Development at the new company. He later claimed that he came up with the microcircuit out of laziness - it seemed rather pointless when, in the process of manufacturing micromodules, silicon wafers were first cut into individual transistors, and then again connected to each other into a common circuit. The process was extremely labor-intensive - all connections were soldered by hand under a microscope! - and dear. By that time, Fairchild employee, also one of the co-founders, Jean Hoerni, had already developed the so-called. planar transistor production technology, in which all working areas are located in the same plane. Noyce proposed isolating individual transistors in a crystal from each other with reverse-biased p-n junctions, covering the surface with an insulating oxide, and making interconnections by sputtering strips of aluminum. Contact with individual elements was carried out through windows in this oxide, which were etched according to a special pattern with hydrofluoric acid.

Moreover, as he found out, aluminum perfectly adhered to both silicon and its oxide (it was the problem of adsorption of the conductor material to silicon that until recently did not allow the use of copper instead of aluminum, despite its higher electrical conductivity). This planar technology has survived to this day in a somewhat modernized form. To test the first microcircuits, a single device was used - an oscilloscope.

Meanwhile, it turned out that Noyce was ahead of him in the noble task of creating the first microcircuit. Back in the summer of 1958, Texas Instruments employee Jack Kilby demonstrated the possibility of manufacturing all discrete elements, including resistors and even capacitors, on silicon.

He did not have planar technology at his disposal, so he used so-called mesa transistors. In August, he assembled a working prototype of a trigger, in which individual elements he made with his own hands were connected with gold wires, and on September 12, 1958, he presented a working microcircuit - a multivibrator with an operating frequency of 1.3 MHz. In 1960, these achievements were demonstrated publicly - at the exhibition of the American Institute of Radio Engineers. The press greeted the opening very coldly. Among other negative features of the “integrated circuit”, non-repairability was called. Although Kilby applied for a patent in February 1959, and Fairchild did so only in July of the same year, the latter was granted a patent earlier - in April 1961, and Kilby - only in June 1964. Then there was a ten-year war about priorities, in the result of which, as they say, friendship won. Ultimately, the Court of Appeal upheld Noyce's claim to technological primacy, but ruled that Kilby was credited with creating the first working microcircuit. In 2000, Kilby received the Nobel Prize for this invention (among the other two laureates was Academician Alferov).

Robert Noyce and Gordon Moore left Fairchild Semiconductor and founded their own company, and Andy Grove soon joined them. The same financier who had previously helped create Fairchild provided $2.5 million, although the one-page business plan, typed by Robert Noyce in his own hand, was not very impressive: a lot of typos, plus statements of a very general nature.

Choosing a name was not an easy task. Dozens of options were proposed, but all of them were rejected. By the way, do the names CalComp or CompTek mean anything to you? But they could belong not to those popular companies that carry them now, but to the largest processor manufacturer - at one time they were rejected among other options. As a result, it was decided to name the company Intel, from the words "integrated electronics". True, we first had to buy this name from the motel group that had previously registered it.

So, in 1969 Intel started working with memory chips and achieved some success, but clearly not enough for fame. In its first year, revenue was only $2,672.

Today, Intel makes chips based on market sales, but in its early years the company often made chips to order. In April 1969, Intel was contacted by representatives of the Japanese company Busicom, which produces calculators. The Japanese heard that Intel has the most advanced chip production technology. For its new desktop calculator, Busicom wanted to order 12 microcircuits for various purposes. The problem, however, was that Intel's resources at that moment did not allow such an order to be completed. The methodology for developing microcircuits today is not very different from what it was in the late 60s of the 20th century, although the tools differ quite noticeably.

In those long, long ago years, very labor-intensive operations such as design and testing were performed manually. Designers drew drafts on graph paper, and draftsmen transferred them to special waxed paper (wax paper). The mask prototype was made by manually drawing lines onto huge sheets of Mylar film. There were no computer systems for calculating the circuit and its components yet. Correctness was checked by “traversing” all the lines with a green or yellow felt-tip pen. The mask itself was made by transferring the drawing from lavsan film onto the so-called rubilite - huge two-layer ruby-colored sheets. Engraving on rubilite was also done by hand. Then for several days we had to double-check the accuracy of the engraving. In the event that it was necessary to remove or add some transistors, this was again done manually, using a scalpel. Only after careful inspection was the rubilite sheet handed over to the mask manufacturer. The slightest mistake at any stage - and everything had to start all over again. For example, the first test copy of “product 3101” turned out to be 63-bit.

In short, Intel physically could not handle 12 new chips. But Moore and Noyce were not only wonderful engineers, but also entrepreneurs, and therefore they really did not want to lose a profitable order. And then it occurred to one of Intel’s employees, Ted Hoff, that since the company did not have the ability to design 12 chips, it needed to make just one universal chip that would replace them all in its functionality. In other words, Ted Hoff formulated the idea of ​​a microprocessor - the first in the world. In July 1969, a development team was created and work began. Fairchild transfer StanMazor also joined the band in September. The customer's controller included the Japanese Masatoshi Shima into the group. To fully ensure the operation of the calculator, it was necessary to manufacture not one, but four microcircuits. Thus, instead of 12 chips, only four had to be developed, but one of them was universal. No one had ever produced microcircuits of such complexity before.

What is a chipset?

Chipset – the basis of the motherboard, is a set of system logic chips. Through the chipset, all PC subsystems interact. Chipsets have a high degree of integration, and are (most often) two microcircuits (single-chip solutions are less common), which implement integrated controllers that ensure the operation and interaction of the main subsystems of the computer.

Almost all modern chipsets have a system logic set consisting of two north and south bridge microcircuits. The name of the microcircuits is due to their position relative to the PSI bus: north is higher, south is lower.

The north bridge chip ensures operation with the fastest subsystems.

It contains: a system bus controller, through which interaction with the processor occurs; a memory controller that works with system memory; an AGP (Accelerated Graphics Port) graphics bus controller that provides interaction with the graphics subsystem (today most chipsets support 1x/2x/4x interfaces, with 8th AGP speed coming soon); communication bus controller with the south bridge (PCI - buses in the classical sense).

The task of the north bridge is to organize servicing of requests to system memory with minimal delays. Solutions to this problem are based on the implementation of a memory controller, which allows you to simultaneously process a large number of requests and data, prioritizing and sequencing access to main memory. To make more efficient use of the memory bus, data buffering is used, ensuring simultaneous operation of the memory of several devices in access time sharing mode.

As mentioned earlier, the classic implementation of a two-bridge architecture involves using the PCI bus as a communication channel between bridges. But the 32-bit PCI bus operating at 33MHz has a peak throughput of 133Mb/s, which is not enough to meet the needs of modern peripheral devices. Therefore, most manufacturers use other interfaces to connect chipset chips, which in turn made it possible to move the PCI bus controller from the north bridge to the south bridge. The pioneer in this area was the hub architecture (Intel 800 series of chipsets). Its essence boils down to the transition to connecting bridges according to the “point-to-point” scheme. In this case, a special 8-bit bus was used, providing a bandwidth of 266 MB/s. The controller of this bus, using proprietary technologies, optimizes work with requests from peripheral devices to main memory. All this makes the operation of the hubs (north and south bridges) more independent and removes the restrictions imposed by the use of the PCI bus as a connecting link. Similar technologies are implemented in chipsets from VIA (V-Link Hub architecture) and in dual-processor solutions from SiS (MnTIOL bus).

The South Bridge ensures work with slower system components and peripheral devices. The following controllers and devices have become standard for the south bridge:

2. USB controller (one or more), providing operation with devices connected to the universal serial bus (USB), USB should replace outdated external interfaces, such as serial RS-232 (COM port) and parallel IEEE-1284 (LPT -port). Disadvantages of old solutions: low bandwidth, inability to hot-swappable and connecting several devices in a chain to the same port, as well as short interface cable length.

3. LPC (Low Pin Count Interface) bus controller, which replaced the outdated ISA. The LPC bus has a 4-bit interface connected to an input/output chip (Super I/O chip), which supports external ports (serial COM and parallel LPT, PS/2 and infrared), as well as a floppy drive controller.

Most modern chipsets implement the AC'97 audio controller (Audio Codec) in their south bridge. The AC’97 specification implies the separation of digital and analog processing processes, each of which is performed by a separate chip, while the AC-Link interface for their interaction is also defined. Thus, the south bridge processes the audio signal in digital form - in other words, it implements a digital part (Digital AC’97 Controller). To implement all the capabilities provided by the AC’97 specification, an AMP controller is integrated into the south bridge chip. The AMP cards (Audio/Modem Riser Card) it supports contain analog circuits of the AC’97 audio codec and/or the MC’97 modem codec (Modem Codec). The use of dual-chip chipsets allows the use of different combinations of north and south bridges, provided that they support the same interface. This makes it possible to create the most productive systems at minimal cost and in the shortest possible time, since to implement the latest specifications it is enough to upgrade only one system logic chip, and not the chipset as a whole.

Intel H55 and H57 Express

Why the chipsets are called “integrated” is obviously already well known: solutions with integrated video are usually called this, but now the graphics processor has left the chipset and moved to the central processor in the same way as the memory controller (in Bloomfield) and the PCI Express controller for graphics ( in Lynnfield) previously. In accordance with this, the Intel product range has slightly changed: the previous letter G has been replaced by H. H55 and H57 are indeed very close in functionality, and H57 of this pair is certainly the older one. However, if you compare the capabilities of the new products with the hitherto lonely chipset for processors of the Socket 1156 - P55 socket, it turns out that the H57 is most similar to it, having only two differences, precisely due to the implementation of the video system. H55 is the youngest PCH in the family, with reduced functionality.

H57 chipset specification

The key characteristics of the H57 are as follows:

· up to 8 PCIEx1 ports (PCI-E 2.0, but with PCI-E 1.1 data transfer speed);

· up to 4 PCI slots;

· 6 Serial ATA II ports for 6 SATA300 devices (SATA-II, second generation of the standard), with support for AHCI mode and functions like NCQ, with the ability to be individually disabled, with support for eSATA and port splitters;

· the ability to organize a RAID array of levels 0, 1, 0+1 (10) and 5 with the Matrix RAID function (one set of disks can be used in several RAID modes at once - for example, on two disks you can organize RAID 0 and RAID 1, for each the array will be allocated its own part of the disk);

· 14 USB 2.0 devices (on two EHCI host controllers) with the ability to individually disable;

P55, the differences between the newcomer were minimal. The architecture has been preserved (one chip, without division into north and south bridges - de facto it is just a south bridge), all traditional “peripheral” functionality has remained unchanged. The first difference is the implementation of a specialized FDI interface in the H57, through which the processor sends the generated screen image (be it a Windows desktop with application windows, a full-screen movie demonstration or 3D games), and the chipset’s task is to pre-configure display devices to ensure timely output of this image on the [required] screen (Intel HD Graphics supports up to two monitors. However, the very fact of additional interfaces between the processor and the chipset (previously between chipset bridges) is nothing new, and when we talk about the DMI bus as the only corresponding communication channel, then we mean only the main channel for transmitting broad-profile data, nothing more, and some highly specialized interfaces have always existed.

