Repeated cases of false positives.

All tips

The recent announcement of a fresh operating system from Redmond caused a great wave of discussions among our readers. Possibility of free updating, collection of information about users, replacement of a seemingly eternal Internet browser Explorer and even jokes about Bill Gates' ability to count to ten, all this became the subject of active discussion and debate. However, today we will focus our attention on a slightly different question that is primarily of interest to the gaming community: “so, what can the “ten” do in games?

Perhaps the most important thing to remember when considering installing Windows 10 is that its main trump card - DirectX 12 - remains unavailable. The first games using the new set of APIs should appear towards the end of 2015, but for now the only application that can demonstrate the potential of DirectX 12 is a synthetic benchmark from Futuremark. Therefore, on at this stage we will compare operating systems from Microsoft in games that are popular on this moment, we’ll find out how quickly and efficiently video card driver developers work and try to understand whether it makes sense to upgrade to Windows 10 now if you spend most of your time on your PC mastering virtual worlds.

Testing methodology

The main participants in testing were three Windows operating systems last generations: Windows 7 Ultimate SP1 x64, Windows 8.1 Pro x64, Windows 10 Pro x64. The official image of each OS was installed on a GoodRAM SSD Iridium Pro 240GB solid state drive, after which all available updates from their Windows Update center were downloaded. Installed in the same way latest versions drivers for hardware components. The role of graphics adapters was taken by ASUS MATRIX-GTX980-P-4GD5 and MSI R9 290X Lightning, which made it possible to use accelerators from two different camps. Both video cards, together with the Intel Core i5-4690K central processor, did not lend themselves to additional overclocking and operated at the nominal frequency set by the manufacturer.

Full configuration test bench as follows:

processor: Intel Core i5-4690K (3.5 GHz, 6 MB);
cooler: Zalman CNPS10X Flex (NF-A15 PWM fan, 140 mm);
motherboard: MSI Z87M Gaming (Intel Z87);
RAM: GoodRAM GY1600D364L10/16GDC (2x8 GB, 1600 MHz, 10-10-10-28-1T);
video card No. 1: ASUS MATRIX-GTX980-P-4GD5 (GeForce GTX 980);
video card No. 2: MSI R9 290X Lightning (Radeon R9 290X);
system disk: GoodRAM SSD Iridium Pro 240GB (240 GB, SATA 6 Gbit/s);
auxiliary storage: ADATA SX900 256GB (256 GB, SATA 6 Gbit/s);
power supply: Chieftec CTG-750C (750 W);
monitor: LG 23MP75HM-P (1920x1080, 23″);

operating system No. 1: Windows 7 Ultimate SP1 x64;
operating system No. 2: Windows 8.1 Pro x64;
operating system No. 3: Windows 10 Pro x64;
GeForce driver: NVIDIA GeForce 355.60;
Radeon driver: ATI Catalyst 15.7.1.

FPS measurements were carried out in two synthetic tests and 10 real gaming applications. The screen resolution was 1920x1080 pixels. The frame rate per second was recorded using in-game benchmarks or using the Fraps utility v.3.5.99. Each test scene was repeated five times to minimize possible error, after which the average result was displayed. The final graphs show the values of minimum and average fps.

All available graphics settings are set to high. Additionally, the command was written in the console fps_max 0 to remove the built-in fps limitation. The test scene covers the beginning of the game: running around the map, starting control of runes, and a fight on the center line.

In GTA V, the maximum image quality parameters are selected, MSSA 2x anti-aliasing and tessellation are enabled. Additionally, options for loading detailed objects and shadows at a longer distance are activated.

Game version - 1.07. Preset - “Extraordinary quality.” All included Extra options in addition to proprietary technologies from NVIDIA: HBAO+ and NVIDIA HairWorks. The test scene is a horse ride through a ruined village at the beginning of the game, lasting a total of 60 seconds.

The image quality is high. Additional options such as HBAO+ and “depth of field” are disabled.

