The whole truth about multi-core processors. What is better nuclear power or frequency? Processor 2 cores

At the end of the day, we took a good look at the features of the new multi-core processor Intel Core 2 Extreme QX6700, focusing on its architectural features. In addition, the first results of the initial testing of this processor were presented. Let us guess that there were more than a few tests conducted by the technical experts of the Intel company at the IDF Forum 2006. Naturally, according to the test data, the four-core Intel Core 2 Extreme QX6700 processor looked even more hostile to the older dual-core Intel Core 2 Extreme X8600 processor. However, the set of tests that were tested, having raised some doubts about their objectivity, we decided to independently conduct comprehensive testing of the Intel Core 2 Extreme QX6700 processor in tandem with the Intel Core 2 Extreme X8600 processor.

Peredmova

We remember that at the IDF 2006 forum, Intel presented a new dual-core processor Intel Core 2 Extreme QX6700 and announced the first results of its testing with the dual-core processor Intel Core 2 Extreme X8600. For testing by Intel engineers, the following benchmarks and programs were selected:

• 3DMark06 v. 1.0.2;
• PCMark05 v. 1.1.0;
• 3DS Max 8 SP2;
• XMPEG 5.03 (DivX 6.2.5 codec);
• POV-Ray 3.7 Beta 15;
• Sony Vegas 7.0 Build 115.

Of course, such a set of tests cannot be considered objective for assessing the productivity and alignment of processors. Really, 3DMark06 v. 1.0.2 is a synthetic gaming test that is used for testing processors and video cards. Unfortunately, based on these results, it is not possible to draw conclusions about processor performance in games. And the fact that the PC demonstrates a high result in the 3DMark06 v test. 1.0.2 does not mean at all that the results will be the same in real PC games.

Test PCMark05 v. 1.1 allows you to conduct a comprehensive analysis of the productivity of the PC and its surrounding subsystems, including the processor. The undoubted advantage of this test is that it does not require a lot of hours for testing; there are few results from this test for a comprehensive objective assessment of PC productivity.

The 3DS Max 8 SP2 can be completely used for testing the processor, but Intel's developers used the testers' final rendering of trivial scenes. Aje robot with 3DS Max 8 SP - it’s not just the rendering, but the process of creating the scene itself. Scripts that imitate the work of a correspondent near the projection windows were also not stagnated during the testing. And although in this case the main emphasis lies on the processor of the graphics card, it would be wrong to say that the results should not be left to the processor.

The POV-Ray 3.7 Beta 15 add-on, which includes a benchmark, again allows you to test the processor for better performance when rendering trivial scenes. There is also the XMPEG 5.03 add-on, which, in tandem with the DivX 6.2.5 codec, was used for converting High Definition video content.

Well, the last addition is Sony Vegas 7.0a Build 115 – suitable for non-linear video editing. In this situation, everything is correct and we are not entitled to the same respect.

Regardless of the fact that the above tests (or supplements) are widely used and traditionally used for testing processors, to obtain any objective information about the performance of the Intel Core 2 Extreme QX6700 processor, Based on the results of this set of tests, it would be possible to not entirely correct. In general, it can be seen that in these specially selected tests, the four-core Intel Core 2 Extreme QX6700 processor demonstrates its superiority over the dual-core Intel Core 2 Extreme X8600 processor, but this does not mean that we can talk about increased productivity when working with other programs. It is possible to show, for example, the results of the video conversion test using XMPEG 5.03 programs paired with the DivX 6.2.5 codec that the Intel Core 2 Extreme QX6700 processor, when working with any video conversion programs, allows mothers increase in productivity in young women with an Intel Core 2 Extreme X8600 processor?

To get a more objective picture of the performance of the Intel Core 2 Extreme QX6700 processor and its class in advance, if we can talk about the absolute superiority of four cores over two, we decided to carry out full-fledged testing of several Nuclear and dual-core processors are subject to a wide range of tests.

First of all, let’s move on to look at the testing methodology and analyze the results, and let’s give a short introduction about the test participants.

Briefly about the Intel Core 2 Extreme QX6700 and Intel Core 2 Extreme X8600 processors

The Intel Core 2 Extreme QX6700 processor is codenamed Kentsfield. From the design point of view, the VIN is two dual-core Conroe processors housed in one processor housing.

The maximum power consumption (TDP) of the four-core Intel Core 2 Extreme QX6700 processor is 130 W, which also requires an effective cooling system, and it is impossible to create a quiet computer based on such a processor. The maximum power consumption (TDP) of the dual-core processor Intel Core 2 Extreme X8600 is lower than 95 W.

The Intel Core 2 Extreme QX6700 processor has a clock frequency of 2.66 GHz and a voltage of 1.238, the FSB frequency is 1066 MHz, and the total volume of L2 cache memory is 8 MB (2x4 MB). The Intel Core 2 Extreme X8600 processor has a clock frequency of 2.93 GHz and a voltage of 1.213 V, the FSB frequency is 1066 MHz, and the L2 cache memory is 4 MB.

Brief technical characteristics of both processors are given in the table. 1.

Table 1. Brief technical characteristics of processors
Intel Core 2 Extreme QX6700 and Intel Core 2 Extreme X8600

Testing technique

To test the Intel Core 2 Extreme QX6700 processor, the stand with the current configuration is used:

• motherboard – Intel D975XBX2 (BIOS BX97510J.86A.1304.2006.0620.1451);
• RAM – DDR2-800 Kingston KHX8000D2K2/2G (2x1024 MB in dual-channel mode);
• memory timings:

CAS Latency - 4,

RAS to CAS Delay - 4,

Row Precharge - 3,

Active to Precharge – 12;

• video system – MSI NX8800GTX video card (NVIDIA GeForce 8800GTX graphics processor); ForceWare video driver version 84.21;
• disk subsystem - two Seagate Barracuda 7200.7 drives with a capacity of 120 GB each, combined in a RAID level 0 array on a Sil 3114 RAID controller; NTFS file structure;
• Operating system – Windows XP Professional SP2.