The second difference is impossible to notice on the block diagram of the chipset - however, it is impossible to notice it in objective reality, since it exists only in the reality of marketing. Here, Intel applies the same approach that segmented the chipsets of the previous architecture: the top chipset (today it is X58) implements two full-speed interfaces for external graphics, the mid-level solution (P55) - one, but divided into two at half speed, and low-end and integrated products of the line - one full-speed, without the ability to use a pair of video cards. It is quite obvious that the actual chipset of the current architecture cannot in any way affect the support or lack of support for two graphical interfaces (yes, by the way, the P45 and P43 were clearly the same chip). It’s just that when the system is initially configured, a motherboard based on H57 or H55 “does not find” options to organize the operation of a pair of PCI Express 2.0 ports, while a motherboard based on P55 manages to do this in a similar situation. The real, “iron” background of the situation generally makes no difference to the ordinary user. So, SLI and CrossFire are available on P55 based systems, but not on H55/H57 based systems.

The key characteristics of the H55 are as follows:

· support for all processors with the Socket 1156 socket (including the corresponding Core i7, Core i5, Core i3 and Pentium families) based on the Nehalem microarchitecture, when connected to these processors via the DMI bus (with a bandwidth of ~2 GB/s);

· FDI interface for receiving a fully rendered screen image from the processor and a unit for outputting this image to the display device(s);

· up to 6 PCIEx1 ports (PCI-E 2.0, but with PCI-E 1.1 data transfer speed);

· up to 4 PCI slots;

· 6 Serial ATA II ports for 6 SATA300 devices (SATA-II, second generation of the standard), with support for AHCI mode and functions like NCQ, with the ability to be individually disabled, with support for eSATA and port splitters;

· 12 USB 2.0 devices (on two EHCI host controllers) with the ability to individually disable;

· Gigabit Ethernet MAC controller and a special interface (LCI/GLCI) for connecting a PHY controller (i82567 for Gigabit Ethernet implementation, i82562 for Fast Ethernet implementation);

· High Definition Audio (7.1);

· harness for low-speed and outdated peripherals, etc.

There are already changes in support for traditional peripherals - although not too significant (it is almost impossible to determine by eye how many USB ports the chipset supports). It is clearly noticeable that the regression in this case “rolls back” the situation to the times of the ICH10/R southbridges: the H55 is deprived of precisely those changes that allowed us to propose the name ICH11R for the P55 at one time. H55 is pure ICH10, and without the letter R: the low-end chipset of the Intel 5x line also did not receive the functionality of a RAID controller. Of course, in this case the FDI interface was added to the list of ICH10 characteristics, and it is equally obvious that the H55 does not have support for SLI/CrossFire, or indeed two [normal] graphical interfaces. To summarize the differences: the most budget solution in the new line has 12 USB ports instead of 14 on the P55/H57, 6 PCI-E ports instead of 8, and does not have RAID functionality. The “peripheral” PCI Express controller still formally complies with the second version of the standard, however, the data transfer speed along its lines is set at the PCI-E 1.1 level (up to 250 MB/s in each of two directions simultaneously) - ICH10, definitely. How bad or good is the peripheral support of the new chipsets? In the case of the H57, this is still the same maximum, but not unique for today, set. In the case of the H55, we must assume that many will notice the absence of RAID (but, of course, not the enormous limitation of the number of USB ports to 12). Actually, buyers might not have noticed (few people still need more than one hard drive at home), but how to sell motherboards without RAID? Well, very cheap microATX models, of course, will be released anyway - Intel, for example, offers such a solution as a reference for the new platform. But more serious products without the usual attribute... hardly. This means that they will solder an additional RAID controller, bringing the already excessive number of SATA ports to 8-10. On the other hand, perhaps the H55 will have its own well-defined niche, and more demanding (or those who don’t know exactly what they want) buyers will be offered models based on the H57. The difference in the selling price of chipsets ($3) is unlikely to significantly affect the price of the final product.

Comparison table of motherboard characteristics

ASUS P7H55-M Pro

ASUS has the widest range of motherboards based on the Intel H55 chipset, which includes six models. Among them, the P7H55-M Pro is a mid-range product without any unique features. Accordingly, its expansion capabilities and functionality will satisfy the needs of most users, as will the price, which is about 3,600 rubles.

Let's start with the fact that the configuration of the ASUS P7H55-M Pro expansion slots is the most optimal, and includes one PEG slot, one PCI Express x1 slot and a pair of PCI slots.

We didn't have any complaints about the rear panel configuration, although we wouldn't mind the additional DisplayPort video output.

The processor power subsystem is made according to a 4-phase circuit, and the memory controller power converter is made according to a 2-phase circuit.

The ASUS P7H55-M Pro motherboard supports a large number of proprietary utilities and technologies. These include the Express Gate shell, the MyLogo 2 POST screen replacement function, as well as the BIOS firmware recovery system - CrashFree BIOS 3. Note the support for BIOS settings profiles - OC Profile:

As well as the multifunctional utility TurboV EVO, which, in addition to overclocking the processor and memory, allows you to overclock the built-in graphics core:

As for the BIOS, the board boasts a very large set of RAM settings.

System monitoring is performed at a very high level. In particular, the board displays the current temperatures of the processor and system, monitors voltages and rotation speeds of all fans, which, using the Q-Fan2 function, can change the rotation speed depending on the temperature of the processor and system.

Overclocking capabilities are concentrated in the "AI Tweaker" section, and do not have any disadvantages:

In particular, on the ASUS P7H55-M Pro board we achieved stable system operation at a Bclk frequency of 190 MHz.

It is quite easy to formulate conclusions about the ASUS P7H55-M Pro motherboard, since the price of the product fully corresponds to its main capabilities, and as a bonus, the user receives support for the ParallelATA protocol, as well as a lot of additional ASUS technologies.

· high stability and performance;

· 6-phase processor power supply;

· support for one P-ATA channel (JMicron JMB368);

· High Definition Audio 7.1 sound and Gigabit Ethernet network controller;

· support for USB 2.0 interface (twelve ports);

· a wide range of proprietary ASUS technologies (PC Probe II, EZ Flash 2, CrashFree BIOS 3, MyLogo 2, Q-Fan, etc.);

· additional set of AI Proactive technologies (AI Overclock, OC Profile (eight profiles), AI Net 2, TurboV EVO, EPU, etc.).

· not detected.

Board Features:

· powerful overclocking functions and fairly high results;

· no support for LPT and FDD interfaces;

· only one PS/2 port.

Conclusion

In this course project, I had to get acquainted with the “Integrated” Intel H55 and H57 chipsets. First of all, you need to understand that incompatibility between different chipsets and processors of this socket is non-fatal. Any of these processors will work in a board on any of these chipsets, the only question is whether its owner will lose the integrated graphics, for which they have already paid anyway. It seems that everything is simple: if you want to use the built-in Clarkdale graphics, take the H57. If you want to create a normal (we don’t say “full”, 2 x16) SLI/CrossFire - take the P55. We can't do it together. And in the most likely intermediate case, when exactly one external video card is planned to be used as video? In this case, there is no difference at all between the P55 and H57, and even the selling price does not play a role here - you will be buying a motherboard in a store, and not a chipset chip near the entrance to the Intel factory.

Today we will look at the first motherboard based on the Intel H55 Express chipset, designed to work in tandem with 1156-pin processors from the same manufacturer. This is the first such board that came into our laboratory, so let's start by introducing this logic set and related ones. And let's go, as usual, from afar :).

In relation to computers intended for household use, the generally accepted classification includes four market segments: flagship, productive, mass and budget.

advertising

When at the end of 2008 Intel introduced the new Nehalem architecture in the form of Core i7 processors based on the Bloomfield core with 1366 pins and the corresponding X58 Express logic set, few would have thought that this would be all there was to it. Several CPU models and a single chipset are all that the world's leading processor manufacturer still offers in the top segment.

However, the rest were completely left to processors with a 775-pin socket, whose history stretches back to 2004, the time of the NetBurst architecture. Intel, indeed, was in no hurry to bring a new platform to the market: its Core 2 CPU still felt very good in the fight against AMD Athlon and Phenom.

But after the appearance of the Phenom II processors, thanks to which the main competitor managed to get closer to the mass and productive solutions of Intel both in terms of specific performance (per GHz) and frequency potential, it was impossible to postpone the announcement of the new platform. Therefore, at the end of last summer 2009, a combination of processors with an LGA 1156 socket and the P55 Express logic set was presented. There are only a few CPU models (all quad-core, Lynnfield core), and again only one set of logic. It seemed that history was repeating itself.

However, the processor socket with 1156 pins was initially conceived as a complete replacement for the “old man” LGA 775. And at the very beginning of 2010, the expected expansion took place. Intel presented a whole “pack” of processors based on the Clarkdale core, as well as several sets of logic intended for them. However, the P55 Express is also compatible with new CPUs - there are no exceptions in terms of support for processors between chipsets (yet). But they still differ significantly from each other. Let's try to summarize these differences in one table.

Briefly about the new processors and chipset

In the last issue of our magazine, in the article “New 32nm Intel Core i5-661 processor,” we spoke in detail about the new Clarkdale processors and the Intel H55 Express chipset, and therefore we will not repeat ourselves once again and will only briefly recall the main features of the new series of processors and the new chipset.

So, the family of all 32nm Intel processors has the common code name Westmere. At the same time, the microarchitecture of the new processors itself remains the same, that is, the cores of these processors are based on the Nehalem processor microarchitecture.

The Westmere family includes desktop, mobile and server processors. Desktop processors include Gulftown and Clarkdale processors.

The six-core Gulftown processor is aimed at high-performance solutions, and the dual-core Clarkdale processors are aimed at low-cost mass solutions.

Clarkdale processors have an integrated dual-channel DDR3 memory controller and normally support DDR3-1333 and DDR3-1066 memory.

Each Clarkdale processor core has a Level 1 (L1) cache, which is divided into an 8-channel 32 KB data cache and a 4-channel 32 KB instruction cache. In addition, each core of the Clarkdale processor is equipped with a unified (same for instructions and data) second level cache (L2) of 256 KB in size. The L2 cache is also 8-channel, and its line size is 64 bytes. Also, all Clarkdale processors have a third level cache (L3) of 4 MB (2 MB for each processor core). The L3 cache is 16-channel and inclusive in relation to the L1 and L2 caches, that is, the L3 cache always duplicates the contents of the L1 and L2 caches.

All Clarkdale processors feature an LGA 1156 socket and are compatible not only with the new Intel H55 Express chipset, but also with the Intel H57 Express and Intel Q57 Express chipsets, as well as the Intel P55 Express chipset.

The Clarkdale processor family includes two series: Intel Core i5 600 series and Intel Core i3 500 series. The 600 series includes four models: Intel Core i5-670, Core i5-661, Core i5-660 and Core i5-650, and the 500 series includes two: Intel Core i3-540 and Core i3-530.

One of the main innovations of Clarkdale processors is that they have an integrated graphics core, that is, both the CPU and GPU will be located in the same case.

A pair of processor cores with 4 MB of third-level cache are manufactured using a 32-nm process technology, and the integrated graphics core and built-in memory controller are manufactured using 45-nm technology.

Of course, the graphics core integrated into the processor cannot compete with discrete graphics and is not intended for use in 3D games. At the same time, support for hardware decoding of HD video is announced, so these processors with integrated graphics can find use in multimedia centers for playing video content.