The frame rate was measured based on the results of the in-game benchmark. Graphics settings according to the “Ultra” preset without any changes.

Screen resolution - 100% (1920x1080 pixels). Maximum graphics settings, including MSAA 4x anti-aliasing and SSAO lighting processing algorithm.

Image profile - “maximum”, SMAA anti-aliasing and lighting processing according to SSBC technology.

Ultra image quality preset. Testing took place both using an in-game benchmark and with the participation of Fraps programs. The latter helped to more accurately determine the minimum fps.

Game version - 9.9. Once again, all sliders are set according to the preset “Maximum” profile.

The in-game “Tank Battle” test was used as a test scene. “Cinematic” picture quality mode. Anti-aliasing in 6x mode is activated. Test results

Before we begin analyzing the test results, it is necessary to clarify the issue of short-term fps drops in games when using Windows 10, which some users experience. In my case, a similar situation could be observed in Battlefield 4 and Grand Theft Auto V. The solution turned out to be quite simple - just disable the p2p update system Windows Update Delivery Optimization (WUDO), which is basically active in the Home and Pro versions. You can do this in the section Windows settings Update. I can’t say 100% that this is the reason, but the steps described above ensured that there were no lags in gaming applications throughout the entire testing.

The final 3DMark Fire Strike graph can be safely used in an advertising campaign aimed at supporting a new product. Regardless of the chosen video card, it was possible to obtain uniform “ladders”, indicating the superiority of newer OS versions over their predecessors, provided that the PC configuration is absolutely identical. In practice, the situation is somewhat different, as we will see as we progress through the material.

The first real gaming application immediately demonstrated mixed results. The NVIDIA video card managed to improve its performance in Windows 10 compared to Windows 7 and Windows 8.1. The maximum advantage reaches 9.7%. A more interesting situation developed with a representative from the red camp: the best minimum fps was also achieved in Windows 10, while the average fps turned out to be higher in the older “seven”. Windows 8.1, in in this case, got the position of an outsider.

In Grand Theft Auto V, the Radeon R9 290X shows, although not large, a stable increase in fps in new generations of the OS. For the GeForce GTX 980, similar behavior is typical only with respect to the minimum frame rate, where the gaps are more noticeable and range from 15% between Windows 8.1 and Windows 10 to an impressive 28.3% between the “ten” and “seven”. Windows 8.1 again brings up the rear.

The results obtained in one of the best RPG games of 2015 - The Witcher 3: Wild Hunt - are different high density. For the Radeon R9 290X, the difference was only one frame per second, which can be attributed to a possible error. The GeForce GTX 980 is most productive when paired with Windows 10, which provides it with an 8.1% handicap over the popular Windows 7 in terms of minimum fps.

Overall, the environment in Dying Light is similar to what we saw in The Witcher 3: Wild Hunt, only now the difference in the results of the GeForce GTX 980 is more obvious. Windows 7 is capable of providing a minimum fps of 77 frames per second, in Windows 8.1 this figure is already 83 fps, and Windows 10 guarantees an increase of another 4 frames. For the Radeon R9 290X, the OS used does not play a significant role.

So we got to one of the most discussed games on Windows 10. As difficult as it may be to guess, the reason for the negativity from users was more low level performance in the environment of the new operating system from Microsoft. Our testing partially confirms this. The graphs show a decrease in the performance of the GeForce GTX 980 by an average of 7%, both in terms of minimum and average fps. If your system has an accelerator from AMD installed, then on the contrary, you can count on an additional boost of 1-3%.

Middle-earth: Shadow of Mordor is the first game where the benefits of installing Windows 10 will be immediately noticeable, although this only applies to owners of AMD graphics adapters. The minimum frame rate increased from 45.5 fps to 66.18 fps, which is an increase of 45%. To ensure the reliability of the result, the built-in benchmark was repeated more than ten times and the game was even reinstalled - the data was confirmed. The GeForce GTX 980 turned out to be completely indifferent to the version of the installed operating system.

