IntroductionNew Intel processors belonging to the family Ivy Bridge, have been on the market for several months now, but meanwhile it seems that their popularity is not too high. We have repeatedly noted that compared to their predecessors, they do not look like a significant step forward: their computing performance has increased slightly, and the frequency potential revealed through overclocking has become even worse than that of the previous generation Sandy Bridge. Intel also notes the lack of rush demand for Ivy Bridge: life cycle The last generation of processors, the production of which uses an older technological process with 32 nm standards, is extended and extended, and not the most optimistic forecasts are made regarding the distribution of new products. More specifically, by the end of this year, Intel plans to bring Ivy Bridge's share of desktop processor shipments to only 30 percent, while 60 percent of all CPU shipments will continue to be based on the Sandy Bridge microarchitecture. Does this give us the right not to consider the new Intel processors as another success for the company?
Not at all. The fact is that everything said above applies only to processors for desktop systems. The mobile market segment reacted to the release of Ivy Bridge in a completely different way, because most of the innovations in the new design were made specifically with laptops in mind. Two main advantages of Ivy Bridge over Sandy Bridge: significantly reduced heat dissipation and power consumption, as well as an accelerated graphics core with support for DirectX 11 - in mobile systems are in great demand. Thanks to these advantages, Ivy Bridge not only gave impetus to the release of laptops with a much better combination of consumer characteristics, but also catalyzed the introduction of a new class of ultraportable systems - ultrabooks. The new technological process with 22-nm standards and three-dimensional transistors has made it possible to reduce the size and cost of manufacturing semiconductor crystals, which, naturally, is another argument in favor of the success of the new design.
As a result, only desktop computer users may be somewhat averse to Ivy Bridge, and the dissatisfaction is not due to any serious shortcomings, but rather to the lack of fundamental positive changes, which, however, no one promised. Do not forget that in Intel’s classification, Ivy Bridge processors belong to the “tick” clock, that is, they represent a simple translation of the old microarchitecture onto new semiconductor rails. However, Intel itself is well aware that fans of desktop systems are somewhat less intrigued by the new generation of processors than their fellow laptop users. Therefore, there is no rush to carry out a full-scale update model range. At the moment, in the desktop segment, the new microarchitecture is cultivated only in older quad-core processors of the Core i7 and Core i5 series, and models based on the Ivy Bridge design are adjacent to the familiar Sandy Bridge and are in no hurry to relegate them to the background. A more aggressive introduction of the new microarchitecture is expected only in late autumn, and until then the question of which quad-core Core processors are preferable - the second (two-thousandth series) or third (three-thousandth series) generation - buyers are asked to decide on their own.
Actually, to facilitate the search for an answer to this question, we conducted special testing, in which we decided to compare Core i5 processors belonging to the same price category and intended for use within the same LGA 1155 platform, but based on different designs: Ivy Bridge and Sandy Bridge.
Third generation Intel Core i5: detailed introduction
A year and a half ago, with the release of the second generation Core series, Intel introduced a clear classification of processor families, which it adheres to currently. According to this classification, the fundamental properties of the Core i5 are a quad-core design without support for Hyper-Threading technology and a 6 MB L3 cache. These features were inherent in the previous generation Sandy Bridge processors, and they are also observed in the new version of the CPU with the Ivy Bridge design.This means that all Core i5 series processors using the new microarchitecture are very similar to each other. This, to some extent, allows Intel to unify its product output: all of today's Core i5 generations of Ivy Bridge use a completely identical 22-nm semiconductor chip with E1 stepping, consisting of 1.4 billion transistors and having an area of about 160 square meters. mm.
Despite the similarity of all LGA 1155 Core i5 processors in a number of formal characteristics, the differences between them are clearly noticeable. A new technological process with 22-nm standards and three-dimensional (Tri-Gate) transistors allowed Intel to reduce the typical heat dissipation for the new Core i5. If previously Core i5 in LGA 1155 version had a thermal package of 95 W, then for Ivy Bridge this value is reduced to 77 W. However, following the reduction in typical heat dissipation, there was no increase in clock frequencies of the Ivy Bridge processors included in the Core i5 family. The older Core i5s of the previous generation, as well as their today's successors, have nominal clock speeds not exceeding 3.4 GHz. This means that in general, the performance advantage of the new Core i5 over the old ones is provided only by improvements in the microarchitecture, which, in relation to CPU computing resources, are insignificant even according to the Intel developers themselves.
Speaking of strengths fresh processor design, first of all you should pay attention to the changes in the graphics core. Third generation Core i5 processors use a new version Intel video accelerator – HD Graphics 2500/4000. She has support software interfaces DirectX 11, OpenGL 4.0 and OpenCL 1.1 and in some cases can offer better 3D performance and faster video encoding high resolution to H.264 format via Quick Sync technology.
In addition, the Ivy Bridge processor design also contains a number of improvements made in the hardware - memory and bus controllers PCI Express. As a result, systems based on the new third-generation Core i5 processors can fully support video cards that use graphics PCI bus Express 3.0, and are also capable of clocking DDR3 memory at higher frequencies than their predecessors.