Despite the presence of an integrated graphics core in Clarkdale processors, they also have a built-in PCI Express v.2.0 interface with 16 lanes for using discrete graphics. When Clarkdale processors are used in conjunction with motherboards based on the Intel H55 Express chipset, the 16 PCI Express v.2.0 lanes supported by the processor can only be grouped as one PCI Express x16 channel.

Naturally, support for the PCI Express v.2.0 interface for using discrete graphics directly by the Clarkdale processor itself deprives it of the need to use a high-speed bus to connect the processor to the chipset. Therefore, Clarkdale processors, just like Lynnfield processors, use a bidirectional DMI (Direct Media Interface) bus with a bandwidth of 20 Gbit/s (10 Gbit/s in each direction) to communicate with the chipset.

Another feature of Clarkdale processors is support for the new generation of Intel Turbo Boost technology. Intel Turbo Boost Technology is only available on Intel Core i5 600 series processors and is not available on Intel Core i3 500 series processors.

For all Intel Core i5 600 series processors, if both processor cores are active, Intel Turbo Boost mode can increase their clock speed by one step (133 MHz), and if only one processor core is active, then its clock speed can be increased by two steps (266 MHz).

Another feature of all Intel Core i5 600 series processors is that they feature hardware acceleration of the Advanced Encryption Standard (AES) encryption and decryption algorithm to ensure data security. Again, the Intel Core i3 500 series processors do not have hardware encryption acceleration.

The next important point: all Clarkdale processors support Hyper-Threading technology, as a result of which the operating system sees a dual-core processor as four separate logical processors.

The differences between the Intel Core i5 600 series processor models are the clock speed, graphics core frequency, their TDP, and support for Intel vPro technology and virtualization technology.

Thus, all Intel Core i5 600 series processors have a graphics core frequency of 773 MHz and a TDP of 73 W, with the exception of the Intel Core i5-661 model, which has a graphics core frequency of 900 MHz and a TDP of 87 W. In addition, all Intel Core i5 600 series processors, except the Intel Core i5-661 model, support Intel vPro technology and virtualization technologies (Intel VT-x, Intel VT-d). The Intel Core i5-661 processor does not support Intel vPro technology and only supports Intel VT-x technology.

All processors of the Intel Core i3 500 series family have a graphics core frequency of 733 MHz and a TDP of 73 W. Additionally, these processors do not support Intel vPro technology and only support Intel VT-x technology.

After a brief overview of the features of Clarkdale processors, let's look at the new Intel H55 Express chipset.

The Intel H55 Express chipset (Fig. 1), or, in Intel's terminology, a platform hub (Platform Controller Hub, PCH), is a single-chip solution that serves as a replacement for the traditional north and south bridges.

Rice. 1. Block diagram of Intel H55 Express chipset

As already noted, in Clarkdale processors the interaction between the processor and the chipset is implemented via the DMI bus. Accordingly, the Intel H55 Express chipset has a DMI controller.

In addition, to support the graphics core built into the Clarkdale processor, the Intel H55 Express chipset provides an Intel FDI (Flexible Display Interface) bus, through which the chipset interacts with the built-in graphics core. Precisely because of the absence of such a bus in the Intel P55 Express chipset, it will not be possible to use the built-in graphics core in Clarkdale processors on boards with the Intel P55 Express chipset.

As already noted, on boards with the Intel H55 Express chipset there can only be one PCI Express x16 slot, that is, 16 PCI Express v.2.0 lanes supported by Clarkdale processors can be combined into only one PCI Express x16 slot. Accordingly, boards with the Intel H55 Express chipset cannot support NVIDIA SLI and ATI CrossFire modes.

Also integrated into the Intel H55 Express chipset is a 6-port SATA II controller. Moreover, this controller only supports AHCI mode and does not allow creating RAID arrays.

The Intel H55 Express chipset supports six PCI Express 2.0 lanes, which can be used by controllers integrated on the motherboard and to organize PCI Express 2.0 x1 and PCI Express 2.0 x4 slots.

Note also that the Intel H55 Express chipset already has a built-in MAC level of a gigabit network controller and a special interface (GLCI) for connecting a PHY controller.

The Intel H55 Express chipset also integrates a USB 2.0 controller. In total, the chipset supports 12 USB 2.0 ports.

Well, of course, the Intel H55 Express chipset has a built-in Intel HDA (High Definition Audio) audio controller, and to create a full-fledged audio system on the board, it is enough to integrate an audio codec, which will be connected via the HD Audio bus to the audio controller integrated into the chipset.

Another interesting feature of the Intel H55 Express chipset is its implementation of Intel QST (Intel Quiet System Technology). Actually, Intel QST technology itself is not new - it was first implemented in the Intel 965 Express chipset. To be more precise, the Intel 965 Express chipset provided for the possibility of hardware implementation of Intel QST technology. However, it cannot be said that this technology was popular among motherboard manufacturers. In fact, until now, none of the motherboard manufacturers (with the exception of Intel itself) have implemented this technology. Moreover, we can assume that Intel QST technology will not be implemented on boards based on the Intel H55 Express chipset, despite the theoretical possibility (except perhaps on boards from Intel itself).

Let us remind you that Intel QST is a technology for intelligent control of fan speed.

In short, Intel QST technology is designed to implement such an algorithm for controlling fan speed in order, on the one hand, to minimize the level of noise they create, and on the other, to ensure effective cooling.

Traditionally, the controller responsible for regulating the rotation speed of the processor cooler fan (Fan Speed ​​Control, FSC) is a separate chip (for example, manufactured by Winbond), which, receiving information about the processor temperature, controls the rotation speed of the processor cooler fan. As a rule, these are multifunctional microcircuits, and fan speed control is just one of the capabilities of such microcircuits. Such specialized microcircuits contain a built-in PWM controller and also allow you to dynamically change the voltage on the fan (for three-pin coolers). The algorithm by which the duty cycle of PWM pulses or the voltage on the fan changes is “stitched” into the controller itself. Motherboard manufacturers are responsible for programming FSC controllers.

An alternative method is to use a controller built into the chipset to control the fan speed rather than a separate specialized chip. Actually, this is what Intel QST technology is all about. However, the use of an FSC controller built into the chipset is not the only difference between Intel QST technology and traditional fan speed control technology based on a separate chip. The fact is that Intel QST technology implements a special PID algorithm that allows you to more accurately (compared to traditional methods) control the temperature of the processor or chipset, correlating it with a certain control temperature Tcontrol, which ultimately allows you to minimize the level of noise generated by the fans. In addition, Intel QST technology is fully programmable.

In order to describe Intel QST technology, let us recall that to monitor the temperature of processors, digital temperature sensors (Digital Temperature Sensor, DTS), which are an integral part of the processor, are used. The DTS sensor converts an analog voltage value into a digital temperature value, which is stored in the processor's internal software-accessible registers.

The digital value of the processor temperature is available for reading via the PECI (Platform Environment Control Interface) interface. Actually, DTS sensors together with the PECI interface represent a single solution for thermal monitoring of processors.

The PECI interface is used by the FSC (Fan Speed ​​Control) controller to control the fan speed.

The main component of Intel QST technology is a PID controller (Proportional-Integral-Derivative), whose task is to select the desired duty cycle of PWM pulses (or supply voltage) based on data about the current processor temperature.

The operating principle of the PID controller is quite simple. The input data of the PID controller is the current process temperature (for example, the temperature of the processor or chipset) and some predefined control temperature Tcontrol. The PID controller calculates the difference (error) between the current temperature and the control temperature and, based on this difference, as well as the rate of its change and knowledge of the value of the difference at previous points in time, using a special algorithm, it will calculate the necessary change in the duty cycle of PWM pulses required to minimize the error. That is, if we consider the difference between the current and control temperatures as a time-dependent error function e(t), then the task of the PID controller is to minimize the error function or, more simply, to change the fan speed in such a way as to constantly maintain the processor temperature at the control level.

The main feature of the PID controller is precisely the fact that the algorithm for calculating the necessary changes takes into account not only the absolute value of the difference (error) between the current temperature and the control one, but also the rate of temperature change, as well as the value of errors at previous points in time. That is, the algorithm for calculating the necessary adjustments uses three components: proportional term (Proportional), integral (Integral) and differential (Derivative). The controller itself is named after these members: Proportional-Integral-Derivative (PID).

The proportional term takes into account the current difference (error) between the current and reference temperature values. The integral term takes into account the value of errors at previous times, and the differential term characterizes the rate of error change.

Proportional term P is defined as the product of the error e(t) at the current moment in time by a certain proportionality coefficient Kp:

P = K p e(t).

Coefficient Kp is a configurable characteristic of the PID controller. The higher the coefficient value Kp, the greater will be the change in the controlled characteristic for a given error value. Values ​​too high Kp lead to system instability, and too low values Kp- insufficient sensitivity of the PID controller.

Integral term I characterizes the accumulated sum of errors over a certain time interval, that is, it takes into account, as it were, the prehistory of the development of the process. The integral term is defined as the product of the coefficient K i to the integral of the error function over time:

Coefficient K i is a configurable characteristic of the PID controller. The integral term, together with the proportional term, allows you to speed up the process of minimizing the error and stabilize temperatures at a given level. At the same time, the large value of the coefficient K i can lead to fluctuations in the current temperature relative to the control one, that is, to the occurrence of temporary overheating (T>T control).

Differential term D characterizes the rate of temperature change and is defined as the derivative of the error function with respect to time, multiplied by the proportionality coefficient Kd

Coefficient Kd is a configurable characteristic of the PID controller. The differential term allows you to control the rate of change of the controlled characteristic of the PID controller (in our case, changing the duty cycle of PWM pulses or supply voltage) and thereby avoid the possibility of temporary overheating caused by the integral term. At the same time, increasing the value of the coefficient Kd also has negative consequences. The point is that the differential term is sensitive to noise and amplifies it. Therefore, the coefficient values ​​are too large Kd lead to system instability.

The block diagram of the PID controller is shown in Fig. 2.

Rice. 2. PID controller block diagram

The algorithm for calculating the required change in the duty cycle of PWM pulses as a response to an error occurs is quite simple:

PWM = –P –I + D.

It should be noted that the efficiency of the PID controller is determined by the optimal selection of coefficients Kp, K i And Kd. The task of configuring the PID controller (its firmware) using specialized Intel software is assigned to the motherboard manufacturer.

All we have to do is tell you how Intel QST technology is implemented at the hardware level. As we have already noted, this is a solution integrated into the chipset. The chipset contains a programmable ME (Memory Engine) unit, designed to develop a PID algorithm for temperature control, as well as an FSC unit, which contains PWM controllers and directly controls the fans.

In addition, the implementation of Intel QST technology also requires an SPI flash memory chip with sufficient space for the Intel QST technology firmware. Note that no separate flash memory chip with an SPI interface is required. The same SPI flash memory is used in which the system BIOS is flashed.