IN Far Cry 4 GeForce card The GTX 980, running Windows 8.1 and Windows 10, showed parity of results, but both systems were inferior to the performance of the “seven”. Count on maximum return from Radeon R9 290X can be used with Windows 10 or Windows 7. It seems that Windows 8.1 is not the best choice for the gamer of today.

For tankers, Windows 10, like its older sister Windows 8.1, is not of particular interest. In the best case, performance will be at the level of the proven Windows 7, and in the worst case, performance will decrease by 6-8%.

New test participant

If we have already begun testing current operating systems, then it is worth remembering one more participant - an alternative for Windows promoted by Valve. We are, of course, talking about SteamOS. Behind Last year The range of games available for this OS has expanded significantly. More and more often you can find announcements of new games, the release of which is planned on two competing platforms at once.

Since SteamOS is built on Linux based, DirectX resources are not available to it as such, but the winner in the “OpenGL vs DirectX” confrontation is known in advance. In any case, we compared the capabilities of SteamOS in two computer games, more recent than Metro Last Light and Half Life 2. They became a simulator of equipment from the Second World War War Thunder and steampunk first-person shooter BioShock Infinite. The decisive factor in choosing test applications was that these games have built-in benchmarks, because full-fledged analogue There is no Fraps utility for SteamOS yet.

First, let's look at the productivity of Microsoft platforms. In the case of the GeForce GTX 980, Windows 10 was 2-5% faster than Windows 7. The final fps indicator of Windows 8 is located somewhere in the middle. The Radeon R9 290X runs faster under Windows 7, outperforming its peers by an average of 3%.

As for SteamOS, it lags and lags noticeably. An average frame rate of 41 fps can be achieved on a significantly cheaper video card than the R9 290X, if you play under Microsoft's OS. And in general, Radeon products are definitely not the best choice for building a Steam Machine. Problems arise already at the stage of installing the operating system, when SteamOS cannot independently select the correct video driver and in order to somehow improve the level of performance, you have to update it in manual mode via terminal.

In BioShock Infinite, Windows 7 proved to be the sole leader. It is noticeably ahead of its competitors in terms of minimum fps, be it GeForce GTX 980 or Radeon R9 290X. SteamOS is lame on both legs and the graphs eloquently reflect this. The gap with products from Redmond can be as much as 73%.

Analytics

Analyzing the graphs presented in the review, it becomes clear that the leader may vary depending on the specific test application. A lot depends on the installed video card and the video drivers it uses. Therefore, to make it easier for readers to decide whether to switch to Windows 10, a general histogram of the “Total Average FPS” was built.

Final average fps

What do we have in the end? Windows 8.1 gains outsider status in the eyes of gamers. It’s quite strange that the operating system, released back in 2013, was never able to surpass the level of its predecessor. Windows 7, on the contrary, feels cheerful and confident. It became the leader in performance when the system was installed Radeon video card R9 290X and was slightly behind in tests with the GeForce GTX 980.

About Windows 10, despite its early stage life cycle, there is nothing bad to say. It is somewhere faster, somewhere slower than the “old woman” Windows 7, but overall it is on the same level with it. Further improvement of the new OS by Microsoft engineers, together with support from video driver developers for current graphics adapters, will definitely secure Windows 10’s position as a leader. The only sad thing is that to achieve this goal, Microsoft itself resorts to not entirely humane methods, removing Windows 7 and 8.1 from the list of supported platforms for its exclusives. The implementation of DirectX 12 is certainly a step forward, but a shadow of doubt about the artificial nature of the limitation still remains.

Speaking of DirectX 12. At the beginning of the material, we mentioned that there is a test that can demonstrate the potential of the API capabilities, and we tested our equipment in it.