From its first debut to the general public until now, the third-generation Core i5 desktop processor family (that is, Core i5-3000 processors) has remained almost unchanged. Only a couple of intermediate models have been added to it, as a result of which, if we do not take into account economical options with a reduced thermal package, it now consists of five representatives. If we add a pair of Ivy Bridge Core i7 based on the Ivy Bridge microarchitecture to this five, we get a complete desktop line of 22 nm processors in LGA 1155 version:
The table above obviously needs to be supplemented to describe the functioning of the technology in more detail. Turbo Boost, allowing processors to independently increase their clock frequency if energy and temperature operating conditions allow it. In Ivy Bridge this technology has undergone certain changes, and the new Core i5 processors are capable of auto-overclocking somewhat more aggressively than their predecessors belonging to the Sandy Bridge family. Against the background of minimal improvements in the microarchitecture of computing cores and the lack of progress in frequencies, this is often what can ensure a certain superiority of new products over their predecessors.
The maximum frequency that Core i5 processors are capable of reaching when loading one or two cores exceeds the nominal by 400 MHz. If the load is multi-threaded, then Core i5 generation Ivy Bridge, provided they are in favorable temperature conditions, can raise their frequency by 200 MHz above the nominal value. At the same time, the efficiency Turbo works Boost for all processors under consideration is absolutely the same, and the differences from the previous generation CPUs are a greater increase in frequency when loading two, three and four cores: in the Sandy Bridge generation Core i5, the auto-overclocking limit in such conditions was 100 MHz lower.
Using the readings of the CPU-Z diagnostic program, let’s take a closer look at the representatives of the Core i5 lineup with Ivy Bridge design.
Intel Core i5-3570K
The Core i5-3570K processor is the crown of the entire third-generation Core i5 line. It boasts not only the highest clock frequency in the series, but also, unlike all other modifications, it has an important feature, emphasized by the letter “K” at the end of the model number - an unlocked multiplier. This allows Intel, not without reason, to classify the Core i5-3570K as a specialized overclocking offering. Moreover, compared to the older overclocking processor for the LGA 1155 platform, Core i7-3770K, Core i5-3570K looks very tempting thanks to a much more acceptable price for many, which can make this CPU almost the best market offer for enthusiasts.
At the same time, the Core i5-3570K is interesting not only for its predisposition to overclocking. For other users, this model may also be interesting due to the fact that it has a built-in older variation of the graphics core – Intel HD Graphics 4000, which has significantly higher performance than the graphics cores of other members of the Core i5 model range.
Intel Core i5-3570
The same name as the Core i5-3570K, but without the final letter, seems to hint that we are dealing with a neo-overclocking version of the previous processor. So it is: the Core i5-3570 operates at exactly the same clock speeds as its more advanced brother, but does not allow unlimited multiplier variation, which is popular among enthusiasts and advanced users.
However, there is one more “but”. The Core i5-3570 did not include a fast version of the graphics core, so this processor is content with the younger version of Intel HD Graphics 2500, which, as we will show below, is significantly worse in all aspects of performance.
As a result, the Core i5-3570 is more similar to the Core i5-3550 than the Core i5-3570K. For which he has very good reasons. Appearing a little later than the first group of Ivy Bridge representatives, this processor symbolizes a certain development of the family. Having the same recommended price as the model that is one line lower in the table of ranks, it seems to replace the Core i5-3550.
Intel Core i5-3550
A decreasing model number once again indicates a decrease in computing performance. In this case, the Core i5-3550 is slower than the Core i5-3570 due to its slightly lower clock speed. However, the difference is only 100 MHz, or about 3 percent, so it should not be surprising that both the Core i5-3570 and Core i5-3550 are rated the same by Intel. The manufacturer’s logic is that the Core i5-3570 should gradually displace the Core i5-3550 from store shelves. Therefore, in all other characteristics, except for the clock frequency, both of these CPUs are completely identical.
Intel Core i5-3470
The younger pair of Core i5 processors, based on the new 22nm Ivy Bridge core, have a recommended price below the $200 mark. These processors can be found in stores at similar prices. At the same time, the Core i5-3470 is not much inferior to the older Core i5: all four computing cores are in place, a 6-MB third-level cache and a clock speed of over 3 gigahertz. Intel chose a 100-MHz clock frequency step to differentiate modifications in the updated Core i5 series, so there is simply no way to expect a significant difference between the models in performance in real tasks.
However, the Core i5-3470 additionally differs from its older brothers in graphics performance. The HD Graphics 2500 video core operates at a slightly lower frequency: 1.1 GHz versus 1.15 GHz for more expensive processor modifications.
Intel Core i5-3450
The youngest variation of the third generation Core i5 processor in the Intel hierarchy, the Core i5-3450, like the Core i5-3550, is gradually leaving the market. The Core i5-3450 processor is smoothly replaced by the Core i5-3470 described above, which operates at a slightly higher frequency. There are no other differences between these CPUs.
How we tested
To get a full breakdown of the performance of modern Core i5s, we tested in detail all five Core i5s of the 3,000th series described above. The main competitors for these new products were earlier LGA 1155 processors of a similar class belonging to the Sandy Bridge generation: Core i5-2400 and Core i5-2500K. Their cost makes it possible to contrast these CPUs with the new Core i5 of the three thousandth series: Core i5-2400 has the same recommended price as Core i5-3470 and Core i5-3450; and the Core i5-2500K is sold slightly cheaper than the Core i5-3570K.In addition, we included in the charts the test results for higher-end processors Core i7-3770K and Core i7-2700K, as well as a processor offered by a competitor, AMD FX-8150. By the way, it is very significant that after the next price reductions, this senior representative of the Bulldozer family costs as much as the cheapest Core i5 of the three thousandth series. That is, AMD no longer harbors any illusions about the possibility of pitting its own eight-core processor against Intel's Core i7 class CPU.