The drivers of all new devices were installed in advance.

As already stated, the dual-core Intel Core 2 Extreme X8600 processor was protested.

To test both processors, we compiled benchmarks and real programs that intensively use the processor and memory and are traditionally used for a comprehensive analysis of system performance:

• Game tests:

Quake 4 Demo ver 1.3,

F.E.A.R. ver 1.07,

Far Cry v.1.33,

Prey ver 1.01,

Company of Heroes ver 1.0,

Serious Sam 2 Demo,

The Chronicles of Riddik;

• PC productivity in general:

Crystal Mark 9.0;

• scientific developments:

Science Mark 2.0,

Super_PI/mod 1.5 XS,

SunGard Adaptive Credit Risk;

• robot with 3D graphics:

3ds Max 8.0 SP3 (script SPECapc 3ds max 8 v.1.3),

Alias ​​WaveFront Maya 6.5 (script SPECapc Maya 6.5 v1.0),

SPECViewperf 9.0

CINEBENCH 9.5

POV-Ray v.3.7 Beta 17 (recall test);

• text recognition: ABBYY FineReader 8.0 Pro;
• processing of digital photographs: Adobe Photoshop CS2;
• audio code: Lame 4.0;
• archive: 7-ZIP 4.42;
• video code:

XMPEG 5.2 Beta 2,

DivX Converter 6.1.1,

TMPGEnc 2.524,

MainConcept MPEG Encoder 1.51,

MainConcept H.264 Encoder v.2.0.15.

All tests were run three times, and the results were calculated using the average value and the reliable range of values ​​with 95% confidence.

Description and setup of tests

Game tests

The group of gaming tests included the most popular dynamic games and the synthetic benchmark 3DMark06 v.1.0.2, which is used to measure PC productivity in gaming applications and is traditionally used for testing video cards. However, the results of this test depend not only on the video card, but also on the capabilities of the central processor.

In order to make the most of the processor itself, and not the video card, when tested, all games and the 3DMark06 v.1.0.2 benchmark were launched at a separate size of 800x600 pixels, and the video driver was adjusted for maximum productivity. In addition, in order to increase the emphasis on the central processor in games, anti-aliasing and anisotropic filtering technologies were not used. All games have been tuned for maximum productivity for all effects that increase the realism of the image, but are indicated for decreased productivity. A description of how to manage a skin infection is a tiresome and tedious task, so it’s easy to remember its main principle: all effects that can be turned on are turned on.

It is significant that the games Quake 4 ver. 1.3 and Prey ver 1.014 we have put together demo versions written specifically for this test, and for all others - those that are included in the igor warehouse.

In gaming tests, the speed of frame processing was measured, that is, the number of frames per second (frames per second, fps).

In the 3DMark06 v.1.0.2 test, the result, which is based on a complex formula, is measured in dimensionless units, and the more, the better.

PC productivity in perspective

PCMark05 and CrystalMark 9.0 have advanced to a group of tests aimed at measuring the overall productivity of a PC.

The first test is intended for a comprehensive analysis of PC productivity. It carries out a number of tests (part 48), which focus on different subsystems of the PC: processor, memory, graphics subsystem, data storage subsystem. The test results are followed by an integral indicator of system productivity, as well as productivity indices of various PC subsystems (CPU Score, Memory, Graphics, HDD).

The results of the PCMark05 test are expressed in unitless units, and the greater the result, the better.

Another test is also complex and is used to analyze the productivity of a PC across all subsystems. This benchmark is carried out in addition to tests with important considerations on the central processor (ALU, FPU), memory (MEM), data storage subsystem (HDD), graphics subsystem (GDI, D2D, OGL).

The test results are used to calculate a dimensionless integral indicator of productivity (Mark), and also indicate indicators of the productivity of various PC subsystems.

Well, we know the result is better, it’s better.

Science development

We have a group of tests that are scientifically designed, including Science Mark 2.0, Super_PI/mod 1.5 XS and SunGard Adaptiv Credit Risk.

The Science Mark 2.0 test measures PC productivity during the course of scientific research. The main considerations are the processor and memory.

Test results are given in sizeless units. Greater results indicate greater productivity.

For the Super_PI/mod 1.5 XS test, the PI number is calculated based on the specified accuracy (the number of digits after the comma). Our tester was given the highest accuracy - 32 M, that is 32 million characters after coma.

The result of the test is an hour of visualization, analysis in seconds. It turns out that less than an hour increases the productivity of the processor.

SunGard Adaptiv Credit Risk is a program that is designed to develop credit risks for a large number of officials with the analysis of a large set of data. It is the industry standard and is used by great corporations. Designed for use with cluster systems and heavy-duty servers, this program supports rich processor capacity and scales well with an increasing number of processors.

The result of the test based on the SunGard Adaptiv Credit Risk program is an hour in seconds. The lower the hour, the higher the productivity of the processor.

Archive

For archiving, vikorystvuvaya rich streaming archiver 7-Zip 4.42. The archive was supplied with a test directory of 135 MB in size, which was compressed to 66.9 MB, and the maximum compression level was set (Ultra).

The result of the test is an hour of archiving, while less than an hour is, of course, shorter.

Audio code

To encode audio files in the WAV format in the MP3 format, the popular Lame 4.0 codec is used. The encoding was supplied by a WAV file with an output size of 195 MB, which was converted into an MP3 file with a size of 17.7 MB. The codec was launched from the command line with standard settings (44.1 kHz, 128 Kbps).

The result of the test is an hour of conversion, expressed in seconds, and the less, the better.

Recognition of the text

To recognize the text, the program ABBYY FineReader 8.0 Pro was used. A 49-sided PDF file was selected as a document for recognition.

The result of the test is an hour of document recognition, expressions in seconds, and even less, then more quickly.