The advantage of DirectX 12 over DirectX 11 (both in single-threaded and multi-threaded mode) is really huge, but the gap over Mantle is not so great, only 14.6%. In any case, the data obtained should not be taken at face value. This in no way means that we will soon witness a tenfold increase in fps in computer games, which can already be seen today in the example of Battlefield 4 using Mantle technology.

conclusions

The Windows 10 operating system from Microsoft has proven itself to be a fast and stable platform for games. First of all, owners of Windows 8.1 should think about the possibility of an upgrade, for whom such a step will mean a free increase in performance in a number of gaming applications, and the control interface will seem familiar and familiar. For adherents of the “people's” Windows 7, there are no compelling reasons to switch to the new OS yet. A massive update among participants in this category will most likely occur at the end of 2015 - beginning of 2016, when several interesting projects with DirectX 12 support will be released.

In the end, I would like to touch upon the issue of further prospects for the development of the SteamOS platform. It's quite difficult to find and highlight any benefits from purchasing or building a Steam Machine yourself. The main problems are the lower performance of system components in gaming applications, a limited list of available games, and difficulties in compatibility with AMD video cards. If the user is faced with the task of getting some kind of analogue of a game console, where the keyboard and mouse will not be in demand, then Windows installation and activating the Big Picture mode in the Steam client looks much more promising, because as soon as you need to customize SteamOS for yourself, you will not only have to connect the hated keyboard, but also get acquainted with the list of commands for the Linux terminal.

Dmitry Chekanov, thg.ru

Of course, many of our readers initially persisted and did not switch from Windows XP, but last year some of them began to give up and install Vista.

Perhaps Microsoft's announcement of the end of life of this famous operating system had a positive effect on the popularity of Windows Vista. Or perhaps many high-end users simply wished to enjoy new features such as visual interface Aero or Media Center, available in Home Premium and Ultimate versions. However, some of Vista's annoying problems remain, including slow boot times, a huge memory hunger, and the need for even more memory to minimize slow boot times via SuperFetch.

Meanwhile, the radical reduction in prices for high-speed and capacious memory has become, apparently, the most probable cause Vista's growing popularity among enthusiasts. New applications and games that require additional memory have arrived just in time to support the increase in memory capacity in computers, but to properly support 4GB or more, a 64-bit operating system is required. Vista was the first Windows OS to offer 64-bit versions en masse for the mass market, while poor driver support under 64-bit Windows XP meant that Vista x64 was an easier choice.

But for many of us, the trade-off of using an OS that supports and requires additional memory seems like a waste of resources, especially since the additional memory does not provide the 100% responsiveness that was felt in the previous OS. Microsoft has finally decided to listen to this OS problem Windows 7, which is an improved version of Vista.

We've already published several articles covering some of the changes Microsoft has made to its latest operating system. However, we are no less interested in the performance of programs and their responsiveness. Of course, we have heard a lot of positive feedback about how Windows 7 improves security, functionality, and the feel of working with the system. But let's narrow the focus to actual performance and see how much faster Windows 7 than the OS it replaces. It's time to measure the difference.

Microsoft Windows 7 / Test configuration

Modern hardware and software deserve each other, so we took some of the latest components to evaluate the difference in performance of the two operating systems.

We used mother's Asus board P7P55D. She supports the latter Intel socket LGA 1156 and is equipped with a P55 Express PCH chipset, while this model, in addition to one more motherboard from Gigabyte, performed well in our last comparison. And better compatibility with the system imaging program we use made Asus an obvious choice.

The reason we took such a recent motherboard model is because it supports the latest Intel processors for socket LGA 1156. Technology Turbo Boost allows Core processor i7-870 increases the CPU multiplier up to 27x (3.6 GHz), 26x (3.46 GHz) and 24x (3.2 GHz) in single-, dual- and multi-threaded applications.

U LGA processors 1156 have an interesting feature - they require higher C-states (sleep states) for three cores in order for the remaining core to set the highest multiplier. If rumors about better support If the C-states in Windows 7 are correct, then the new OS will give a significant performance boost in single-threaded applications.