As a result, the test systems included the following software and hardware components:
Processors:
AMD FX-8150 (Zambezi, 8 cores, 3.6-4.2 GHz, 8 MB L3);
Intel Core i5-2400 (Sandy Bridge, 4 cores, 3.1-3.4 GHz, 6 MB L3);
Intel Core i5-2500K (Sandy Bridge, 4 cores, 3.3-3.7 GHz, 6 MB L3);
Intel Core i5-3450 (Ivy Bridge, 4 cores, 3.1-3.5 GHz, 6 MB L3);
Intel Core i5-3470 (Ivy Bridge, 4 cores, 3.2-3.6 GHz, 6 MB L3);
Intel Core i5-3550 (Ivy Bridge, 4 cores, 3.3-3.7 GHz, 6 MB L3);
Intel Core i5-3570 (Ivy Bridge, 4 cores, 3.4-3.8 GHz, 6 MB L3);
Intel Core i5-3570K (Ivy Bridge, 4 cores, 3.4-3.8 GHz, 6 MB L3);
Intel Core i7-2700K (Sandy Bridge, 4 cores + HT, 3.5-3.9 GHz, 8 MB L3);
Intel Core i7-3770K (Ivy Bridge, 4 cores + HT, 3.5-3.9 GHz, 8 MB L3).
CPU cooler: NZXT Havik 140;
Motherboards:
ASUS Crosshair V Formula (Socket AM3+, AMD 990FX + SB950);
ASUS P8Z77-V Deluxe (LGA1155, Intel Z77 Express).
Memory: 2 x 4 GB, DDR3-1866 SDRAM, 9-11-9-27 (Kingston KHX1866C9D3K2/8GX).
Graphic cards:
AMD Radeon HD 6570 (1 GB/128-bit GDDR5, 650/4000 MHz);
NVIDIA GeForce GTX 680 (2 GB/256-bit GDDR5, 1006/6008 MHz).
Hard drive: Intel SSD 520 240 GB (SSDSC2CW240A3K5).
Power supply: Corsair AX1200i (80 Plus Platinum, 1200 W).
Operating system: Microsoft Windows 7 SP1 Ultimate x64.
Drivers:
AMD Catalyst 12.8 Driver;
AMD Chipset Driver 12.8;
Intel Chipset Driver 9.3.0.1019;
Intel Graphics Media Accelerator Driver 15.26.12.2761;
Intel Management Engine Driver 8.1.0.1248;
Intel Rapid Storage Technology 11.2.0.1006;
NVIDIA GeForce 301.42 Driver.
When testing a system based on the AMD FX-8150 processor, patches operating system KB2645594 and KB2646060 were installed.
The NVIDIA GeForce GTX 680 video card was used to test the speed of processors in a system with discrete graphics, while the AMD Radeon HD 6570 was used as a benchmark when studying the performance of integrated graphics.
The Intel Core i5-3570 processor did not participate in testing systems equipped with discrete graphics, since in terms of computing performance it is completely identical to the Intel Core i5-3570K, operating at the same clock speeds.
Computational performance
Overall PerformanceTo evaluate processor performance in common tasks, we traditionally use the Bapco SYSmark 2012 test, which simulates user work in common modern office programs and applications for creating and processing digital content. The idea of the test is very simple: it produces a single metric characterizing the weighted average speed of the computer.
In general, Core i5 processors belonging to the three thousandth series demonstrate quite expected performance. They are faster than the previous generation Core i5, and the Core i5-2500K processor, which is almost the fastest Core i5 with a Sandy Bridge design, is inferior in performance to even the youngest of the new products, the Core i5-3450. However, at the same time, fresh Core i5s are not able to reach the Core i7, due to the lack of Hyper-Threading technology in them.
A deeper understanding of SYSmark 2012 results can be provided by familiarizing yourself with the performance estimates obtained in various system usage scenarios. The Office Productivity scenario simulates typical office work: preparing text, processing spreadsheets, working with by email and visiting Internet sites. The script uses the following set of applications: ABBYY FineReader Pro 10.0 Adobe Acrobat Pro 9 Adobe Flash Player 10.1 Microsoft Excel 2010, Microsoft Internet Explorer 9, Microsoft Outlook 2010, Microsoft PowerPoint 2010, Microsoft Word 2010 and WinZip Pro 14.5.
In the script Media Creation simulates the creation of a commercial using pre-shot digital images and videos. For this purpose, popular Adobe packages are used: Photoshop CS5 Extended, Premiere Pro CS5 and After Effects CS5.
Web Development is a scenario within which the creation of a website is modeled. Applications used: Adobe Photoshop CS5 Extended, Adobe Premiere Pro CS5, Adobe Dreamweaver CS5 Mozilla Firefox 3.6.8 and Microsoft Internet Explorer 9.
The Data/Financial Analysis scenario is dedicated to statistical analysis and forecasting of market trends, which is performed in Microsoft Excel 2010.