So, in conclusion, we once again emphasize that Intel QST technology has a number of advantages over traditional fan speed control technologies, however, as we have already noted, it is not popular among motherboard manufacturers. The fact is that the traditional method of controlling fan speed uses separate microcircuits on motherboards. However, controlling the fan speed is only one of the functions of such microcircuits, and even if you do not use this particular function of the microcircuit, you still cannot refuse it. Well, if the chip still has to be integrated on the board, then why not assign it the function of controlling fans (since it is still present) and not bother with Intel QST technology?

Motherboard overview

ASRock H55DE3

The ASRock H55DE3 board based on the Intel H55 Express chipset turned out to be the only model in our review that is made in the ATX form factor. It can be positioned as a board for universal or multimedia PCs.

To install memory modules, the board has four DIMM slots, which allows you to install up to two DDR3 memory modules per channel (in dual-channel memory mode). In total, the board supports up to 16 GB of memory, and it is optimal to use two or four memory modules with it. In normal operation, the board is designed for DDR3-1333/1066 memory, and in overclocking mode the manufacturer claims support for DDR3-2600/2133/1866/1600 memory. Of course, you should not assume that in overclocking mode any memory labeled as DDR3-2600/2133/1866/1600 will work on the ASRock H55DE3 board. In this case, not everything depends on the board itself. After all, the main thing is whether the memory controller integrated into the processor can support its operation at such a speed. Consequently, the ability of memory to operate in overclocking mode largely depends on the specific processor instance.

If you use the graphics core built into the Clarkdale processor, connecting the monitor to the ASRock H55DE3 board is possible via VGA, DVI-D and HDMI interfaces.

In addition, the board has another slot of the PCI Express 2.0 x16 form factor, which operates at x4 speed and is implemented through four PCI Express 2.0 lanes supported by the Intel H55 Express chipset. This slot is optimally used for installing expansion cards, however, support for the ATI CrossFire mode is also declared when installing a second video card in the second slot with the PCI Express 2.0 x16 form factor. Naturally, to implement the ATI CrossFire mode, both video cards must have ATI GPUs.

As for the advisability of using two video cards in ATI CrossFire mode on the ASRock H55DE3 board, the same can be said here as regarding a similar solution on the Gigabyte H55M-UD2H board. That is, firstly, you need to remember that the ASRock H55DE3 board does not belong to the gaming category, for which the ability to combine video cards is relevant, and secondly, you need to take into account that the second slot with the PCI Express 2.0 x16 form factor operates at x4 speed, and communication between the two video cards occurs via the DMI bus, which connects the chipset to the processor, which, of course, negatively affects the performance of the graphics subsystem in ATI CrossFire mode.

In addition to the PCI Express 2.0 x16 slot running at x4 speed, the ASRock H55DE3 board has two traditional PCI 2.2 slots and one PCI Express 2.0 x1 slot.

To connect internal hard drives and optical drives, the ASRock H55DE3 board provides four SATA II ports, which are implemented via a controller integrated into the Intel H55 Express chipset. To connect external drives, there are two more eSATA ports, which are also implemented through a controller integrated into the chipset. Let us remind you that the SATA controller of the Intel H55 Express chipset does not support the ability to create RAID arrays. eSATA ports have shared USB connectors, which is very convenient because there is no need to additionally connect an external drive with an eSATA interface to the USB connector to provide power.

In addition, the board integrates a Winbond W83667HG controller, through which a serial port and a PS/2 port are implemented. It is also responsible for monitoring the supply voltage and controlling the fan speed.

To connect a variety of peripheral devices, the ASRock H55DE3 board has 12 USB 2.0 ports. Six of them are output to the rear panel of the board (two ports are combined with eSATA ports), and the remaining six can be output to the rear side of the PC by connecting the corresponding dies to three connectors on the board (two ports each).

The audio subsystem of this motherboard is based on the VIA VT1718S audio codec, and on the back of the motherboard there are five mini-jack audio connectors and one S/PDIF optical connector (output).

The board also integrates a Realtek RTL8111D gigabit network controller.

If we count the number of controllers integrated on the ASRock H55DE3 board that use PCI Express 2.0 lanes, and also take into account the presence of a PCI Express 2.0 x4 slot (in the PCI Express 2.0 x16 form factor) and a PCI Express 2.0 x1 slot, we get that all six PCI lanes are used Express 2.0 supported by the Intel H55 Express chipset. Four of them are used to organize a PCI Express 2.0 x4 slot (in the PCI Express 2.0 x16 form factor), another line is used to organize a PCI Express 2.0 x1 slot, and the remaining line is used to connect the Realtek RTL8111D controller. All other controllers integrated on the board do not use the PCI Express bus.

The board's cooling system consists of one heatsink based on the Intel H55 Express chipset.

To connect fans, the ASRock H55DE3 board has one four-pin and two three-pin connectors. The four-pin one is for connecting the processor cooler, and the three-pin one is for additional case fans.

The ASRock H55DE3 board uses a 5-phase (4+1) switching voltage regulator for the processor, based on the ST Microelectronics ST L6716 four-phase PWM controller. This controller combines three MOSFET drivers, and in addition, another ST L6741 MOSFET driver is used. This controller supports technology for dynamic switching of the number of power phases (two, three or four power phases).

In addition, the board contains a single-phase PWM controller ST L6716 from STMicroelectronics with an integrated MOSFET driver, which is apparently used to organize the power supply circuit for the graphics controller and memory controller built into the processor.

The options for customizing the BIOS of the ASRock H55DE3 board are quite wide, which is typical for all ASRock boards. It is possible to overclock the processor both by changing the multiplication factor (in the range from 9 to 26 for the Intel Core i5-661 processor) and by changing the reference frequency in the range from 100 to 300 MHz. Memory can also be overclocked by changing the divider value or the reference frequency.

By changing the value of the divider, you can set the memory frequency to 800, 1066 or 1333 MHz (with a reference frequency of 133 MHz).

Naturally, it is possible to change memory timings, supply voltage and much more.

To control the rotation speed of the processor cooler fan, the BIOS settings provide the CPU FAN Setting menu. The CPU FAN Setting parameter can be selected as Automatic Mode or Full On. When you select Full On, the cooler will always spin at maximum speed, regardless of the processor temperature, and when Automatic Mode is selected, two more parameters become available: Target CPU Temperature and Target FAN Speed. Unfortunately, the Target CPU Temperature parameter is not described anywhere in the documentation. Moreover, despite the declared possibility of changing this parameter in the range from 45 to 65 °C, it does not change - its value is 50 °C.

The Target FAN Speed ​​parameter allows you to select one of nine operating modes of the processor cooler, which are designated as Level 1, Level 2, etc. What is known about these operating modes is that a higher level corresponds to a higher rotation speed of the processor cooler fan.

It would be natural to assume that the difference between speed modes lies in the minimum processor temperature, upon reaching which the duty cycle of PWM pulses begins to change.

However, during testing it turned out that the different operating modes of the cooler do not depend in any way on the processor temperature and only determine the duty cycle of PWM pulses, which does not depend on the processor temperature. So, Level 1 mode corresponds to a duty cycle of 10%, Level 2 mode - 20%, etc. in 10% increments. That is, we can state that the technology for intelligent control of the rotation speed of the processor cooler fan on the ASRock H55DE3 board is not implemented at all. In passing, we note that the same drawback is also characteristic of other AsRock boards.

The ASRock H55DE3 board comes bundled with several proprietary utilities. In particular, the ASRock OC Tuner utility is designed to overclock the system in real time. It allows you to change the system bus frequency, multiplier, and processor supply voltage. In addition, this utility provides system monitoring and changing the rotation speed of the processor cooler fan (by changing the value of the Target FAN Speed ​​parameter).

The ASRock H55DE3 board has only one BIOS chip and no BIOS recovery features, which, of course, makes it vulnerable and the update procedure unsafe. The procedure for flashing the BIOS on the ASRock H55DE3 board is quite simple using proprietary ASRock Instant Flash technology, which allows you to start the BIOS update process from flash media before booting the system.

ASUS P7H55-M PRO

The ASUS P7H55-M PRO board based on the Intel H55 Express chipset has a microATX form factor and is aimed at home universal or multimedia PCs.

To install memory modules, the board has four DIMM slots, which allows you to install up to two DDR3 memory modules per channel (in dual-channel memory mode). In total, the board supports installation of up to 16 GB of memory (chipset specification), and it is optimal to use two or four memory modules with it. At the same time, the manufacturer claims support not only for memory at standard frequencies (DDR3-1333/1066), but also for faster memory up to DDR3-2133. However, as we have already noted, the possibility of using memory in overclocking mode depends not only on the board itself, but also on the specific processor instance into which the memory controller is integrated.

To install a video card, the board has a PCI Express 2.0 x16 slot, which is implemented through 16 PCI Express 2.0 lanes supported by Lynnfield and Clarkdale processors. When using the graphics core built into the Clarkdale processor, connecting a monitor is possible via VGA, DVI-D or HDMI interfaces, the connectors of which are located on the rear of the board.

In addition, the board has another PCI Express 2.0 x1 slot, which is implemented through one of the six PCI Express 2.0 lanes supported by the Intel P55 Express chipset. The ASUS P7H55-M PRO board also has two traditional PCI slots.

To connect drives, the ASUS P7H55-M PRO board has six SATA II ports, which are implemented through the controller built into the Intel HP55 Express chipset and do not support the ability to create RAID arrays.

To connect a variety of peripheral devices, the ASUS P7H55-M PRO board has 12 USB 2.0 ports (the Intel H55 Express chipset supports a total of 12 USB 2.0 ports). Six of them are output to the rear panel of the board, and six more can be output to the back of the PC by connecting the corresponding dies to three connectors on the board (two ports per die).

The audio subsystem of the ASUS P7H55-M PRO board is based on a 10-channel Realtek ALC889 audio codec, providing a signal-to-noise ratio of 108 and 104 dB (ADC), as well as playback and recording of 24 bit/192 kHz on all channels. Accordingly, on the back side of the motherboard there are six mini-jack audio connectors and one S/PDIF optical connector (output).

The board also integrates a Realtek RTL8112L gigabit network controller, which uses one PCI Express 2.0 line, and a Winbond W83667HG-A controller, which implements a serial port and a PS/2 port. The same controller is responsible for monitoring the supply voltage and controlling the fan speed.

If we count the number of controllers integrated on the ASUS P7H55-M PRO board that use PCI Express 2.0 lines, and also take into account the presence of a PCI Express 2.0 x1 slot, it turns out that out of the six lines supported by the Intel H55 Express chipset, only three are used (PCI Express slot 2.0 x1, JMicron JMB368 and Realtek RTL8112L controllers), while others remain unoccupied.

The cooling system of the ASUS P7H55-M PRO board is quite simple: one heatsink is installed on the chipset, and another decorative one is installed on the MOSFET transistors of the processor supply voltage regulator. Moreover, not all MOSFET transistors are covered by a heatsink, but only six out of 12. In addition, the board has two four-pin and one three-pin connector for connecting fans.

To configure fan speed control modes, the BIOS menu provides several options. To set the CPU cooler fan speed control mode, you first need to specify the Enable value for the CPU Q-Fan Control parameter. After this, you can select one of four control modes (CPU Fan Profile) for the processor cooler fan - Standard, Silent, Turbo or Manual.

When studying the implementation of fan speed control, it turned out that for Silent and Standard modes, the minimum duty cycle of PWM control pulses is 20%. The difference between the Silent and Standard modes lies in the temperature range in which a dynamic change in the duty cycle of the PWM signal is realized.