The new Thermalright MUX-120 cooler delivers Cogage True Spirit-level performance for LGA 1156 processors.

Microsoft Windows 7 / Tests and settings

Microsoft Windows 7 / 3D games

Windows 7 shows a slight increase in performance in our Crysis test, but results that don't fundamentally affect the smoothness of the game can hardly be called significant.

We don't see any noticeable difference in performance between the two operating systems in Far Cry 2.

S.T.A.L.K.E.R.: Clear Sky gives a negligible performance increase under Windows 7; the frame rate can only be called unacceptable in the most “heavy” mode.

In World in Conflict we got strange results: scaling is worse under Windows 7, but we get better performance at low settings.

Microsoft Windows 7 / Audio and video encoding

Audio encoding tests run faster under Windows 7 for both iTunes and Lame.

TMPGEnc does not seem to work very well under Windows 7, so we will wait for an update to this program. Such a significant difference usually indicates some kind of incompatibility.

MainConcept works a little faster under Vista.

Microsoft Windows 7 / Applications

Adobe Photoshop CS4 under Windows Vista is slightly faster, although the difference is within the margin of error of a few runs. So it could just be a coincidence.

In 3ds Max, the difference between Vista and Windows 7 is again one second, although in this short test it seems longer. However, one second can still be attributed to an error.

AVG under Windows 7 runs a little faster, but the difference can again be attributed to chance.

The more interesting WinRAR test leans towards Vista. But WinZip is not.

Microsoft Windows 7 / Performance Results Analysis

Before we move on to synthetic tests, let's take a closer look at what real-world application tests have revealed to us.

In games we see minimal growth Windows performance 7, although productivity application tests run faster under Windows Vista. This difference in the average performance of encoding applications is due to the TMPGEnc problem that we already pointed out above. But what happens if we don't take into account the results of the TMPGEnc encoding test?

If we exclude the TMPGEnc encoding test, the performance difference drops to 0.41%, and the average difference across all applications is 0.57% in favor of Vista. But in this case, the apps don't say anything about loading times. To evaluate it, we need to turn to two synthetic tests PCMark and SYSmark.

Microsoft Windows 7 / Synthetic tests

Sandra is relatively responsive to the OS, so running tests should show that the hardware's performance has not changed.

Now that we have confirmed that everything is in order with the hardware, let's see how the remaining synthetic tests will react to changes in the software component.

PCMark tests reveal several aspects of system performance, including boot times. Productivity test results show that Windows 7 will "feel" 10% faster in everyday work.

The SYSmark 2007 Preview synthetic test shows a significant 25% gain under Windows 7 in the Video Creation run, although the average performance improvement is a more realistic 7%. By the way, such a 7% improvement in performance from a hardware upgrade usually costs hundreds of dollars. Unlike most other synthetic benchmarks, SYSmark uses real software packages such as Adobe Creative Suite and Microsoft Office, so users can consider a 7% increase as a good indicator of how faster system will be felt in everyday work.

Microsoft Windows 7 / Conclusion

From a performance testing perspective, moving from Windows Vista to Windows 7 is a line change in the configuration table. The only test that showed a significant difference was slower under Windows 7 due to the incompatibility of one of the program components.

Yes, most conventional tests are designed to show differences in performance between different hardware, so they don't do a very good job of demonstrating how users will experience their day-to-day experience. The game certainly runs smoother at 40 frames per second (fps) than at 30 fps, but the test doesn't show how long it will take to load a level. Rendering a 3D animation in 24 seconds is of course a technical time saver compared to 25 seconds, but again, it doesn't show how much time the user loses launching a program or navigating between different menus.

To better assess how users feel about the operation of the system and programs, it is necessary to measure loading time. Of course we could use a stopwatch, but electronic timer more accurate, and tests like PCMark and SYSmark use it. These programs show that Windows 7 feels 7-10% faster compared to Windows Vista, and for that alone we can be positive about the new OS, even if we don't see much of a difference in most other tests.