The 3D Modeling script is all about creating three-dimensional objects and rendering static and dynamic scenes using Adobe Photoshop CS5 Extended, Autodesk 3ds Max 2011, Autodesk AutoCAD 2011 and Google SketchUp Pro 8.
The last scenario, System Management, involves creating backups and installing software and updates. There are several different Mozilla versions Firefox Installer and WinZip Pro 14.5.
In most scenarios, we are faced with a typical picture where the Core i5 3000 series is faster than its predecessors, but inferior to any Core i7, both based on the Ivy Bridge microarchitecture and Sandy Bridge. However, there are also cases of processor behavior that is not entirely typical. Thus, in the Media Creation scenario, the Core i5-3570K processor manages to outperform the Core i7-2700K; when using 3D modeling packages, the eight-core AMD FX-8150 performs unexpectedly well; and in the System Management scenario, which generates mainly a single-threaded load, the previous generation Core i5-2500K processor almost catches up with the performance of the fresh Core i5-3470.
Gaming Performance
As you know, the performance of platforms equipped with high-performance processors in the vast majority of modern games is determined by the power of the graphics subsystem. That is why, when testing processors, we try to conduct tests in such a way as to remove the load from the video card as much as possible: the most processor-dependent games are selected, and tests are carried out without turning on anti-aliasing and with settings that are not at the highest resolutions. That is, the results obtained make it possible to evaluate not so much the level of fps achievable in systems with modern video cards, but how well processors perform with a gaming load in principle. Therefore, based on the results presented, it is quite possible to speculate about how processors will behave in the future, when more quick options graphics accelerators.
In our numerous previous tests, we have repeatedly characterized the Core i5 family of processors as being well suited for gamers. We do not intend to abandon this position now. In gaming applications, Core i5 is strong due to its efficient microarchitecture, quad-core design and high clock speeds. Their lack of support for Hyper-Threading technology can play a good role in games that are poorly optimized for multi-threading. However, the number of such games among the current ones is decreasing every day, which we can see from the results presented. Core i7, based on the Ivy Bridge design, is higher than similar ones in all charts internal structure Core i5. As a result, the gaming performance of the 3,000-series Core i5 is at the expected level: these processors are definitely better than the Core i5 of the 2,000-series, and sometimes they can even compete with the Core i7-2700K. At the same time, we note that AMD’s senior processor cannot compete with modern Intel offerings: its lag in gaming performance can, without any exaggeration, be called catastrophic.
In addition to the gaming tests, we also present the results of the synthetic benchmark Futuremark 3DMark 11, launched with the Performance profile.
The synthetic test Futuremark 3DMark 11 does not show anything fundamentally new either. The performance of the third-generation Core i5 falls exactly between the Core i5 with the previous design and any Core i7 processors that have support for Hyper-Threading technology and slightly higher clock speeds.
Tests in applications
To measure the speed of processors when compressing information, we use the WinRAR archiver, with which we archive a folder with various files with a total volume of 1.1 GB with the maximum compression ratio.
IN latest versions The WinRAR archiver has significantly improved support for multithreading, so that now the archiving speed has become seriously dependent on the number of available CPU computing cores. Accordingly, Core i7 processors, enhanced by Hyper-Threading technology, and eight-core AMD processor The FX-8150 demonstrate the best performance here. As for the Core i5 series, everything is as always with it. Core i5 with Ivy Bridge design is definitely better than the old ones, and the advantage of the new products over the old ones is about 7 percent for models with the same nominal frequency.
Processor performance under cryptographic load is measured by the built-in test of the popular TrueCrypt utility, which uses AES-Twofish-Serpent “triple” encryption. It should be noted that this program not only is it capable of efficiently loading any number of cores with work, but also supports a specialized set of AES instructions.
Everything is as usual, only the FX-8150 processor is again at the top of the chart. It is helped in this by the ability to execute eight computational threads simultaneously and good speed execution of integer and bit operations. As for the Core i5 of the three thousandth series, they are again unconditionally superior to their predecessors. Moreover, the difference in CPU performance with the same declared nominal frequency is quite significant and is about 15 percent in favor of new products with Ivy Bridge microarchitecture.
With the release of the eighth version of the popular scientific computing package Wolfram Mathematica, we decided to return it to the list of used tests. To evaluate the performance of systems, it uses the MathematicaMark8 benchmark built into this system.
Wolfram Mathematica has traditionally been one of the applications that struggles with Hyper-Threading technology. That is why in the above diagram the first position is occupied by the Core i5-3570K. And the results of other Core i5 3000 series are quite good. All these processors not only outperform their predecessors, but also leave behind the older Core i7 with Sandy Bridge microarchitecture.
We measure performance in Adobe Photoshop CS6 using our own test, a creative reworking of the Retouch Artists Photoshop Speed Test, which involves typical processing of four 24-megapixel images taken with a digital camera.
The new Ivy Bridge microarchitecture provides an approximately 6 percent advantage over the similarly clocked third-generation Core i5 over its earlier counterparts. If we compare processors with the same cost, then the carriers of the new microarchitecture find themselves in an even more advantageous position, winning over 10 percent of performance from the Core i5 of the 2000 series.
Performance in Adobe Premiere Pro CS6 is tested by measuring the rendering time in the H.264 Blu-Ray format of a project containing HDV 1080p25 video with various effects applied.