3D graphics

The group of tests that reveal the productivity of the processor when working with 3D add-ons includes SPECapc 3ds max8 v.1.3, SPECapc for Maya 6.5, POV-Ray 3.7 Beta 17, CINEBENCH 9.5 and SPECViewperf 9.0.3.

Test SPECapc 3ds max8 v.1.3 is a script for the Autodesk 3DS max 8.0 SP3 program and is intended for testing the platform with priority considerations for the processor and video card. This is how the rendering of end 3D scenes with the most important demands on the central processor is carried out, as well as the typical creation and editing of the scene with the main demands on the video card processor. In order to shift the main focus to the processor and minimize the input of the video card to the final test result, for the SPECapc 3ds max8 v.1.3 program. The software video driver is stuck.

The unique characteristic of the SPECapc 3ds max8 v.1.3 test is an hour of testing tasks. Based on the time required to carefully create and edit the scene, an integral indicator of the productivity of the video card, standards and the results of a given reference PC is developed. Similarly, based on the time of rendering of the final scenes, an integral indicator of the productivity of the central processor is calculated, which is also normalized to the results of a given reference PC.

The SPECapc for Maya 6.5 benchmark is intended for testing the platform of the Alias ​​WaveFront Maya 6.5 add-on, focusing on the processor, video card and disk subsystem. The test consists of 30 subtests.

The test result is given in the form of two standardized warehouses: standardized processor productivity and integral standardized productivity. When expanding the integral productivity, the following coefficients are introduced: for testing the video card – 0.7; for tests on the processor - 0.2 and for tests on the disk subsystem - 0.1.

To analyze the standardized test results, a reference PC with an Intel Xeon 2.4 GHz processor, 2 GB of PC800 ECC RDRAM memory and an NVIDIA Quadro FX 1000 video card is used.

Benchmark POV-Ray 3.7 Beta 17 purposes for assessing rendering speed and the main focus in the test lies on the processor. The test fails to use benchmarks, and the result is smooth PPS (Pixel Per Second) rendering.

The CINEBENCH 9.5 test is used for testing graphics cards and processors and allows for faster rendering. This test is tested by the CPU Benchmark, and the final result is the speed of rendering when all processors in the system are idle (for multi-processor systems), which is reflected in the dimensionless units of CINEBENCH.

SPECViewperf 9.0.3 is a test designed to improve the productivity of the graphics subsystem in professional OpenGL applications. It is traditionally used for testing graphics stations and professional video cards, whose results largely depend on the productivity of the processor.

The results of the test are absolute mental units (unitless), which indicate how many times the test has a PC performance that is tested above the performance of a standard PC.

Processing digital photographs

To evaluate the productivity of the processor during the hour of work with digital photo editing programs, the Adobe Photoshop CS2 script was used. Each image (digital photograph) in the TIFF format has filters sequentially applied to it and the total time of finishing all operations is calculated. The result of the test is an hour of victory, changes in seconds.

Video coding

A group of tests to evaluate the productivity of video encoding was composed of popular software converters and codecs. Five add-ons were selected: XMPEG 5.0.3, DivX 6.4 Converter, TMPGEnc 2.524, MainConcept MPEG Encoder 1.51 and MainConcept H.264 Encoder v. 2.0.

The XMPEG 5.0.3 utility was paired with the DivX 6.4.1 Codec. For an additional video clip of 24 with a size of 51.8 MB in the MPEG-2 format with a separate format of 1920x1980 pixels and a bitrate of 18,000 kbps, converted into an HD video file worldwide 36.5 MB with separate storage 7800 K1/s8.

The DivX 6.4 Converter utility was used to convert a 51.8 MB video clip in the MPEG-2 format with a separate 1920x1980 pixels and a bitrate of 18,000 Kbps for a Div video file X size 11 MB and separate building 1280x. The DivX 6.1.1 Converter utility has a High Definition profile.

The TMPGEnc 2.524 utility is designed for converting AVI files to MPEG format for burning to DVDs. Our version has an output AVI file of 416 MB in size and a size of 2 x 1 s, converted to an MPEG-2 video file (m2v+wav) of 115 MB in size in DVD 4:3 NTSC format. Allowed frames set at 720x480 pixels, bitrate - 8000 Kbps, video speed - 29.97 fps.

The MainConcept MPEG Encoder 1.51 utility is also designed for converting AVI files to MPEG format for burning to DVDs. Our version has an output AVI file of 416 MB in size and a size of 2 x 1 s, converted to an MPEG-2 (mpg) video file of 111 MB in size in DVD 4:3 NTSC format. Separate frame size - 720x480 pixels, production speed - 29.97 fps, video speed - 8000 Kbps.

For additional utilities MainConcept H.264 Encoder v. 2.0 the output AVI file with a size of 416 MB and three dimensions 2 x 1 using the additional H.264 High codec was converted to an MPEG-2 (mpg) video file with a size of 295 MB in the DVD 4: 3 NTSC format. The separate frame size was set to 720x480 pixels, the production speed was 29.97 fps.

Test results

The results of routine testing of processors are shown in the table. 2.

Table 2. Results of routine testing of processors

It was realized that it was difficult to analyze such a large amount of data, so we decided to divide the test results into logical groups and calculate the integral standard indicator of productivity from the skin group of tests. In order to normalize the results, the results of the Intel Core 2 Extreme X8600 were used, so the results demonstrated by this processor were taken as one.

The first logical group of tests is gaming programs. In this period, the integral indicator of productivity was measured as a geometric average compared to the standardized results for all games (the 3DMark06 benchmark did not perform well). We decided to make the 3DMark06 test incorrectly, since its result is poorly correlated with what is expected in real games.

The logical group was attacked by video coding tests. It was preceded by XMPEG 5.0.3, DivX Converter 6.4, TMPEGEnc 2.524, MainConcept H.264 Encoder v.2.0 and MainConcept MPEG Encoder v.1.51. The integral indicator of productivity was assessed as the geometric average of the standardized results for all tests. Other tests tended not to follow logical groups, which is due to their diversity, and to produce different results that were weakly correlated with one another.