Adobe Photoshop CS4 under Windows Vista is slightly faster, although the difference is within the margin of error of a few runs. So it could just be a coincidence.

In 3ds Max, the difference between Vista and Windows 7 is again one second, although in this short test it seems longer. However, one second can still be attributed to an error.

AVG under Windows 7 runs a little faster, but the difference can again be attributed to chance.

The more interesting WinRAR test leans towards Vista. But WinZip is not.

Microsoft Windows 7 | Performance Results Analysis

Before we move on to synthetic tests, let's take a closer look at what real-world application tests have revealed to us.

In gaming, we see minimal performance gains in Windows 7, although productivity application tests run faster under Windows Vista. This difference in the average performance of encoding applications is due to the TMPGEnc problem that we already pointed out above. But what happens if we don't take into account the results of the TMPGEnc encoding test?

Microsoft Windows 7 | Synthetic tests

Sandra is relatively responsive to the OS, so running tests should show that the hardware's performance has not changed.

Now that we have confirmed that everything is in order with the hardware, let's see how the remaining synthetic tests will react to changes in the software component.

PCMark tests reveal several aspects of system performance, including boot times. Productivity test results show that Windows 7 will "feel" 10% faster in everyday work.

The SYSmark 2007 Preview synthetic test shows a significant 25% gain under Windows 7 in the Video Creation run, although the average performance improvement is a more realistic 7%. By the way, such a 7% improvement in performance from a hardware upgrade usually costs hundreds of dollars. Unlike most other synthetic benchmarks, SYSmark uses real-world productivity suites like Adobe Creative Suite and Microsoft Office to measure productivity, so users can consider the 7% increase a good indicator of how much faster the system will feel in everyday work.

Microsoft Windows 7 | Conclusion

To better assess how users feel about the operation of the system and programs, it is necessary to measure loading time. Of course, we could have used a stopwatch, but an electronic timer is more accurate, and tests like PCMark and SYSmark use it. These programs show that Windows 7 feels 7-10% faster compared to Windows Vista, and for that alone we can be positive about the new OS, even if we don't see much of a difference in most other tests.

Note: We published an article demonstrating the difference in responsiveness of Windows 7 compared to Vista - an article on the power consumption of the Core i7 Mobile. In it, we conducted a full run of PCMark Vantage under Win7 and Vista, while power consumption was measured every two seconds. Yes, Windows test 7 finishes faster, but also uses more energy during the run.

As sad as it may sound, the planks random access memory, called RAM, or, in English, RAM, can cause the appearance huge amount problems, ranging from the usual delay in loading the operating system and ending with its complete crash (especially if the most primitive conditions that need to be paid attention to are not met). In general, a RAM test (let's take Windows 7 64 bit as an example) will have to be performed in any case when installing new modules or replacing old ones. And at the same time, it is worth taking into account several main factors that in the future may influence the identification of the causes of malfunctions or the occurrence of conflicts.

Why do you need a Windows 7 64 bit RAM test?

First of all, you need to clearly understand that not every RAM module can fit into the motherboard slot and work without problems. The motherboard itself may simply not support the installation of certain types of brackets or even similar equipment from unknown manufacturers (although it may meet all the stated criteria).

The question is different: most users try to perform a RAM test only when too obvious problems are detected in the computer. And when checking equipment, most of us do not pay any attention to such parameters as clock operating frequency, timing, operating voltage, supported bus, throughput(which, by the way, must be compatible with the processor), etc.

Preliminary equipment check

And if problems arise, which are most often expressed in the fact that when the operating system starts, failure messages begin to appear (most often this is a blue screen indicating an error code indicating that the failure occurred not in software, but in physical level) everyone is trying to reboot the system, considering the error to be a temporary short-term phenomenon. This is wrong.