Nonlinear video editing is a highly parallelizable task, so the new Core i5 with Ivy Bridge design is not able to reach the Core i7-2700K. But they outperform their classmate predecessors using the Sandy Bridge microarchitecture by about 10 percent (when comparing models with the same clock frequency).
To measure the speed of video transcoding into the H.264 format, x264 HD Benchmark 5.0 is used, based on measuring the processing time of source video in MPEG-2 format, recorded in 1080p resolution at 20 Mbps. It should be noted that the results of this test are of great practical importance, since the x264 codec used in it underlies numerous popular transcoding utilities, for example, HandBrake, MeGUI, VirtualDub, etc.
The picture when transcoding high-resolution video content is quite familiar. The advantages of the Ivy Bridge microarchitecture result in approximately 8-10 percent superiority of the new Core i5 over the old ones. What’s unusual is the high result of the eight-core FX-8150, which even outperforms the Core i5-3570K in the second encoding pass.
At the request of our readers, the used set of applications has been supplemented with another benchmark that shows the speed of working with high-resolution video content - SVPmark3. This is a specialized test of system performance when working with the SmoothVideo Project package, aimed at improving the smoothness of video by adding new frames to the video sequence containing intermediate positions of objects. The numbers shown in the diagram are the result of a benchmark on real FullHD video fragments without involving the power of the graphics card in the calculations.
The diagram is very similar to the results of the second pass of transcoding with the x264 codec. This clearly suggests that most tasks associated with processing high-definition video content create approximately the same computational load.
We measure computing performance and rendering speed in Autodesk 3ds max 2011 using the specialized test SPECapc for 3ds Max 2011.
To be honest, nothing new can be said about the performance observed in the final rendering. The distribution of results can be called standard.
Testing the final rendering speed in Maxon Cinema 4D is performed using a specialized test called Cinebench 11.5.
The Cinebench results chart doesn't show anything new either. The new Core i5 of the three thousandth series once again turns out to be noticeably better than their predecessors. Even the youngest of them, Core i5-3450, confidently outperforms Core i5-2500K.
Energy consumption
One of the main advantages of the 22-nm process technology used to produce Ivy Bridge generation processors is the reduced heat generation and power consumption of semiconductor crystals. This is also reflected in the official specifications of the third generation Core i5: they are equipped with a 77-watt thermal package rather than a 95-watt one, as before. So the superiority of the new Core i5 over its predecessors in terms of efficiency is beyond doubt. But what is the scale of this gain in practice? Should the efficiency of the 3,000-series Core i5 series be considered a serious competitive advantage?To answer these questions, we conducted special testing. The new Corsair AX1200i digital power supply we use in our test system allows us to monitor the electrical power consumed and output, which is what we use for our measurements. The following graphs, unless otherwise noted, show the total system consumption (without monitor), measured “after” the power supply and representing the sum of the power consumption of all components involved in the system. The efficiency of the power supply itself is not taken into account in this case. During measurements, the load on the processors was created by the 64-bit version of the LinX 0.6.4-AVX utility. In addition, to properly estimate idle power consumption, we activated turbo mode and all available energy-saving technologies: C1E, C6 and Enhanced Intel SpeedStep.
When idle, systems with all processors participating in the tests show approximately the same power consumption. Of course, it is not completely identical, there are differences at the level of tenths of a watt, but we decided not to transfer them to the diagram, since such an insignificant difference is more likely related to measurement error than to the observed physical processes. In addition, in conditions of similar processor consumption values, the efficiency and settings of the power converter begin to have a serious impact on the overall power consumption motherboard. Therefore, if you are really concerned about the amount of power consumption at rest, you should first look for motherboards with the most efficient power converter, and, as our results show, any processor from among the LGA 1155-compatible models can be suitable.
A single-threaded load, in which processors with turbo mode increase the frequency to maximum values, leads to noticeable differences in consumption. The first thing that catches your eye is the completely immodest appetites of the AMD FX-8150. As for LGA 1155 CPU models, those based on 22 nm semiconductor crystals are indeed noticeably more economical. The difference in consumption between quad-core Ivy Bridge and Sandy Bridge, operating at the same clock speed, is about 4-5 W.
The full multi-threaded compute load exacerbates the consumption differences. The system, equipped with third-generation Core i5 processors, is more economical than a similar platform with processors of the previous design of about 18 W. This perfectly correlates with the difference in theoretical heat dissipation figures declared by Intel for its processors. Thus, in terms of performance per watt, Ivy Bridge processors have no equal among desktop CPUs.
GPU performance
When considering modern processors for the LGA 1155 platform, attention should also be paid to the graphics cores built into them, which with the introduction of the Ivy Bridge microarchitecture have become faster and more advanced in terms of available capabilities. However, at the same time, Intel prefers to install in its processors for the desktop segment a stripped-down version of the video core with the number of actuators reduced from 16 to 6. In fact, full graphics are present only in Core i7 and Core i5-3570K processors. Most of the 3,000-series Core i5 desktops will obviously be quite weak in 3D graphics applications. However, it is quite likely that even the existing reduced graphics power will satisfy a certain number of users who do not intend to consider the integrated graphics as a 3D video accelerator.We decided to start testing integrated graphics with the 3DMark Vantage test. Results obtained in different versions of 3DMark are a very popular metric for assessing the weighted average gaming performance of video cards. The choice of the Vantage version is due to the fact that it uses DirectX version 10, which is supported by all video accelerators tested, including the graphics of Core processors with Sandy Bridge design. Note that in addition to the full set of processors of the Core i5 family working with their integrated graphics cores, we included in the tests and performance indicators of systems based on the Core i5-3570K with a discrete graphics card Radeon HD 6570. This configuration will serve as a kind of benchmark for us, allowing imagine the place of Intel graphics cores HD Graphics 2500 and HD Graphics 4000 in the world of discrete video accelerators.