The normalized results in such a simplified form are presented in the diagram.

Now we analyze the data extraction.

First of all, let's take a look at the results of testing in games. As you can see, a four-core processor not only does not outperform a dual-core processor, but loses approximately 10% in productivity. Therefore, the assertion that a four-core processor is oriented towards heavy PC gaming is no more than a myth. Today there are no possibilities that could take advantage of the stagnation of a four-core architecture.

This, of course, does not mean that the stink won’t show up tomorrow. Prote for daily use of a multi-core processor is ineffective.

The results of the gaming synthetic test 3DMark06 are up to the highest standards. The increase in productivity in 3DMark06 CPU Score became 58%, which is quite impressive. However, its integral result (3DMark Score) is modest - the increase in productivity is only 5%, so we are talking about an increase, not a decrease in productivity. Once again, we remind you that the 3DMark06 test is a lot of time to live and work on the basis of its results based on the fact that the Intel Core 2 Extreme QX6700 processor has superiority in gaming applications, but it would still be wrong.

Offensive test – PCMark05. However, the results are still mixed. In PCMark05 CPU Score, the Intel Core 2 Extreme QX6700 processor showed a 56% increase in productivity, but the integral result of this test (PCMark05 Score) is new for both processors. On the right, the increase in the PCMark05 CPU Score result is offset by the decrease in the PCMark05 Memory and PCMark05 Graphics results. Since this test is used as a comprehensive test that analyzes the performance of the PC as a whole, it should be noted that for the set of requirements to be tested in the PCMark05 test, a system based on a four-core Intel Core 2 Extreme QX6700 processor does not have an edge Lively with the system based on Intel Core 2 Extreme X8600 processor

In the CrystalMark 9.0 test, the Intel Core 2 Extreme QX6700 demonstrated a significant increase in productivity. Thus, the final result (Mark) increased by 26%, and the results of tests oriented towards the processor (ALU, FPU) increased by 77%.

Now let's look at the results of testing with 3D add-ons (SPECapc 3ds max8 v.1.3, SPECapc for Maya 6.5, POV-Ray 3.7 Beta 17, CINEBENCH 9.5 and SPECViewperf 9.0.3).

In the SPECapc 3ds max8 v.1.3 test, in tasks related to the rendering of final scenes, the Intel Core 2 Extreme QX6700 processor allowed a productivity increase of 46%, which is a very good showing. At the same time, in factories that work on projection windows (rotations, transformations, scaling, etc.), it was not an increase, but a 10-hundred-hundred-hundred-acre loss in productivity.

The SPECapc for Maya 6.5 test, which does not render final scenes, showed a similar picture – a 7% drop in productivity.

In the POV-Ray 3.7 Beta 17 test, which also means faster rendering, it was found that the Intel Core 2 Extreme QX6700 processor provided an 81% increase in productivity.

A similar picture was observed in the CINEBENCH 9.5 test, where rendering speed was observed. A four-core processor allows rendering to speed up an hour by 48% compared to a dual-core processor.

The SPECViewperf 9.0.3 test has an integral result for the Intel Core 2 Extreme QX6700 processor, which we determined to be the geometric average of the standardized results of all tests, 5% less than the Intel Core 2 Extreme X8600 processor. This test is primarily intended for testing professional video cards, but, as we have already stated, the result depends on the processor, and in this case, the presence of a multi-core processor does not mean increased productivity.

Tests that are scientifically based have mixed results. In Science Mark 2.0 and Super_PI/mod 1.5 XS tests, the Intel Core 2 Extreme QX6700 demonstrated a decrease in productivity by 7 and 3%, similar. However, the problem is, rather, the tests themselves, not the processor. The right thing is that these tests are single-threaded and poorly parallelized across a number of cores. It is not possible to realize that a rich-nuclear architecture can deprive them of productivity gains.

The SunGard Adaptiv Credit Risk test is no longer a costless utility, but a serious program, designed for use in large corporations and primarily aimed at high-processor servers. At this time, the Intel Core 2 Extreme QX6700 processor once again reveals its superiority – the productivity increase is 79%!

When working with the Adobe Photoshop CS2 add-on, the Intel Core 2 Extreme QX6700 processor allows you to achieve at least a significant increase (only 16%), but still an increase in productivity. And with the ABBYY FineReader 8.0 Pro add-on for text recognition, the situation is turning around. The use of the Intel Core 2 Extreme QX6700 processor resulted in a 5% decrease in productivity.

Archiving data with the help of the archiver 7-Zip 4.42 also resulted in a slight (4%) drop in productivity when the four-core processor was installed, and in tasks for audio conversion with the local Lame 4.0 codec it was already 9%.

Before testing data archiving and audio conversion, one thing needs to be done. In principle, these tests can be used to reveal the advantages of multi-core architecture. To do this, you need to launch a number of program sessions at a time. If, for example, you need to convert a bunch of WAV files, you can launch a bunch of sessions at the same time (without being afraid of writing a corresponding BAT file) and convert the same file from the same session. It is better, of course, to know the proper software shell for the codec, so that it can work automatically. In this case, a multi-core processor can really speed up the time spent converting audio files.

The remaining group of tests that we didn’t have time to look at were programs for video encoding. In this case, in all additions, the four-core processor has demonstrated its superiority. Depending on the specific program and video format, the increase in productivity increased from 10 to 66%.

Visnovki

So, what kind of benefits can be obtained based on the results of the test? The four-core Intel Core 2 Extreme QX6700 processor did not live up to our expectations. But, perhaps, the whole point is that under the enemies of the new homeland of dual-core Intel Core 2 Duo processors, the stinks were completely protected.

At this time, the potential capabilities implemented in the Intel Core 2 Extreme QX6700 processor are simply impossible to open, and the fragments of additional components that can be used to extract benefits from the multi-core architecture of the processor are not so abundant today. The blame lies with the video encoding and final rendering of three-dimensional scenes, with most of them during testing, the advantage of the multi-core processor was not blocked. Obviously, the Intel Core 2 Duo would be correctly positioned as a processor for graphics stations and PCs that are important for video processing. In other cases, the effectiveness of stagnating a four-core processor is even doubtful.