Thus, the RAM test (Windows 7 64 bit in our case) should initially be limited to checking the physical condition of the motherboard slots, the performance of the RAM strips and the consistency of the equipment characteristics with each other.

To begin with, really try just restarting your computer terminal or laptop. Sometimes (when too heavy load RAM) the system freezes precisely because of this. This means that in this moment too many applications are running and system processes, which occupy the maximum amount of RAM, and additional virtual memory, for which the pagefile.sys file is responsible, backing up free place on system and logical drives, is also missing.

Digressing a little, it is worth saying that even if there are logical partitions, a page file is created for each of them.

The simplest method of checking for blue screens on desktop PCs

Now let's turn to desktop computers, since in them checking the performance of any strip is completely elementary. By at least, the proposed operational test Windows memory 7 64 bit - or any other system - will give results immediately. The idea is to first remove all RAM modules from their respective slots while the computer is turned off, and then insert them one at a time and then reboot. If the system starts in operating mode, there are no problems with memory. Otherwise, the bar itself will have to be replaced either due to a breakdown or due to a discrepancy with the parameters of the mother chip.

Are there built-in system tools?

The Windows 7 64 bit RAM test, if anyone doesn’t know, is initially turned on when the computer itself starts. In other words, when voltage is applied to the motherboard, the primary hardware detection system comes into play. BIOS devices or more modern version UEFI, which has a graphical interface. That's not the point. Have you noticed that the stationary unit emits a signal from the system speaker when turned on? A short one-time signal just corresponds to the fact that everything system components in order.

If something is wrong, you should go into the settings of the primary I/O system and check the RAM parameters (in particular, this concerns the voltage supplied to the strips, the bridge used, etc. - overlockers understand what we are talking about).

How to use built-in diagnostics?

But self-diagnosis in the BIOS is far from the most ideal option for determining the full functionality of the RAM. The Windows 7 64 bit RAM error test itself can cause incorrect data to appear in reports. The problem is that both BIOS and Windows can perceive RAM completely differently.

Here you will have to use at least the built-in system tool. First, you need to call the “Start” menu and enter the line mdsched in it, after which the application that appears will offer two options for checking: without restart and at the next start. It is better to choose the second method, despite the system's recommendations. It will take a little longer to reboot, but the problems will be identified for sure.

Windows 7 64 bit RAM test: Memtest86+ program and AIDA64 application

But Windows-based systems quite often lose out to third-party software. The Mentest86+ utility is recognized as one of the most powerful programs.

Its only drawback is that it only works in DOS mode and produces access errors to RAM sectors, similar to how scanning is performed hard drives. The average user will understand little about this, but the error addresses can tell a lot to an advanced user.

The second utility (in Russian) performs the Windows 7 64 bit RAM test in such a way that even when launched, it displays the basic characteristics of the motherboard and RAM modules compatible with it. If you use the SPD section, you can learn a lot of interesting things about your motherboard and RAM sticks. For example, it will clearly indicate what maximum volume is supported, which manufacturer is suitable, which slots with their electrical parameters can be used, etc. In general, it is generally accepted (both among users and among experts) that the “AIDA” program performs the RAM test of Windows 7 64 bit best (although none of the proposed utilities eliminates problems, we're talking about namely diagnostics). But that's not all.

Checking RAM from a USB drive

Even if there are problems with the hard drive, memory or starting the PC, the Windows 7 64 bit RAM test (errors will only be detected, but not corrected) can be performed even when booting from removable media, for example, a regular flash drive.

The most important condition is to download not only the Memttest86+ utility, but also its additional installer, but not in the form of an ISO image, but in the form of a special USB installer, from which you will need to create bootable USB flash drive. After this, when you reboot the system, you will first have to enter the BIOS settings and set the USB device you are using as the first one when you start the computer, then boot and perform the appropriate test.