The HD Graphics 2500 graphics core installed by Intel in most of its desktop processors is similar in 3D performance to the HD Graphics 3000. But the older version of Intel graphics from Ivy Bridge processors, HD Graphics 4000, looks like a huge step forward, its performance is more than doubled exceeds the speed of the best embedded core of the previous generation. However, any of the available Intel HD Graphics options cannot yet be called having acceptable 3D performance by desktop standards. For example, a Radeon HD 6570 video card, which belongs to the lower price segment and costs about $60-70, can offer significantly better performance.
In addition to the synthetic 3DMark Vantage, we also ran several tests in real gaming applications. In them, we used low graphics quality settings and a resolution of 1650x1080, which we currently consider to be the minimum of interest to desktop users.
In general, the games show approximately the same picture. The older version of the graphics accelerator built into the Core i5-3570K provides an average number of frames per second at a fairly good level (for an integrated solution). However, the Core i5-3570K remains the only third-generation Core i5 processor whose video core is capable of delivering acceptable graphics performance, which, with some relaxations in picture quality, may be enough to comfortably play a significant number of current games. All other CPUs in this class, which use the HD Graphics 2500 accelerator with a reduced number of execution units, produce almost half the speed, which is clearly not enough by modern standards.
The advantage of the HD Graphics 4000 graphics core over the built-in accelerator of the previous generation HD Graphics 3000 varies widely and averages about 90 percent. The previous flagship integrated solution can easily be compared with the younger version of graphics from Ivy Bridge, HD Graphics 2500, which is installed in most Core i5 desktop processors of the three thousandth series. As for the previous version of the commonly used graphics core, HD Graphics 2000, its performance now looks extremely low; in games it lags behind the same HD Graphics 2500 by an average of 50-60 percent.
In other words, the 3D performance of the graphics core of Core i5 processors has indeed increased significantly, but compared to the number of frames that the Radeon HD 6570 accelerator is capable of producing, all this seems like fuss. Even the HD Graphics 4000 accelerator built into the Core i5-3570K is not very good alternative low-level desktop 3D accelerators, while the more common version of Intel graphics, one might say, is generally not applicable for most games.
However, not all users consider the video cores built into processors as 3D gaming accelerators. A significant portion of consumers are interested in HD Graphics 4000 and HD Graphics 2500 due to their media capabilities, which simply do not have alternatives in the lower price category. Here we primarily mean Quick Sync technology, designed for fast hardware encoding video in AVC/H.264 format, the second version of which is implemented in processors of the Ivy Bridge family. Since Intel promises a significant increase in transcoding speed in new graphics cores, we separately tested the functioning of Quick Sync.
During practical tests We measured the transcoding completion time of one 40-minute episode of a popular TV series encoded in 1080p H.264 at 10 Mbps for viewing on an Apple iPad2 (H.264, 1280x720, 3Mbps). For the tests, we used the Cyberlink Media Espresso 6.5.2830 utility, which supports Quick Sync technology.
The situation here is radically different from what was observed in the games. If previously Intel did not differentiate Quick Sync in processors with different versions graphics core, now everything has changed. This technology in HD Graphics 4000 and HD Graphics 2500 operates at approximately twice the speed. Moreover, conventional Core i5 processors of the three thousand series, in which the HD Graphics 2500 core is installed, transcode high-resolution video via Quick Sync with approximately the same performance as their predecessors. Progress in performance is visible only in the results of the Core i5-3570K, which has an “advanced” HD Graphics 4000 graphics core.
Overclocking
Overclocking Core i5 processors belonging to the Ivy Bridge generation can proceed according to two fundamentally different scenarios. The first of them concerns overclocking the Core i5-3570K processor, which was initially aimed at overclocking. This CPU has an unlocked multiplier, and increasing its frequency above the nominal values is carried out according to a typical algorithm for the LGA 1155 platform: by increasing the multiplication factor, we raise the processor frequency and, if necessary, achieve stability by feeding it to the CPU high voltage and improving its cooling.Without raising the supply voltage, our copy of the Core i5-3570K processor overclocked to 4.4 GHz. All that was required to ensure stability in this mode was simply switching the motherboard's Load-Line Calibration feature to High.
An additional increase in the processor supply voltage to 1.25 V made it possible to achieve stable operation at a higher frequency - 4.6 GHz.
This is a quite typical result for Ivy Bridge generation CPUs. Such processors usually overclock a little worse than Sandy Bridge. The reason is believed to lie in the reduction in the area of the semiconductor processor chip that followed the introduction of 22-nm production technology, raising the question of the need to increase the heat flux density during cooling. At the same time, the thermal interface used by Intel inside processors, as well as the commonly used methods of removing heat from the surface of the processor cover, do not help solve this problem.