For home PC users based on a four-core processor, this is, in fact, exotic, or, of course, one of the ways is to get drunk in your eyes, otherwise there is no need for it.

Most often, a computer with an Intel Core 2 Extreme QX6700 processor sacrifices the productivity of a PC with an Intel Core 2 Extreme X8600 processor. Therefore, it is still premature to position it as a processor for high-performance home PCs. The initial, minor drop in productivity that comes with games and other programs cannot be overlooked. Prote computer based on the Intel Core 2 Extreme QX6700 processor is a highly productive solution. Power is only available to those who need it, as a dual-core processor allows you to achieve greater productivity with the highest demands, except for rendering and video encoding, and for less money and less energy living room

However, the computer is bought not for one price, but for the multi-core architecture and for the future. The Intel Core 2 Extreme QX6700 processor is ahead of its time, but the situation may change tomorrow. Now the software infrastructure is not ready to cope with the advantages of multi-core processors. There is no doubt about the fact that soon all new programs will support rich nuclear power.

Most often it is said that the advantages of a rich-nuclear architecture can be quickly achieved today - without waiting for the bright future. All you need to do is to start working in multi-tasking mode, if a number of different programs are simultaneously installed on the computer, such as anti-virus scanning, audio coding or games. It's true that it's true, but... it's too chastkovo. Seriously, this is nothing more than a marketing myth. To get to this point, try launching a custom video converter (for example, converting a movie for PocketPC) and running it in Microsoft Word, or simply playing solitaire on your computer. Tsikavo, how many weeks will it take you to get back together? Tim, it is important to note that most video converters for PocketPC (for example, Omniquiti Lathe 1.5) are single-threaded and do not utilize a number of processor cores at once, so one core is completely consumed, and the solution is simple yuyut. It would seem that no one cares about what kind of task is placed on other kernels. Yakbi is not just “ale”. On the other hand, in such scenarios, the speed of the system as a whole depends not only on the capabilities of the processor - but also on the hard drive, memory and bus width with a shared bandwidth. It is very likely that two or more add-ons that come together at the same time will begin to compete for the same (not processor) PC resources, which will not allow productivity to increase.

Previously, we realized that the effectiveness of a multi-core processor in home PCs is controversial. However, we have not yet considered another important aspect - marketing, which, apparently, is the driver of progress. It doesn’t matter whether the new processor is rotten or good, - if the company needs wines for marketing purposes, then there will certainly be releases.

However, Intel is in such a hurry to release a new, multi-core processor, because the processors of the Intel Core 2 Duo family have become crazy market leaders? The food is by no means more banal and easier to finish. First of all, for the Intel company, the Intel Core 2 Extreme QX6700 processor is a kind of by-product of production, which does not require serious financial investment: from a technological point of view, the production of dual-core processors Conroe and four-core Kentsfield It's teasing. The difference is limited at the packaging stage, which is produced in specialized factories in Malaysia. Intel has already perfected this process: due to the technology of packaging two dual-core chips in one case, multi-core Xeon server processors are not differentiated from Intel Core 2 Extreme QX6700 processors.

In fact, since the production of four-core processors does not require additional financial expenditures, why not publish it?

In other words, the appearance of a four-core processor is a legacy of Intel’s ambitious plans. When it comes to winning the title of industry leader and introducing a product to the market that is not available from competitors, it costs a lot. And there is no doubt that in terms of technology, the Intel Core 2 Extreme QX6700 is truly a great step ahead.

And, in our opinion, this is another, third reason for such a hasty release of a multi-core processor. In the competitive struggle between Intel and AMD, one company has to work on its progress. And of course, AMD could not help but react to the release of the multi-core Intel Core 2 Extreme QX6700 processor. Intel knew about this miraculously, as well as about those that have nothing to do with AMD. What came of it? The release of the four-core processor Intel Core 2 Extreme QX6700 by AMD has created a surprising and a priori failure of a solution code-named AMD 4x4, which uses two dual-core processors instead of one four-core processor. Why is this amazing? Throughout history, AMD has been trying hard to extend its leadership in the field of low-power processors. In addition, she has already said that increasing the clock frequency is not a method of increasing the productivity of processors. The release of the AMD 4x4 solution follows the policies pursued by the company. On the other hand, there is no need for energy-saving platforms, since they consume a lot of electrical energy and require an even more efficient (and therefore even cooler) cooling system. In addition, the new dual-core AMD processors (FX-70, FX-72 and FX-74) are nothing more than an upgraded version of the old processors in a new case, equipped with a Socket F (1207 FX).

Standardized results of routine testing of processors

The first test results of the AMD 4x4 solution, obtained in American and European test laboratories, allow such developments to be made. For the productivity of the AMD 4x4 solution with two dual-core AMD Athlon 64 processors, the FX-74 beats the solution based on one four-core Intel Core 2 Extreme QX6700 processor in almost all tests. Given that the energy consumption of an AMD 4x4 system is approximately twice as high, it also requires heavy-duty (not lower than 600 W) power supply units. The performance of AMD 4x4 is significant, as it is based on the Intel Core 2 Extreme QX6700 processor. In this manner, the answer to the question: “Who needs this?” is clear. The appearance of the four-core Intel Core 2 Extreme QX6700 processor prompted AMD to waste money and release a new family of dull dual-core processors onto the market, which were doomed to fail.

Hello everyone. So today we’ll talk about kernels, and more importantly about their strength. Not everything is as simple as it seems at first glance. As it turns out, 8 cores are definitely better than 4, so I think it’s easy to understand why, well, more cores mean more strength.