Burning and using an optical disc

If someone doesn’t like this option (for example, a person is used to using exclusively CD/DVD disks, or the settings of the primary system in terms of perceiving the first device from optical media look simpler), you can use the same Memtest86+ utility, launched exclusively in compatibility mode with Windows (and not DOS), after which, when specifying the media on which the main package and boot data will be recorded, the appropriate drive is selected in the menu.

Possible problems

Sometimes you may notice that when you download the above utility from removable media Users often note that the scan goes quite quickly at first, but at some stage it just freezes. In principle, as the developers themselves note, this is normal. The fact is that testing is carried out on the principle of several passes. In this case, one or two are enough. If there are RAM errors, they will be detected immediately. If they are not there, if the utility freezes, the testing process can be forcibly terminated.

Conclusion

As you can see, this is exactly what the Windows 7 64 bit RAM test looks like. What program can be used? The question is quite complex, because none of the currently created utilities can eliminate errors or problems (unlike the same testing programs, error correction, or even remagnetization of the surface of hard drives). Alas, if problems are detected, you will have to change the memory sticks, since they can cause physical damage hardware, and provoke the appearance of hardware conflicts even at the software level of the primary BIOS/UEFI systems, not to mention the main components of the Windows OS, which are related to RAM only in terms of system requirements and the maximum available volume for downloaded software modules in the form of heavy dynamic libraries which are for serious software products are required in the main set.

Even the most polished and most secure operating system is by no means guaranteed against failures perceived by the user as errors. Particularly annoying are unknown and unexpectedly occurring errors, sometimes fraught with extremely unpleasant consequences.

It is not difficult to imagine the emotions of a person working on complex document, which is lost as a result of a Windows system crash. To avoid similar problems— it’s better to prepare for them in advance. And to do this, you need to be aware of how Windows 7 OS is checked for errors. Let us consider separately two serious questions related to the stated topic:

Checking Windows system files and registry.
Control hard state computer disk.

Old timers computer world People well remember the utilities of the world-famous Peter Norton, not only the author of excellent books, but also the creator of programs under the “Norton utilities” brand. The first versions of these programs worked even before the advent of Windows - in the then popular MS DOS operating system. These utilities made it possible to detect the presence of errors on the hard drive and identify the presence of other DOS problems. We will be interested in similar tools for Windows.

Files and Registry

Checking OS files can be done in two ways: either using standard Windows tools, or using a third-party software. The OS has built-in file monitoring tools. To use their capabilities you need to open a window command line and type in it sfc team with the /scannow parameter, like this:

As a result, system files will be scanned to detect errors in them. The result will be displayed immediately in the command line window.

The information obtained can be analyzed (at least using the Internet), which can be useful for determining the degree of wear and tear of the OS and equipment. During operation, the program will try to correct all detected violations in system files on the disk.

Additional service, more high quality scanning and reliable adjustments are provided by third-party software.

For example, the same package “Norton Utilities” (NU) for Windows. Although this remedy and is the most popular and powerful software package for dealing with OS and computer faults - it still costs a lot of money. Especially in its “professional” configuration. Nowadays, you can find many free analogues of this creation on the Internet.

The Windows registry is the very place where, in addition to the information needed by the OS, a lot of all sorts of rubbish and garbage accumulates. Periodic cleaning of the registry is our direct responsibility. For this purpose, it is worth installing and periodically running the time-tested CCleaner utility (although NU also copes well with this task). Search for it online and download it.

HDD

The appearance of errors in the file structure on a hard drive is caused by wear on the disk surface, failures in Windows disk services (drivers), and positioning errors of the laser subsystem of the hard drive. Check and cure HDD possible in ways similar to the previous ones. We will look at the easiest way - launch standard program disk error checker that comes with Windows. For this:

Open the “My Computer” shortcut, select the icon of any of the partitions (for example, “Local Disk C”).
Open context menu right click mice.
Select Properties. A window with tabs will open.
Go to the "Service" tab
Click on the “Run check” button.

The program will examine the partition and fix any problems. Perform the same operation with the remaining partitions.