However, be that as it may, overclocking to 4.6 GHz is a very good result, especially if you take into account the fact that Ivy Bridge processors at the same clock frequency as Sandy Bridge produce approximately 10 percent better performance due to their microarchitectural improvements.
The second overclocking scenario concerns the remaining Core i5 processors, which do not have a free multiplier. Although the LGA 1155 platform has an extremely negative attitude towards increasing the frequency of the base clock generator, and loses stability even when the generating frequency is set 5 percent higher than the nominal value, it is still possible to overclock Core i5 processors that are not related to the K-series. The fact is that Intel allows you to increase their multiplier to a limited extent, increasing it by no more than 4 units above the nominal value.
Considering that the Turbo Boost technology remains operational, which for Core i5 with Ivy Bridge design allows for 200 MHz overclocking even when all processor cores are loaded, the clock frequency can generally be “increased” by 600 MHz above the standard value. In other words, the Core i5-3570 can be overclocked to 4.0 GHz, the Core i5-3550 to 3.9 GHz, the Core i5-3470 to 3.8 GHz, and the Core i5-3450 to 3.7 GHz. This we have successfully confirmed during our practical experiments.
Core i5-3570:
Core i5-3550:
Core i5-3470:
Core i5-3450:
It must be said that such limited overclocking is even easier than with the Core i5-3570K processor. A not so significant increase in clock frequency does not entail stability problems even when using the rated supply voltage. Therefore, most likely, the only thing required to overclock Ivy Bridge processors of the Core i5 line, not related to the K-series, is to change the multiplier value in Motherboard BIOS fees. The result achieved in this case, although it cannot be called a record, will most likely be quite satisfactory for the vast majority of inexperienced users.
conclusions
We have already said more than once that the Ivy Bridge microarchitecture has become a successful evolutionary update Intel processors. 22nm semiconductor manufacturing technology and numerous microarchitectural improvements have made the new products both faster and more cost-effective. This applies to any Ivy Bridge in general and to the 3,000-series Core i5 desktop processors discussed in this review in particular. Comparing the new line of Core i5 processors with what we had a year ago, it is not difficult to notice a whole bunch of significant improvements.Firstly, the new Core i5, based on the Ivy Bridge design, has become more productive than its predecessors. Despite the fact that Intel has not resorted to increasing clock speeds, the advantage of new products is about 10-15 percent. Even the slowest third-generation Core i5 desktop processor, the Core i5-3450, outperforms the Core i5-2500K in most tests. And the older representatives of the new line can sometimes compete with higher-class processors, Core i7, based on the Sandy Bridge microarchitecture.
Secondly, the new Core i5 has become noticeably more economical. Their thermal package is set at 77 Watt, and this is reflected in practice. Under any load, computers using Core i5 with Ivy Bridge design consume several watts less than similar systems using Sandy Bridge CPUs. Moreover, with the maximum computing load, the gain can reach almost two dozen watts, and this is a very significant saving by modern standards.
Thirdly, the new processors have a significantly improved graphics core. The junior version of the graphics core of Ivy Bridge processors works at least as well as the HD Graphics 3000 from the older second-generation Core processors, and besides, supporting DirectX 11, it has more modern capabilities. As for the flagship integrated accelerator HD Graphics 4000, which is used in Core processor i5-3570K, then it even allows you to get quite acceptable frame rates in fairly modern games, albeit with significant relaxations in the quality settings.
The only controversial point that we noticed with the third-generation Core i5 is its slightly lower overclocking potential than that of Sandy Bridge-class processors. However, this drawback manifests itself only in the only overclocking model Core i5-3570K, where the change in the multiplication coefficient is not artificially limited from above, and moreover, it is fully compensated by the higher specific performance developed by the Ivy Bridge microarchitecture.
In other words, we don’t see any reason why, when choosing a mid-class processor for the LGA 1155 platform, preference should be given to “oldies” using semiconductor crystals of the Sandy Bridge generation. Moreover, the prices set by Intel for more advanced modifications of the Core i5 are quite humane and close to the cost of aging processors of the previous generation.
08/21/2017, Mon, 09:36, Moscow time , Text: Vladimir Bakhur
Intel announced the addition of eighth-generation Core chips to its line of U-series mobile processors. A new generation of Coffee Lake processors for desktop PCs will also appear this year, but later.The first four processors of the new eighth generation
Intel introduced four new Core i5 and Core i7 mobile processors in the U line. All new chips have four computing cores with support for Hyper-Threading technology, which in total allows for up to eight computing threads per chip.
Previous generations of mobile Core processors were released with two physical cores and supported four threads with Hyper-Threading technology.
The official working name of the new mobile processors is Kaby Lake Refresh, that is, they are based on the improved seventh generation Kaby Lake architecture.
All 8th generation Core processors (Kaby Lake Refresh) presented today, like their predecessors, are manufactured in compliance with the 14 nm technological process, but “with improved characteristics,” which led to the announcement of the new 8th generation. According to Intel, the transition to 10 nm process standards will take place later in the fall, but within the same eighth generation.
The “real” next-generation architecture, working title Coffee Lake, will be presented even later and will join the list of 8th generation Core chips. However, these chips will also be produced according to 14 nm standards.