Ale os chomu rich. An 8-core processor from AMD costs less than a 4-core processor from Intel. And Intel, until socket 2011-3, there are no eight-core processors! What? Well, never mind! And the four cores that support Hyper-threading technology are like eight-cores in Windows. Tobto, bachita, not everything is so simple. Another important point is that an AMD processor with 8 cores will outperform an Intel processor with 4 cores. Then you see, the kernels are in turmoil, so move on. Although I think everyone already knows that Intel’s products are more optimized and more productive than AMD’s, there is no doubt about it.

So you started thinking, which is better? Here it is also important to understand more clearly why. Let's take a look at the robots first, Intel has three main models, i3 (2 cores/4 threads), i5 (4 cores), i7 (4 cores/8 threads). For games, you can take i7, which is enough not only for current games, but for future ones, as for me. i5 can calmly approach, it has 4 cores, try all the daily games. This i3 is more powerful at mid-range, and even at high-range, since the video card does not pump up.

I can’t say that it’s much better for 8 cores or 4. If you choose Intel, and it’s the cores themselves, and not the threads, then it’s much better for 8 cores. It’s amazing how funny it is. Zagalom of a lot of kernels is good, but still good. For example, you took an i7 and play the game, everything is fine. If you had taken i5 and kicked it out, then the result would have been the same as if you had taken out i7! I still have stock for future games. 4 cores with a high frequency, for example 4.6 GHz, work a little better with one resource-intensive tasks, well, then I play, lower i7 with a frequency of 3.8 GHz. Still, the i5 is cheaper than the i7

High frequency and number of nuclei are not the same thing. For example, for an office computer, you can take an i5, everything will be fine. Or you can take the Pentium G3258 for example, speed it up to 4.6 GHz, or a little less, and everything will be fine, even if there are two cores. A lot of kernels are allowed to be fired immediately. A high frequency allows you to conquer one task, or even more quickly. So, roughly speaking, it is absolutely possible to launch the program.

For office programs, I don’t rely on too many cores. At most two, but at a high frequency. For daily games, like mine, it’s best to have 4 cores at a high frequency. For any kind of Photoshop, resource-intensive programs, it’s imperative to get the i7.

Before speaking, I’m not sure, but the socket 2011-3 is supported by processors of the i7 family only, so they are the most productive.

Another axis at any moment, you can take the i7 on 1155 sockets, well, like a butt. Or you can take an i5 with 1151 sockets. In principle, so brutal in appearance, it would immediately seem that the i5 will be much weaker. So, everything is correct, but it’s NOT OK, on ​​the right is that 1155 socket is not old, and 1151 is a new and current socket. An i5 on socket 1151 will be close to an i7 on socket 1155. And if the i5 is still rozignat, then it will be a beauty. What am I doing? Cores by cores, but choose not only the number of them, but also marvel at the existence of the core, which is my pleasure for you

Well, guys, the axis will handle it, a little bit of porridge will come out, because I still can’t believe it’s better for 4 cores or 8 cores. So I’ll write again that Intel (excluding platforms 2011-3) does not have processors with 8 cores, a maximum of 6 cores and those with the old socket 1366. Otherwise, there are also full-fledged 8-core AMD processors, which can be pushing 4 -oh nuclear Intel. Well, and the most important thing: for current games, it’s better to take the i5 and sign it (models that are angry, go with the letter K), the socket with which is 1151. If you need to work in hard programs, then it’s better to use the i7 and instead . If you don’t have a lot of finances, but you want to spend money, then take i3. The entire Core I* family has become more productive processors no matter what.

QX | 22 Lipnya 2015, 14:45
The frequency and technical process are not less. Today's 2-core processors at 3 GHz are not on par with the first 2-core processors, which are also at 3 GHz. The frequency is the same, but the old ones are simply motorized galma equalized with the new ones. As a result, the current 2-core i3 is much smaller, and the lower 4-core Quad Q6600. The Pentium G is fresher than the old Quad.

QX | 11 Lipnya 2015, 12:18
Here the difference in frequency is not great, 3.5 versus 3 GHz. That's why there are 4 cores. However, other characteristics also do not appear. A lot of cores are required for archiving, video encoding, etc. By taking 2 nuclear weapons, you can protect yourself easily. More food, as you practice a lot on new food. It would be better if you still named the model yourself. And so, I would please you with a thicker and fresher Core i3.

MaKos007 | 30 February 2015, 16:00
Here I am rolling my thoughts on the tree. I’ll tell you straight away - your choice is a dual-core processor with a higher frequency. If the theory is not valid, you don’t have to read further.

The processor frequency depends on the number of operations performed per hour. Thus, the higher the frequency, the more events are generated per second, for example.

Why do we have so many cores... For the presence of more than one core, the processor can handle more than one task. These are like lines of a conveyor belt. One line of the conveyor runs smoothly, and two parallel lines on which the operation is carried out will produce greater results. Also, theoretically, dual-core solutions perform twice as fast as single-core solutions.

This theory, as well as with conveyors, means two flows that need to be developed. In this case, live correctly so that the skin trim works to its best advantage. In combination with processors, this depends on the architecture of the program and its very richness. If the addition allows you to separate a given number of threads (read - to improve the processor’s rich core capacity), then the rich core capacity can give a significant increase in the speed of executing commands. And if it’s not possible, or the data is such that it’s impossible to separate, there aren’t even too many cores in the CPU.

In fact, feeding the optimal number of kernels is easier. What is also important here is the architecture of the cores themselves and the connections between them. Thus, the first multi-core processors are significantly less functional devices, lower daily. In addition, it should be noted that the current operating systems Windows 7 and Windows 8 (I do not consider here * nix systems and their support of multi-core processors - a different topic) have been found to be very well parallelized. Thus, the richness of the core helps not to interfere with the main processes (based on the programs and games) through the control of background tasks. Thus, the anti-virus protection and firewall do not bother (more precisely, the least harm) when the Photoshop program is running.