Thus, for the Silent mode, as the processor temperature increases, the duty cycle of PWM control pulses changes only in the temperature range from 53 to 80 °C, that is, up to 53 °C, the duty cycle of PWM pulses does not change and is 21%. With a further increase in processor temperature, the duty cycle of the pulses begins to increase smoothly, reaching 100% at 80 °C. When the processor temperature decreases, the duty cycle of the control PWM pulses changes in the temperature range from 76 to 45 °C, that is, up to 76 °C, the duty cycle of the PWM pulses does not change and is 100%, and with a further decrease in the processor temperature it begins to gradually decrease, reaching values ​​of 20% at a processor temperature of 45 ° C.

For the Standard mode, the duty cycle of PWM control pulses changes in the temperature range from 45 to 69 °C as the temperature increases and in the range from 66 to 37 °C as the temperature decreases.

For Turbo mode, the minimum duty cycle of PWM control pulses is already 40%. As the processor temperature increases, the duty cycle of PWM control pulses changes in the temperature range from 40 to 60 °C, and when it decreases, from 57 to 35 °C.

In the Manual mode, the speed mode of the cooler is manually adjusted. In this mode, you need to set the upper value of the processor temperature in the range from 40 to 90 ° C and select for it the maximum duty cycle of PWM pulses in the range from 21 to 100%. In this case, when the processor temperature exceeds the set upper value, the duty cycle of PWM pulses will be the specified maximum value. Then you need to select the minimum value of the duty cycle of PWM pulses in the range from 0 to 100%, corresponding to the lower value of the processor temperature, which does not change and is 40 ° C. In this case, when the processor temperature is below 40 °C, the duty cycle of PWM pulses will be the selected minimum value. In the temperature range from 40 °C to the selected upper value, the duty cycle of PWM pulses will change in proportion to the change in processor temperature.

In addition to setting the operating modes of two four-pin fans via the BIOS, it is possible to program the fan speed using the ASUS AI Suite utility, supplied with the board, which allows for more fine-tuning.

This utility allows you to select one of the preset fan speed control profiles (Silent, Standard, Turbo, Intelligent, Stable), as well as create your own control profile (User). Different profiles differ from each other both in the minimum duty cycle of PWM pulses and in the temperature range in which the duty cycle changes. In the custom User profile, the user is given the opportunity to set the minimum and maximum duty cycle of PWM pulses and set the temperature range for changing the duty cycle of PWM pulses and even the rate of change in the duty cycle of PWM pulses within the selected temperature range at three points. The only limitation in this case is that the minimum duty cycle of PWM pulses cannot be lower than 21%, and the maximum processor temperature cannot exceed 74 ° C.

Another feature of the ASUS P7H55-M PRO board is the use of a 6-channel (4+2) switching voltage regulator.

Traditionally, ASUS boards use a circuit that includes a power phase control controller EPU2 ASP0800 and a 4-phase PWM controller PEM ASP0801 to control all power phases.

However, on the ASUS P7H55-M PRO board the processor voltage regulator circuit is arranged somewhat differently. To control all power phases, the same EPU2 ASP0800 controller is used, but paired with a 4-phase PWM controller RT8857 from Richtek Technology. The RT8857 PWM controller integrates two MOSFET drivers, and it also supports dynamic power phase switching technology.

Two more power channels are organized on the basis of a single-channel PWM controller APW1720.

Apparently, four power phases based on the RT8857 controller are used to organize the power supply circuit for the processor cores, and two more power channels based on the APW1720 controller are used to organize power supply to the memory controller and integrated graphics controller.

In conclusion, we note that the ASUS P7H55-M PRO board contains only one BIOS chip (although wiring is provided for installing a second chip). However, in the case of the ASUS P7H55-M PRO board this is not a problem. The fact is that this board supports ASUS CrashFree BIOS 3 BIOS backup recovery technology. The ASUS CrashFree BIOS 3 function automatically starts in the event of a BIOS crash or a checksum mismatch after a failed firmware update. At the same time, it looks for a BIOS image on a CD/DVD, USB flash drive or floppy disk. If a file is found on some media, the recovery procedure starts automatically.

The procedure for updating the BIOS on the ASUS P7H55-M PRO board is very simple. In principle, there are various ways to update the BIOS (including using a utility from under the loaded operating system), but the easiest way is to update the BIOS using a flash drive and the EZ Flash 2 function built into the BIOS. That is, you just need to enter the BIOS menu and select EZ Flash 2.

Naturally, the ASUS P7H55-M PRO board also implements various other proprietary ASUS technologies, and all the necessary utilities are included in the kit. In particular, the board has all sorts of means for overclocking the system. Thus, the ASUS GPU Boost function allows you to overclock the graphics controller integrated into the processor in real time by changing its frequency and supply voltage.

The ASUS Turbo Key feature allows you to redefine the computer's power button, making it a system overclocking button. After the appropriate settings, when you press the power button, the system will automatically accelerate without interrupting the operation of the PC.

To overclock a system based on the ASUS P7H55-M PRO board, you can also use the ASUS TurboV utility, which allows you to overclock in real time when the operating system is loaded and without the need to restart the PC.

ECS H55H-CM

The ECS H55H-CM board, made in the microATX form factor, can be positioned as an inexpensive solution for universal mid-level home computers or office PCs.

To install memory modules, the board has four DIMM slots, which allows you to install up to two DDR3 memory modules per channel (in dual-channel memory mode). In total, the board supports installation of up to 16 GB of memory (chipset specification), and it is optimal to use two or four memory modules with it. In normal operation, the board is designed for DDR3-1333/1066/800 memory.

To install a video card, the board has a PCI Express 2.0 x16 slot, which is implemented using 16 PCI Express 2.0 lanes, supported by Clarkdale and Lynnfield processors. When using the graphics core built into the Clarkdale processor, connecting a monitor is possible via VGA or HDMI interfaces, the connectors of which are located on the rear plate of the board.

In addition, the ECS H55H-CM board has two more PCI Express 2.0 x1 slots, implemented through two PCI Express 2.0 lanes supported by the Intel H55 Express chipset, as well as one traditional PCI slot.

To connect hard drives and optical drives, the ECS H55H-CM board has six SATA II ports, which are implemented using the controller integrated into the Intel P55 Express chipset and do not support the ability to create RAID arrays.

To connect a variety of peripheral devices, the board has 12 USB 2.0 ports. Six of them are output to the rear panel of the board, and the remaining six can be output to the rear side of the PC by connecting the corresponding dies to three connectors on the board (two ports each).

The board also has an Intel 82578DC gigabit network controller, which allows you to connect a PC based on this board to a local network segment to access the Internet.

The audio subsystem of the ECS H55H-CM board is built on the basis of a six-channel Realtek ALC662 audio codec, and three mini-jack audio connectors are installed on the back of the board.

In addition, the board has connectors for connecting two serial ports, which are implemented on two UTC 75232L chips.

The board also has a connector for connecting a 3.5-inch floppy drive, and a parallel port is located on the back of the board. Note that parallel and serial ports, and a connector for connecting a 3.5-inch floppy drive are practically no longer used in home PCs and can only be in demand in office computers, and even then in rare cases.

The board's cooling system includes only one heatsink based on the Intel H55 Express chipset.

In addition, the board has a four-pin connector for connecting a processor cooler fan and a three-pin connector for connecting an additional case fan.

The ECS H55H-CM board uses a 5-phase (4+1) switching voltage regulator for the processor supply. The processor supply voltage regulator is based on the ON Semiconductor NCP5395T 4-phase PWM controller, which also includes MOSFET drivers. This controller supports technology for dynamic switching of the number of power phases (two, three or four power phases).

In addition, the board contains an NCP5380 single-phase PWM controller with an integrated MOSFET driver, which apparently serves to organize the power supply circuit for the graphics controller built into the processor and, possibly, the memory controller.

As you can see, the processor power supply circuits on the ECS H55H-CM and Intel DH55TC boards are similar. And in general, in terms of its functionality, the ECS H55H-CM board is very similar to the Intel DH55TC board.

As for the BIOS functionality on the ECS H55H-CM board, its overclocking capabilities are very limited. You can, for example, change the system bus frequency and the processor clock multiplier (in the range from 9 to 25 for the Intel Core i5-661 processor), but you cannot change the supply voltage. The same goes for memory. You can set the memory frequency value by changing the divider (800, 1066, 1333 or 1600 MHz with a system bus frequency of 133 MHz), and also change the memory timings, but you cannot change the memory supply voltage.

To control the rotation speed of the processor cooler fan, the BIOS settings provide a Smart Fan Function menu with the ability to fine-tune the speed mode of the processor cooler.

When you set the value of the CPU SMART FAN Control parameter to Enable, you can select one of three (Quite, Silent, Normal) preset operating modes of the processor cooler or configure the cooler operating mode manually. For each of the three speed modes of the cooler, the following parameters are set:

• CPU SMART Fan start PWM;
• SMART Fan start PWM TEMP (-);
• Delta T;
• SMART Fan Slope PWM Value.

When setting the speed mode of the cooler manually, you need to set the value of each of the named parameters. Alas, their values ​​are not commented on anywhere, which, of course, makes it difficult to independently configure the cooler’s operating mode. Only armed with an oscilloscope and a utility for testing coolers, we were able to understand the meaning of these parameters.

The CPU SMART Fan start PWM parameter sets the minimum duty cycle of PWM control pulses for the processor cooler fan.

The SMART Fan start PWM TEMP (-) parameter determines the difference between the current and critical processor temperature, upon reaching which the duty cycle of PWM pulses begins to change.

The SMART Fan Slope PWM Value parameter specifies the rate of change in the duty cycle of PWM pulses - by what percentage does the duty cycle of PWM pulses change when the processor temperature changes by 1 °C.

The only parameter that we could not identify was Delta T. However, despite this, having experimented with various options for setting the speed mode of the processor cooler, we concluded that this implementation of the cooler rotation speed control system is very effective and allows you to create both very quiet PCs and high-performance computers with an effective processor cooling system.

In conclusion, we note that the ECS P55H-A board comes with the eJIFFY utility, which is a stripped-down version of the Linux-like operating system. This utility is installed on the PC’s hard drive and, when the computer boots, it allows you to quickly load not a full-fledged operating system, but a lightweight version of it and get quick access to some applications from under it. Actually, the idea is not new, and ASUS has been using it for a long time. The advantage of this solution lies only in the loading speed of a stripped-down version of the operating system, but the relevance of this solution is very doubtful. In addition, it is worth considering that the Linux-like operating system has only an English interface.

We also note that the ECS H55H-CM board, like the Intel DH55TC board, uses only one BIOS chip and does not provide BIOS emergency recovery tools, which, of course, makes it vulnerable and its update procedure unsafe. However, this procedure is quite complicated on all ECS boards. First you need to download the utility for flashing the BIOS from the manufacturer's website. Moreover, each BIOS type (AMI, AFU, AWARD) uses its own version of the utility. Flashing the BIOS is possible both from the Windows operating system and using bootable media with the DOS operating system, and each flashing option uses its own version of the utility. You can begin the BIOS flashing procedure only after reading the instructions. In general, everything is complicated and unsafe.