For some programs, being rich in cores is important. After spending an hour on the Internet, you can see that it speeds up the conversion of video and audio; rendering 3D models, encrypting signals etc. You don't need 4 cores for Photoshop and video editing. It’s entirely enough, as I already said, two, except for the greatness of the skin of them.

teleport | 21 April 2013, 01:30
A simple productivity calculation shows: for a 2-core one the productivity is 2 x 3.5 = 7, for a 4-core one - 4 x 3 = 12. So the 4-core one is twice as powerful. In addition, it is melodiously modern, and therefore economical and productive. And if only one core is vikorized, it heats up less, since the frequency of one core is slightly lower, but there is no need for heating.

For video editing, the processor is generally not critical; it mainly takes up the resources of the video card or a special payment for video editing. If the processor has the same fate as a 2-core one, then other systems (various antivirus programs) will fight for the core, which will be lost, which will lead to motor dullness. In short, the rich nuclear content is short.

yang | 11 April 2013, 20:22
In this case, a dual-core processor will be more efficient and economical in all devices.

Nowadays, the number of cores plays a significant role in choosing a computer. Even with the very cores installed in the processor, the strength of the computer is affected, and its speed is high when processing data and producing the result. The number of cores in the processor chip is rotating, and their number can currently reach from one to four.

In that “old time”, when four-core processors were not yet asleep, those dual-core processors were amazing, the speed of the computer’s power varied in the clock frequency. The processor processed only one stream of information, and as you understand, until the result of processing reached the operator, it took nearly an hour. Now, the multi-core processor, with the help of specially shortened programs, divides the processing of data into several separate, independent threads, which significantly speeds up the result and increases the load on the computer. However, it is important to know that if the program is not tuned to work with a high-core processor, then the speed will be lower than that of a single-core processor with a high clock frequency. So how do you find out how many cores a computer has?

The central processor is one of the most important parts of any computer, and that means how many cores there are in it, and it is a very difficult task for a computer genius, which is the beginning of what can be used for your successful transformation into edited computer bison . Well, that means how many cores your computer has.

Reception No. 1

• To do this, click on the computer bear from the right side, clicking on the “Computer” icon, or in the context menu, displayed on the desktop, on the “Computer” icon. Select the “Power” item.

• A window will open on the left and find the “Device Manager” item.
• In order to open the list of processors available on your computer, click on the arrow located left-handed under the main items, including the “Processors” item.

• Having realized how many processors are on the list, you can easily tell how many cores the processor has, and also the skin core that allows for repetitions, or other recordings. The butt presented to you can be seen to have two cores.

This method is suitable for Windows operating systems, and the axis on Intel processors that use hyper-threading technology (Hyper-threading technology), this method, for all intents and purposes, can be divided into two threads, Square one kind of one. In addition, a program that is suitable for one operating system works on the skin behind the core, and as a result, removes an eight-core processor. If your processor supports Hyper-threading technology, go to a special diagnostic utility.

Reception No. 2

There are cost-free programs for those who worry about the number of cores in the processor. So, the unpaid CPU-Z program can completely cope with your tasks. To quickly get started with the program:

• go to the official website cpuid.com, and access archives from CPU-Z. It is better to use the faster version, which does not require installation on the computer; this version is marked “no installation”.
• Next, unzip the program and launch it from the compressed file.
• In the main window of the program that has opened, on the “CPU” tab, in the lower part you will find the “Cores” item. The axis here will indicate the exact number of cores of your processor.

You can find out how many cores a computer with Windows installed has by asking the task manager.

Reception No. 3

The evilness of the action is like this:

• You can launch the dispatcher by clicking on the right side of the bear on the quick launch panel, which is displayed at the bottom.
• A window will open, you will see the item “Start task manager”

• At the very top of the Windows Manager there is a “Swedcode” tab, its axis, in addition to the chronological display of central memory, and you can see the number of cores. And the skin also means the core, showing your affection.

Priyom No. 4

And another opportunity for updating the computer cores, for which you will need documentation for the computer with a full range of components. Find an entry about the processor. If the processor is upgraded to AMD, then turn on the symbol X and the number, which costs a lot. If it costs X2, that means you get a processor with two cores, etc.

In Intel processors, a number of cores are written in words. If it costs Core 2 Duo, Dual, then there are two cores, if it’s Quad, it’s just that.

Of course, you can fix the kernels by logging into the motherboard through the BIOS, but it will work if the described methods give a clear picture of the power supply that you need, and you can verify what the store told you the truth. Wow, how many cores are there? Your computer is independent.

P.S. Well, that’s all, now we know how to identify how many cores are in a computer, in several ways, and which one to stagnate is your decision 😉

In contact with

We discovered an unacceptable problem between clock frequencies. Having reached the threshold of 3 GHz, manufacturers have seen significant increases in the energy efficiency and thermal performance of their products. The level of technology in 2004 did not allow radical changes in the size of transistors on a silicon crystal, and the way out of the situation was to try not to increase the frequency, but to increase the number of operations that would be completed in one tact. Having adopted the latest server platforms, once the multiprocessor layout had already been tested, it was possible to combine two processors on one chip.

Almost an hour later, CPUs with two, three, four, six and even more cores became widely available. However, the main part of the market is occupied by 2 and 4-core models. AMD is trying to change the situation, but its Bulldozer architecture has not proven its promise and budget eight-core processors are still not that popular in the world. That's why foodWhat's better: 2 or 4-core processor, everything is no longer relevant.

The difference between 2 and 4-core processors

On the hardware levelthe main capacity of a 2-core processor versus a 4-core one- Number of functional blocks. The cutaneous core, in essence, is the CP, equipped with its own computational nodes. 2 or 4 such CPUs are interconnected with an internal fast bus and an underground memory controller for interaction with RAM. Other functional units may also be hidden: most modern CPUs have individual cache memory of the first (L1) and another (L2) level, blocks of whole calculations and operations with a floating coma. The L3 cache, which is incredibly valuable, is one and accessible to all cores. It is also possible to note that AMD FX (as well as Athlon CPUs and A-series APUs) have the same cache memory and controller, and the blocks are calculated using a floating coma: such a module can be placed on two cores at the same time.

Scheme of a multi-core AMD Athlon processor

From the point of view of the Koristuvachsdifference between 2 and 4-core processorThere are only so many tasks that the CPU can process in one clock cycle. With the new architecture, the theoretical difference becomes 2 times for 2 and 4 cores or 4 times for 2 and 8 cores, apparently. Thus, when many processes are running at the same time, an increase in the number of processes is likely to cause an increase in the speed code of the system. If you replace two operations on four core CPUs at one time, you can also cancel several times.

Why the popularity of dual-core CPUs is considered

It would seem that an increase in the number of cores causes an increase in productivity, but on models with as many as six or eight cores, dual-core processors have no better chance. Tim no less, the leading leader in the CPU market, Intel, is rapidly updating its product range and releasing new models of everything with a pair of cores (Core i3, Celeron, Pentium). And despite the fact that in smartphones and tablets with such CPUs, consumers are surprised and unsure. To understand why the most popular models are processors with two cores, consider a number of main factors.

Intel Core i3 - the most popular 2-core processors for home PCs

The problem of madness. During the development of software, developers are trying to develop the technology so that it can function like new computers and existing models of CPUs and DPs. Looking at the assortment on the market, it is important to ensure that the game runs normally both on two cores and on eight. Most home PCs are equipped with a dual-core processor, which is why the greatest respect is given to such computers.

Foldability. To ensure the efficient acquisition of all cores, counting, and generation of programs in the robot process, divide the data into equal threads. For example, the task is to optimally engage all the cores, seeing each of them in one or two processes - simultaneous compression of several videos. With games it’s more complicated, because all the operations are related to each other. Regardless of the fact that the main robot is powered by the graphics processor of the video card, the information for forming a 3D image is prepared by the CPU itself. It is difficult to work in such a way that the cutaneous core produces its portion of data, and then supplies it to the GP synchronously with others. The more one-hour flows that need to be calculated, the more important the implementation of the task.

Reduced capacity of technology. Software vendors are relying on the same principles for their new projects, which are recognized by multiple modernizations. In some cases, it can get to the point where such technologies are 10-15 years behind their roots. The development, based on a ten-year-old project, is even reluctant to give in to radical reworking for perfect optimization. As a result, we must be careful not to release software that rationally exploits the hardware capabilities of the PC. Gra S.T.A.L.K.E.R. Pripyat's call, which was released in 2009 (during the era of the rise of multi-core CPUs), was based on the 2001 engine, which cannot accommodate more than one core.

S.T.A.L.K.E.R. completely affects only one core of a 4-core CPU

The same situation is with the popular online RPG World of Tanks: the Big World engine, on which it is based, was launched in 2005, if the rich-core CPUs have not yet caught up with the only possible way to develop.

World of Tanks is also unable to divide the vantage into cores incrementally

Financial difficulties. The last point of this problem is the first point. If you create a skin program from scratch, without using advanced technologies, its implementation will cost a fortune. For example, the cost of developing GTA V totaled over 200 million dollars. With this technology, however, they were not created “from scratch”, but were based on previous projects, since the game was written for 5 platforms at a time (Sony PS3, PS4, Xbox 360 and One, as well as PC).

GTA V is optimized for high-core power and a progressively faster processor

All these nuances do not allow us to fully exploit the potential of multi-core processors in practice. The interdependence of hardware vendors and software vendors creates a closed circle.

Type of processor: 2 or 4-core

Obviously, despite all the advantages, the potential of multi-core processors remains unrealized until the end. These tasks cannot be distributed evenly and work in one flow, while others fail with average efficiency, and only a small part of the program fully interacts with everything ma kernels. That's why foodFastest processor, 2 or 4 coresbuy, it means a good understanding of the flow situation.

There are products from two manufacturers on the market: Intel and AMD, which differ in the specifics of their implementation. Advanced Micro Devices traditionally talks about being rich in cores, since Intel does not want to go to such lengths and increase the number of cores as long as it does not lead to a decrease in productivity per core (it is very difficult to avoid this).

An increase in the number of nuclei reduces the productivity of the skin

As a rule, the theoretical and practical productivity of a multi-core CPU is lower than that of a similar one (based on the same microarchitecture, with the same technical processor) with one core. It is important to note that the kernels use different resources, and are not particularly indicated on the speed code. In this way, you can’t just add a heavy-duty six-core processor with a range to those that will definitely not be weaker than a dual-core processor from the same series. In some situations it will be, in which case it is obvious. As a rule, you can run old games on a computer with an eight-core AMD FX processor: FPS is lower than on a similar PC, but with a four-core CPU.

What is the need for rich nuclear power today?

Does this mean that a lot of cores are not needed? They don’t care about those that again seem natural, no. Light day-to-day tasks (such as web surfing or working with many programs at once) respond positively to an increase in the number of processor cores. For this very reason, smartphone manufacturers are focusing on efficiency, relegating productivity to another level. Opera (and other browsers based on the Chromium engine), Firefox launch each open tab in the same process, obviously, the more cores there are, the faster the transition between tabs. File managers, office programs, software developers are not power-hungry and resource-intensive. If you need to frequently switch between them, the multi-core processor will increase the productivity of the system.

The Opera browser provides a complex process for every user

Intel knows that HuperThreading technology, which allows the core to process other threads using resources that are not wasted, appeared in the Pentium 4. However, it does not allow one to compensate for the lack of productivity again. .

In "Watch Manager" a 2-core processor with Huper Threading appears as a 4-core processor

The creators of the game, at this time, must act step by step. The advent of new generations of Sony Play Station and Microsoft Xbox consoles stimulated retailers to pay more attention to richness. Since the consoles are based on eight-core AMD chips, programmers no longer need to spend a lot of effort on optimization when playing ported games on a PC. Due to the growing popularity of these consoles, those who were disappointed in the AMD FX 8xxx could be relieved. Nuclear-rich traders are forced to take double positions above the market, so that they can jump from the butt of their gaze.