Gigabyte GA-H55M-UD2H

The Gigabyte H55M-UD2H board based on the Intel H55 Express chipset can be positioned as a board for inexpensive home universal or multimedia PCs. It is made in microATX format and can be placed in a compact multimedia case.

To install memory modules, the board has four DIMM slots, which allows you to install up to two DDR3 memory modules per channel (in dual-channel memory mode). In total, the board supports installation of up to 16 GB of memory (chipset specification), and it is optimal to use two or four memory modules with it. In normal operation, the board is designed for DDR3-1333/1066/800 memory, and in overclocking mode it also supports DDR3-1666 memory.

If you use the graphics core built into the Clarkdale processor, connecting the monitor is possible via VGA, DVI-D, HDMI or DisplayPort interfaces.

To install a discrete video card, the board has one PCI Express 2.0 x16 slot, which is implemented through 16 PCI Express 2.0 lanes supported by Clarkdale and Lynnfield processors.

In addition, the board has another slot of the PCI Express 2.0 x16 form factor, which is implemented through four PCI Express 2.0 lanes supported by the Intel H55 Express chipset and operates at x4 speed. Formally, it can be used to install a second discrete video card, and in the case of using video cards on ATI GPUs, support for the ATI CrossFire mode is declared. However, the feasibility of such a solution is rather doubtful. Firstly, the Gigabyte H55M-UD2H board is by no means a gaming solution. Secondly, you need to take into account that the second slot with the PCI Express 2.0 x16 form factor operates at x4 speed, and communication between the two video cards will occur via the DMI bus connecting the chipset to the processor, which, of course, will negatively affect the ATI CrossFire mode. and therefore the presence of two PCI Express 2.0 x16 slots on the Gigabyte H55M-UD2H board is more of a marketing ploy than a sought-after necessity.

For installing additional expansion cards, the board also contains two more traditional PCI 2.2 slots.

To connect hard drives and optical drives, the Gigabyte H55M-UD2H board has six SATA II ports, implemented through a controller integrated into the Intel H55 Express chipset. Let us remind you that this SATA controller does not support the ability to create RAID arrays.

Five SATA II ports are designed for connecting internal hard drives and optical drives, and one port is made in the eSATA connector and is located on the rear panel of the board.

The board also integrates a JMicron JMB368 controller, through which an IDE connector is implemented (ATA-133/100/66/33 interface). It can be used to connect optical drives or hard drives with this legacy interface.

In addition, the board also integrates the iTE IT8720 controller, through which a connector for connecting a 3.5-inch floppy drive, as well as a serial port and a PS/2 port are implemented. The same controller is responsible for monitoring the supply voltage and controlling the fan speed.

To connect a variety of peripheral devices, the Gigabyte H55M-UD2H board has 12 USB 2.0 ports, six of which are located on the rear panel of the board, and the remaining six can be output to the back of the PC by connecting the corresponding dies to three connectors on the board (two ports for each ).

The board also includes a T.I. FireWire controller. TSB43AB23, through which two IEEE-1394a ports are implemented, one of which is located on the rear panel of the board, and a corresponding connector is provided for connecting the second.

The audio subsystem of this motherboard is based on a 10-channel (7.1+2) Realtek ALC889 audio codec. Accordingly, on the back side of the motherboard there are six mini-jack audio connectors and an S/PDIF optical connector (output), and on the board itself there are S/PDIF input and S/PDIF output connectors.

In addition, the board integrates a Realtek RTL8111D gigabit network controller.

If we count the number of controllers integrated on the Gigabyte H55M-UD2H board that use PCI Express 2.0 lanes, and also take into account the presence of a PCI Express 2.0 x4 slot (in the PCI Express 2.0 x16 form factor), we get that all six PCI Express 2.0 lanes supported are used Intel H55 Express chipset. Four of them are used to organize a PCI Express 2.0 x4 slot (in the PCI Express 2.0 x16 form factor), and two more are used to connect JMicron JMB368 and Realtek RTL8111D controllers. All other controllers integrated on the board do not use the PCI Express bus.

The cooling system of the Gigabyte H55M-UD2H board is very simple and consists of one heatsink based on the Intel H55 Express chipset.

To connect fans, the Gigabyte H55M-UD2H board has two four-pin connectors, one of which is designed to connect a processor cooler, and the second is for connecting an additional case fan.

Unfortunately, the documentation for the Gigabyte H55M-UD2H board does not say anything about the organization of the processor power system. But understanding the circuitry of the switching supply voltage regulator used turned out to be very difficult. A detailed examination of the board allows us to make the following assumption. To power the processor cores, a 4-phase switching voltage regulator is used, built on the basis of the Intersil ISL6334 control chip in combination with three Intersil ISL6612 MOSFET drivers and one Intersil ISL6622 driver. Note that the Intersil ISL6334 controller supports dynamic power phase switching technology to optimize the efficiency of the voltage regulator.

In addition, the board has two more control controllers: Intersil ISL6322G and Intersil ISL6314, the first of which is two-phase with integrated MOSFET drivers, and the second is single-phase with an integrated MOSFET driver. Apparently, one of them is used in the power circuit of the memory controller built into the processor, and the second is used in the power circuit of the graphics core.

The options for customizing the BIOS of the Gigabyte H55M-UD2H board are quite functional, which is typical for all Gigabyte boards. It is possible to overclock the processor both by changing the multiplication factor (in the range from 9 to 26 for the Intel Core i5-661 processor) and by changing the reference frequency (in the range from 100 to 600 MHz). Memory can also be overclocked by changing the divider value or the reference frequency. Naturally, it is possible to change memory timings, supply voltage and much more.

The Gigabyte H55M-UD2H board comes with a proprietary Easy Tune 6 utility, designed to overclock system components. With its help you can overclock the processor, memory and discrete video card. The processor is overclocked by changing the system bus frequency in the range from 100 to 333 MHz in 1 MHz steps. You can also change the memory frequency, and the range of changes in the memory frequency depends on the set value of the system bus frequency. In addition, you can change the PCI Express bus frequency in the range from 89 to 150 MHz in 1 MHz steps, as well as the supply voltage of various system components. In general, this utility, in its functionality, largely replicates the capabilities of the BIOS for overclocking the system, but its use does not require rebooting the system every time. The only thing that the Easy Tune 6 utility does not allow is changing memory timings, as well as overclocking the graphics controller built into the processor. The advantages of this utility include the ability to save created overclocking profiles and, if necessary, load them.

Another undeniable advantage of this utility is the ability to configure the speed mode of the processor cooler fan. To control its rotation speed, the board's BIOS settings provide the CPU Smart Fan Control option. When you select Enable for this option, the rotation speed of the processor cooler fan is dynamically changed depending on its current temperature. True, there are no settings for the fan speed mode in this case.

Using the Easy Tune 6 utility, you can set the correspondence between the temperature range of the processor and the range of changes in the duty cycle of PWM pulses. The minimum duty cycle of PWM pulses can be set to 10% and tied to a certain processor temperature value. That is, if the processor temperature is less than the set value, the duty cycle of PWM pulses will be 10%. Similarly, the maximum duty cycle of PWM pulses can be set to 100% and tied to a certain value of the processor temperature so that at temperatures exceeding the set value, the duty cycle of PWM pulses will be 100%. Well, when the processor temperature is in the range between two specified values, the duty cycle of PWM pulses will change in proportion to the temperature change.

In general, it should be noted that the implementation of fan speed control through the Easy Tune 6 utility is very successful and functional. It allows you to configure coolers for both quiet multimedia PCs and overclocked computers.

Also note that the Gigabyte H55M-UD2H board contains two BIOS chips (proprietary DualBIOS technology), that is, there is a main and backup BIOS chip. In normal operation, the main BIOS is used, but in the event of an emergency (when an incorrect BIOS is flashed or a failure occurs during flashing), the backup BIOS is used, automatically copied to the main chip. Thus, the BIOS on the Gigabyte H55M-UD2H board is almost impossible to “kill,” and the procedure for flashing the BIOS is very simple using proprietary Gigabyte utilities or even a special BIOS option.

Intel DH55TC

The Intel DH55TC board, made in the microATX form factor, can be positioned as a board for the mass market of inexpensive home PCs or as a board for the corporate segment of the market.

The board has four DIMM slots for installing memory modules. In total, the board supports up to 16 GB of memory (chipset specification). In normal operation it is designed for DDR3-1333/1066 memory.

To install a video card, the board has a PCI Express 2.0 x16 slot, which is implemented using 16 PCI Express 2.0 lanes supported by Clarkdale and Lynnfield processors. If you use the graphics core built into the Clarkdale processor, connecting the monitor is possible via VGA, DVI-D or HDMI interfaces.

In addition, the Intel DH55TC board has two more PCI Express 2.0 x1 slots and one traditional PCI slot.

To connect hard drives and optical drives, the Intel DH55TC board has six SATA II ports, implemented using a controller integrated into the Intel P55 Express chipset and does not support the ability to create RAID arrays.

To connect a variety of peripheral devices, the board has 12 USB 2.0 ports, six of which are located on the rear panel of the board, and the others can be output to the back of the PC by connecting the corresponding dies to three connectors on the board (two ports each).

The board also has an Intel 82578DC gigabit network controller, which allows you to connect a PC based on this board to a local network segment to access the Internet.

The audio subsystem of the Intel DH55TC board is built on the basis of the Realtek ALC888 audio codec with support for eight-channel (5.1+2) audio, and on the back of the board there are three mini-jack audio connectors.

In addition, the board has connectors for connecting serial and parallel ports, which are implemented on the basis of the Winbond W83627DHG multifunctional I/O chip.

Note that in addition to supporting serial and parallel ports, the Winbond W83627DHG chip allows you to control the supply voltage and control the fan speed, but the Intel DH55TC board uses Intel QST technology to control the fan speed.

The board's cooling system is implemented quite simply and consists of just one heatsink based on the Intel H55 Express chipset. In addition, the board has three four-pin fan connectors, one of which is designed to connect the processor cooler.

The Intel DH55TC board uses a 5-phase switching voltage regulator. The processor supply voltage regulator is based on the ON Semiconductor NCP5395T 4-phase PWM controller, which also includes MOSFET drivers. This controller supports technology for dynamic switching of the number of power phases (two, three or four power phases). In addition, the board contains an NCP5380 single-phase PWM controller with an integrated MOSFET driver, which is apparently used to organize the power supply circuit for the graphics controller built into the processor, and possibly the memory controller.

As for the options for customizing the BIOS of the Intel DH55TC board, there are practically none. In fact, the board uses the same BIOS capabilities as those on regular laptops. The BIOS of the Intel DH55TC board does not provide for setting the fan speed control mode, as well as overclocking the processor and RAM. Let's make a reservation right away that we are talking about BIOS version TCIBX10H.86A.0023. To make sure that the problem only affects a specific BIOS version, we decided to update it, and at the same time check how easy it is to flash the BIOS on the Intel DH55TC board.

On the manufacturer's website you can download a new version of the BIOS, integrated with its installation utility. Actually, the flashing procedure is very simple: we run the BIOS flashing utility from under the Windows 7 operating system and just wait for the result. The computer should reboot itself and begin the flashing procedure. However, at the last stage we were completely disappointed. Despite the message that the BIOS flashing procedure was successfully completed, with the new BIOS version the board stopped booting at all. Alas, further testing of it became impossible. Note that the Intel DH55TC board does not have a copy of the BIOS and does not provide any means for emergency BIOS recovery (for motherboards from other manufacturers, various means for emergency BIOS recovery have long been available). So, if the BIOS flashing is unsuccessful, it will be impossible to revive this board on your own, which is one of its most serious drawbacks.

MSI H55M-E33

The MSI H55M-E33 can be positioned as a board aimed at the mass segment of universal home or multimedia PCs. Like most boards based on the Intel H55 Express chipset, it is made in the microATX form factor.

The board has four DIMM slots for installing memory modules. In total, it supports up to 16 GB of memory (chipset specification). In normal operation, the board is designed for DDR3-1333/1066/800 memory, and in overclocking mode it also supports DDR3-1600 memory.

To install a video card, the board has a PCI Express 2.0 x16 slot, which is implemented using 16 PCI Express 2.0 lanes, supported by Lynnfield and Clarkdale processors. If you use the graphics core built into the Clarkdale processor, connecting the monitor is possible via VGA, DVI-D and HDMI interfaces, the connectors of which are located on the rear of the board.

In addition, the board has two more PCI Express 2.0 x1 slots, which are implemented through two of the six PCI Express 2.0 lanes supported by the Intel H55 Express chipset. The MSI H55M-E33 board also has a traditional PCI slot.

To connect drives, the MSI H55M-E33 board has six SATA II ports, which are implemented through the controller built into the Intel HP55 Express chipset and do not support the ability to create RAID arrays.

The board also integrates a JMicron JMB368 controller, through which an IDE connector (ATA-133/100/66/33 interface) is implemented, which can be used to connect optical drives or hard drives with this outdated interface.

To connect a variety of peripheral devices, the MSI H55M-E33 board has 12 USB 2.0 ports, six of which are located on the rear panel of the board, and the rest can be output to the back of the PC by connecting the corresponding plugs to three connectors on the board (two ports per plug ).

The board's audio subsystem is based on a 10-channel (7.1+2) Realtek ALC889 audio codec. Accordingly, on the back side of the motherboard there are six mini-jack audio connectors.

The board also contains a Realtek RTL 8111DL gigabit network controller for connecting a PC to a local network segment (for example, to access the Internet).

In addition, the board has two connectors for connecting serial ports and a connector for connecting a parallel port. These ports are implemented through the Fintek F71889F chip, which is also responsible for monitoring voltages and controlling fan speed.

Note that of the six PCI Express 2.0 lanes supported by the Intel H55 Express chipset, only three are used on the board: two lines for two PCI Express 2.0 x1 slots, and one more for the Realtek RTL 8111DL controller.

The board's cooling system is based on a miniature radiator installed on the Intel P55 Express chipset. In addition, the board has two three-pin (SYS_FAN1, SYS_FAN2) and one four-pin (CPU_FAN) fan connectors. The four-pin one is for connecting the CPU cooler fan, and the three-pin one is for additional fans.

The switching voltage regulator for the processor supply on the MSI H55M-E33 board is unconventional for MSI boards. As a rule, MSI boards use a supply voltage regulator made using DrMOS technology, which involves combining two MOSFET transistors and a driver chip for switching these transistors within one DrMOS chip (hence the name of this technology: DrMOS means Driver+MOSFET). However, on the MSI H55M-E33 board the five-phase (4+1) processor supply voltage regulator is made according to a traditional design.

The processor supply voltage regulator is based on the uP6206 4-phase control controller from uPI Semiconductor with integrated MOSFET drivers. This controller supports technology for dynamic switching of the number of power phases.

In addition, the board contains an Intersil ISL8314 single-phase PWM controller with an integrated MOSFET driver, which is apparently used to organize the power circuit for the graphics controller and memory controller built into the processor.

Naturally, the four-phase processor supply voltage regulator supports APS (Active Phase Switching) technology, which allows you to minimize system power consumption by dynamically switching the number of active phases depending on the current processor load.

As for the BIOS features of the MSI H55M-E33 board, it is worth paying attention to two circumstances. Firstly, the BIOS on the board provides various means for overclocking the system, and secondly, it is possible to fine-tune the speed mode of the processor cooler fan.

In particular, the BIOS of the MSI H55M-E33 board allows you to overclock the processor not only in the traditional way by changing the system bus frequency, but also in a semi-automatic mode, when you set the initial system bus frequency, the desired maximum system bus frequency and the number of system bus overclocking stages. In this case, when the system starts, the system bus frequency will automatically accelerate from the specified initial value to the maximum possible value (not exceeding the set maximum frequency).

Another possibility for overclocking the processor provided in the BIOS is the fully automatic overclocking mode of the system bus frequency, when when the system boots, the maximum possible system bus frequency is automatically detected and set.

In general, it should be noted that the MSI H55M-E33 board has no equal in terms of overclocking capabilities - everything is very functional and well thought out.

To control the rotation speed of three-pin fans, you can set the following supply voltage values ​​in the BIOS settings: 100% (12 V), 75% (9 V) and 50% (6 V). The CPU cooler fan speed is adjusted as follows. The board's BIOS specifies a temperature threshold (CPU Smart Fan Target), upon reaching which the fan rotation speed will increase from the minimum to the maximum value. The temperature threshold can be selected from 40 to 70 °C in 5 °C increments. In addition, it is possible to set the minimum fan speed (CPU Min. FAN Speed) as a percentage in the range from 0 to 87.5% in steps of 12.5%.

During testing of the board, it turned out that the minimum fan rotation speed, set as a percentage, is nothing more than the duty cycle of the PWM control pulses supplied to the fan.

The MSI H55M-E33 board comes with a disk with all the necessary drivers and proprietary utilities. In particular, the MSI Control Center utility allows you to monitor the system status (supply voltage, fan speed, processor clock speed, etc.), as well as in real time (without rebooting the operating system) change the system bus frequency and supply voltage of various components system board.

In conclusion, the MSI H55M-E33 board has only one BIOS chip, so the BIOS update process is not safe. The procedure for flashing the BIOS is very simple - through the M-Flash option, which can be accessed through the BIOS. This option allows you to flash the BIOS using flash media. In addition, you can use the MSI Live Update utility, which makes it possible to check for new BIOS versions via the Internet on the technical support site, download them and update them while the operating system is loaded. This utility also allows you to check for new driver versions, which is very convenient.

Biostar TH55XE

The Biostar TH55XE board based on the Intel H55 Express chipset is made in the microATX form factor and belongs to the T-Series of Biostar boards designed for high-performance mass-market PCs.

To install memory modules, the board has four DIMM slots, which allows you to install up to two DDR3 memory modules per channel (in dual-channel memory mode). In total, the board supports installation of up to 16 GB of memory (chipset specification), and it is optimal to use two or four memory modules with it. In normal operation, the board is designed for DDR3-1333/1066/800 memory, and in overclocking mode it also supports DDR3-1600/2000 memory.

To install a discrete video card, the board has a PCI Express 2.0 x16 slot, which is implemented through 16 PCI Express 2.0 lanes supported by Lynnfield and Clarkdale processors.

If you use the graphics core built into the Clarkdale processor, connecting the monitor is possible via VGA, DVI-D or HDMI interfaces, the connectors of which are located on the rear plate of the board.

In addition, the board has a PCI Express 2.0 x4 slot, which is implemented through four of the six PCI Express 2.0 lanes supported by the Intel H55 Express chipset. The Biostar TH55XE board also has two traditional PCI slots.

To connect drives, the Biostar TH55XE board has five SATA II ports and one eSATA port (used to connect external drives), which are implemented through the controller built into the Intel HP55 Express chipset and do not support the ability to create RAID arrays.

The board also integrates a JMicron JMB368 controller, through which an IDE connector (ATA-133/100/66/33 interface) is implemented, which can be used to connect optical drives or hard drives with this interface.

To connect a variety of peripheral devices, the Biostar TH55XE board has ten USB 2.0 ports, four of which are located on the rear panel of the board, and the rest can be output to the back of the PC by connecting the corresponding dies to three connectors on the board (two ports for each).

The board also contains an LSI FW322 FireWire controller, through which two IEEE-1394a ports are implemented, one of which is located on the rear panel of the board, and a corresponding connector is provided for connecting the other.

The audio subsystem of this motherboard is based on a 10-channel (7.1+2) Realtek ALC888 audio codec, and there are six mini-jack audio connectors on the rear panel of the motherboard. In addition, the board itself has an S/PDIF connector (output) for connecting a coaxial port, and an optical S/PDIF connector is located on the back of the board.

The board also integrates a Realtek RTL8111DL gigabit network controller. In addition, there are connectors for connecting serial and parallel ports. These ports are implemented through the ITE IT8721F chip, which is also responsible for monitoring voltages and controlling fan speed.

Note that of the six PCI Express 2.0 lanes supported by the Intel H55 Express chipset, only five are used on the board: four for the PCI Express 2.0 x4 slot and one for the Realtek RTL 8111DL controller.

The cooling system of the Biostar TH55XE board consists of three radiators that are not connected to each other. Two heatsinks are used to cool the MOSFET transistors of the processor voltage regulator located near the LGA 1156 processor socket, and another one is installed on the Intel H55 Express chipset.

To connect fans, the Biostar TH55XE board has two three-pin and one four-pin connectors. The four-pin one is used to connect the processor cooler fan, and the three-pin one is for additional fans installed in the PC case.

The switching voltage regulator of the processor supply on the Biostar TH55XE board is six-channel (4+2). To power the processor cores, a 4-phase voltage regulator is used based on the uP6219 4-phase controller from uPI Semiconductor with three integrated MOSFET drivers and one external uP6281 MOSFET driver.

In addition, the board has another voltage regulator based on a two-phase uP6203 controller with two integrated MOSFET drivers, which is used to provide power to the memory controller and graphics core built into the processor.

Note that the uP6219 4-phase controller supports dynamic power phase switching technology to optimize the efficiency of the voltage regulator and, accordingly, reduce its power consumption.

Now let's look at the features of setting up the BIOS on the Biostar TH55XE board. In the BIOS settings, there is a Smart Fan Configuration option to control the fan speed. It should be noted that the implementation of fan speed control on the Biostar TH55XE board is exactly the same as on other Biostar boards (we have already seen such an implementation scheme, for example, on the Biostar TPOWER I55 board). However, if on the Biostar TPOWER I55 board the cooler control actually did not work, then on the Biostar TH55XE board everything functions properly.

In the Smart Fan Configuration menu, you can enable or disable the use of CPU cooler fan speed control. To enable this function, you must set the CPU Smart FAN parameter to Auto. Next, you need to carry out the cooler calibration procedure (Smart Fan Calibration) and select one of three control profiles (Control Mode): Performance, Quite or Manual.

As it turned out during testing, the Performance and Quite modes are generally the same thing. In these modes, if the difference between the critical and current processor temperature is more than 55 °C, the duty cycle of the PWM control pulses is zero. As soon as the difference between the critical and current processor temperature becomes less than 55 ° C, the duty cycle of WPM pulses begins to increase from 20% in proportion to the decrease in the difference between the critical and current processor temperature, reaching a value of 100% with a difference of 5 ° C.

When you select Manual mode, four additional setting options appear:

• FAN Ctrl OFF (°C);
• FAN Ctrl ON (°C);
• Fan Ctrl Start value;
• Fan Ctrl Sensitive.

For all of these parameters (except for the Fan Ctrl Start parameter), valid values ​​range from 1 to 127.

It was not so easy to understand the meaning of all these parameters, and the user manual will not help here. For example, as follows from the description in the user manual, the FAN Ctrl OFF parameter sets the processor temperature value, below which PWM control is disabled and the processor cooler fan rotates at minimum speed. The FAN Ctrl ON parameter specifies the processor temperature at which PWM control of the processor cooler fan speed is turned on. The Fan Ctrl Start value parameter sets the initial rotation speed of the processor cooler fan, and the Fan Ctrl Sensitive parameter sets the rate of change in the rotation speed of the processor cooler fan. There are a lot of illogical and incomprehensible things in this description of the values ​​for setting the speed mode of the processor cooler fan. For example, if FAN Ctrl OFF sets the processor temperature value below which PWM control is disabled, and FAN Ctrl ON sets the processor temperature value at which PWM control is turned on, then the question arises why they do not match and what will happen if you set FAN Ctrl OFF equal to 40 °C, and FAN Ctrl ON - 50 °C?

The value of the Fan Ctrl Start value parameter is also unclear. If this is the initial fan speed, what is it measured in? It would be logical to assume that the initial fan speed is set by the duty cycle of the PWM pulses, but the range of possible values ​​for this parameter is from 1 to 255, and the duty cycle cannot exceed 100%.

In addition, it is not clear in what units the rate of change in the fan rotation speed is set (apparently, this parameter determines the rate of change in the duty cycle of PWM pulses).

Only after arming ourselves with an oscilloscope and experimenting with various options for setting the manual mode for controlling the rotation speed of the processor cooler fan, we were able to understand the purpose of these parameters. First of all, it should be noted that the units of measurement of all these parameters are dimensionless and conventional. For example, the FAN Ctrl OFF and FAN Ctrl ON parameters, for which valid values ​​are in the range from 1 to 127, do indeed set some processor temperature values, but not in degrees Celsius (°C), but in some conventional units, and like these The conventional units are related to the actual temperature of the processor, it is not possible to understand.

As it turned out, the FAN Ctrl OFF parameter sets the processor temperature value, below which PWM control is disabled, that is, the duty cycle of PWM pulses is 0.

In the processor temperature range from FAN Ctrl OFF to FAN Ctrl ON, the duty cycle of PWM pulses corresponds to the value specified in the Fan Ctrl Start value parameter, and as soon as the processor temperature rises above the FAN Ctrl ON value, the duty cycle of PWM pulses increases from the Fan Ctrl Start value proportional to the change in processor temperature at a rate determined by the value of the Fan Ctrl Sensitive parameter.

The problem with manually setting the cooler rotation speed on the Biostar TH55XE board is that, without an oscilloscope at hand, it is impossible to configure this mode, since the values ​​of all settings are specified in dimensionless conventional units. Alas, the only thing the user can do in this case is to use the Performance or Quite modes (which are the same thing).

If we talk about the BIOS capabilities of the Biostar TH55XE board for overclocking, they are quite typical. You can overclock the processor either by changing the multiplication factor (in the range from 9 to 26 for the Intel Core i5-661 processor) or by changing the reference frequency (in the range from 100 to 800 MHz). The memory can also be overclocked by changing the divider value (DDR3-800/1066/1333) or the reference frequency. Naturally, it is possible to change memory timings, supply voltage and much more.

In addition, for novice users there is an automatic overclocking mode (Automate OverClock). In fact, we are talking about three preset overclocking profiles (V6-Tech Engine, V8-Tech Engine and V12-Tech Engine). When using the V6-Tech Engine profile, the system bus frequency increases to 135 MHz, the V8-Tech Engine profile - up to 140 MHz and the V12-Tech Engine profile - up to 145 MHz.

The Biostar TH55XE board comes with two proprietary utilities: TOverclocker and Green Power Utility. The TOverclocker utility allows you to control the main parameters of the system: processor clock frequency, system bus frequency, supply voltage, etc. In addition, it provides real-time overclocking of the processor by changing the system bus frequency and supply voltage. At the same time, the memory operating frequency increases. Using the TOverclocker utility you can also configure the cooler operating mode, however, as it turned out, this option does not work.

The Green Power Utility is designed to configure and monitor the operating mode of the processor voltage regulator. In general, this utility does not make much sense, and its readings are highly doubtful. However, both utilities often do not start.

Motherboard testing

To test motherboards based on the Intel H55 Express chipset, we used a bench with the following configuration:

• processor - Intel Core i5-661;
• Intel Chipset Device Software - 9.1.1.1025;
• memory - DDR3-1066 (Qimonda IMSH1GU03A1F1C-10F PC3-8500);
• memory capacity - 2 GB (two modules of 1024 MB each);
• memory operating mode - DDR3-1066, dual-channel;
• memory timings - 7-7-7-20;
• video card - integrated into the processor;
• video driver version - 15.16.6.2025;
• hard drive - Western Digital WD2500JS;
• power supply - Tagan 1300W;
• operating system - Microsoft Windows 7 Ultimate (32-bit).

Let us recall that the clock frequency of the Intel Core i5-661 processor is 3.33 GHz, and in Turbo Boost mode it can be 3.46 GHz with two active processor cores or 3.6 GHz when only one core is active. The frequency of the graphics core integrated into the Intel Core i5-661 processor is 900 MHz, and its TDP is 87 W.

The technical characteristics of the compared motherboard models are presented in table. 1 .

When testing the boards, we focused on measuring not performance, which is determined by the installed processor, chipset and memory, but energy consumption, and also looked at the implementation of controlling the rotation speed of the processor cooler fan.

We talked about the implementation of controlling the rotation speed of the processor cooler fan on each of the tested boards when describing the board itself. Let us only note that a digital oscilloscope was used to monitor the duty cycle of control PWM pulses in various operating modes of the cooler.

To measure energy consumption, a digital wattmeter was used, to which the power supply was connected. We emphasize that we measured the power consumption of the entire system based on the tested board, taking into account the power supply, hard drive and memory modules. Energy consumption was measured in two modes of system operation: full load and idle.

Product release date.

Lithography

Lithography indicates the semiconductor technology used to produce integrated chipsets and the report is shown in nanometer (nm), which indicates the size of the features built into the semiconductor.

Design power

Thermal design power (TDP) indicates the average performance in watts when the processor's power is dissipated (running at base frequency with all cores engaged) under a challenging workload as defined by Intel. Read the requirements for thermoregulation systems presented in the technical description.

Available options for embedded systems

Available options for embedded systems indicate products that provide extended purchasing availability for intelligent systems and embedded solutions. Product specifications and conditions of use are provided in the Production Release Qualification (PRQ) report. Contact your Intel representative for details.

Integrated Graphics‡

Integrated graphics deliver stunning graphics quality and performance, as well as flexible display options without the need for a separate graphics card.

Graphics output

The graphics output defines the interfaces available for interacting with the device's displays.

Intel® Clear Video Technology

Intel® Clear Video Technology is a set of video encoding and processing technologies built into the processor's integrated graphics. These technologies make video playback more stable and graphics clearer, brighter and more realistic.

PCI support

PCI support indicates the type of support for the Peripheral Component Interconnect standard

PCI Express Edition

The PCI Express edition is the version supported by the processor. PCIe (Peripheral Component Interconnect Express) is a high-speed serial expansion bus standard for computers to connect hardware devices to it. Different versions of PCI Express support different data transfer rates.

PCI Express Configurations‡

PCI Express (PCIe) configurations describe the available PCIe channel configurations that can be used to map PCIe PCHs to PCIe devices.

Max. number of PCI Express channels

The PCI Express (PCIe) lane consists of two differential signal pairs for receiving and transmitting data, and is also the basic element of the PCIe bus. The number of PCI Express lanes is the total number of lanes that the processor supports.

USB version

USB (Universal Serial Bus) is an industry standard connection technology for connecting peripheral devices to a computer.

Total number of SATA ports

SATA (Serial Storage Interface) is a high-speed standard for connecting storage devices such as hard drives and optical drives to the motherboard.

Integrated network adapter

The integrated network adapter assumes the MAC address of the Intel built-in Ethernet device or the LAN ports on the motherboard.

Integrated IDE adapter

The IDE interface is an interface standard for interconnecting storage devices, which indicates that the drive controller is integrated into the drive rather than being a separate component on the motherboard.

T CASE

The critical temperature is the maximum temperature allowed within the processor's integrated heat spreader (IHS).

Intel® Virtualization Technology for Directed I/O (VT-d)‡

Intel® Virtualization Technology for Directed I/O complements virtualization support in IA-32 architecture-based processors (VT-x) and Itanium® processors (VT-i) with I/O device virtualization capabilities. Intel® Virtualization Technology for Directed I/O helps users increase system security, reliability, and I/O device performance in virtual environments.

Intel® vPro™ Platform Compliant

The Intel vPro® platform is a set of hardware and technologies used to create business computing endpoints with high performance, built-in security, advanced management features and platform stability.

Intel® ME Firmware Version

Intel® Management Engine (Intel® ME) software leverages the built-in capabilities of the platform and management and security applications to remotely manage networked computing resources out of band.

Intel® Remote PC Assist Technology

Intel® Remote PC Assist Technology allows you to request remote technical assistance from your service provider when you encounter a PC problem, even if the OS, network software, or applications are not working. This service ceased to be provided in October 2010.

Intel® Quick Resume Technology

The Intel® Quick Resume Technology Driver (QRTD) enables your Intel® Viv™ Technology-based PC to be used as a consumer electronics device that can be turned on and off instantly (after the initial boot if the feature is enabled).

Intel® Quiet System Technology

Intel® Quiet System technology reduces system noise and heat generation through intelligent fan speed control algorithms.

Intel® HD Audio Technology

Intel® High Definition Audio supports more channels at higher quality than previous integrated audio systems. Additionally, Intel® High Definition Audio integrates the technologies needed to support the latest audio formats.

Intel® AC97 Technology

Intel® AC97 Technology is an audio codec standard that defines a high-quality audio architecture with surround sound support for PCs. It is the predecessor to Intel® High Definition Audio.

Intel® Matrix Storage Technology

Intel® Matrix Storage Technology provides security, performance, and expandability for desktop and mobile PC platforms. When using one or more hard drives, users can benefit from increased performance and reduced power consumption. When using multiple drives, the user receives additional protection against data loss in the event of a hard drive failure. Predecessor to Intel® Rapid Storage Technology

Intel® Trusted Execution Technology‡

Intel® Trusted Execution Technology enhances secure command execution through hardware enhancements to Intel® processors and chipsets. This technology provides digital office platforms with security features such as measured application launch and secure command execution. This is achieved by creating an environment where applications run in isolation from other applications on the system.

Anti-Theft Technology

Intel® Anti-Theft Technology helps keep data on your laptop safe if it's lost or stolen. To use Intel® Anti-Theft Technology, you must subscribe to an Intel® Anti-Theft Technology service provider.