New 8th generation Intel Core processors
The transition to 10nm standards will be the next step and will debut with the Cannon Lake architecture. Thus, the list of eighth-generation Core processors will include i7/i5/i3-8xxx chips of three different architectures: Kaby Lake Refresh, Coffee Lake and Cannon Lake. Previously, there were usually two types of architectures per Core generation.
Architecture Details
The new eighth-generation Core processors operate at relatively low main clock frequencies (no higher than 1.9 GHz for the older i7-8650U model), thanks to which all models fit into a thermal package (TDP) of up to 15 W with four computing cores.
Appearance of the 8th generation Core processor
At the same time, thanks Intel technologies Turbo Boost Technology 2.0, the chips are capable of dynamically increasing the clock frequency by more than twice (up to 4.2 GHz for the older model i7-8650U), which allows you to significantly increase system performance as needed and remain in a “cold” state in standby mode.
Basic characteristics of the first four 8th generation Core processors
All new mobile processors The 8th generation Intel Core is equipped with an integrated graphics core Intel UHD Graphics 620 with support for up to three independent displays, inherited with some changes from the 7th generation processors (Kaby Lake, Intel HD Graphics 620 graphics). The built-in UHD Graphics 620 supports HEVC and VP9 codecs and allows you to work with 4K video with 10-bit color depth.
Photo of the chip of the new 8th generation Intel Core chip
The new 8th generation mobile processors received 8 MB or 6 MB L3 cache, as well as a fast 2-channel memory controller with support for DDR4-2400 and LPDDR3-2133 modules.
About productivity and savings
According to the company's internal tests, the new eighth-generation Core i7 and i5 mobile chips provide performance gains of up to 40% compared to the previous generation chips, and are twice as fast as chips from five years ago, for example, when comparing the new Core i5-8250U with the Core i5- 3317U.
At the end of the summer of this year, new U-series processors based on the Kaby Lake Refresh architecture were released on the market. New items are designed for laptops and other mobile devices and are built on the 14 nm+ process technology, having two cores each. The American manufacturer did not say anything about the timing of the appearance of desktop models of the new series, indicating that the new items would be available soon. Today, September 25, after almost a month, Intel held a presentation of the eighth generation Core desktop processors for PCs and at the same time announced their release date. The line is already known to us as Coffee Lake.
Traditionally, the new line is represented by three main models: manufacturers are offered Core i3, Core i5 and the flagship Core i7. All presented processors have switched to an updated 14 nm++ process technology and an increased number of cores compared to Kaby Lake Refresh: Core i3 is now quad-core (for the first time in history), and Core i5 and Core i7 are six-core. In addition to the classic series, Intel will also sell unlocked versions of the chips with the "K" suffix. These processors support up to 40 PCIe 3.0 lanes per socket, 4K HDR, and Thunderbolt 3.0. The motherboard uses the new Intel Z370 chip ( dynamic memory DDR4-2666, built-in USB 3.1 with data transfer speeds up to 5 Gbps).
Technical characteristics of the new eighth generation Intel Core processors for PCs:
- Core i7-8700K: 6 cores / 12 threads, clock speed from 3.8 GHz (base) to 4.7 GHz (in Turbo mode Boost), 12 MB L3 cache, 95 W TDP.
- Core i7-8700: 6 cores / 12 threads, clock speed from 3.2 GHz (base) to 4.6 GHz (Turbo Boost), 12 MB L3 cache, 65 W TDP.
- Core i5-8600K: 6 cores / 6 threads, clock speed from 3.6 GHz (base) to 4.3 GHz (Turbo Boost), 9 MB L3 cache, 95 W TDP.
- Core i5-8400: 6 cores / 6 threads, clock speed from 2.8 GHz (base) to 4.0 GHz (Turbo Boost), 9 MB L3 cache, 65 W TDP.
- Core i3-8350K: 4 cores/4 threads, 4.0 GHz base clock, 6 MB L3 cache, 91 W TDP.
- Core i3-8100: 4 cores/4 threads, 3.6 GHz base clock, 6 MB L3 cache, 65 W TDP.
It's not easy to surprise the many denizens of technology forums all over the Internet. When Intel recently released its 6-core 8th generation Core processors, many were not impressed. In their opinion, Intel offers slightly redesigned old products with a new cover.
Perhaps the new processors have become derivatives of the previous ones, but this does not detract from their advantages. There are enough differences that many reviewers are calling them worthy of an upgrade from previous-generation chips. This hasn't happened often in recent years. In support of this point of view, test results will be given below.
What are 8th generation Intel Core?
As usual, understanding Intel products is not at all easy. First came the 8th generation Core i7 Coffee Lake S for desktops. Then came the 8th generation Core i7 Kaby Lake R for ultraportable laptops. Why they weren't called Coffee Lake U is unknown.
Now we are talking about the 8th generation Core i7 Coffee Lake H for larger and gaming laptops. They can be considered an improved version of the 6th generation Skylake processors, which appeared in laptops back in 2015.
Since then, engineers have made many improvements. For example, Kaby Lake's video processing engine has been significantly improved. Clock speeds have also increased compared to Skylake. The 14 nm process technology was finally brought to fruition, earning the title 14++.
MSI GS65 Stealth Thin RE
How the testing was performed
IN desktop computers You can control cooling, power consumption, memory and disk space. Laptops do not have this freedom, which significantly affects productivity. Some laptops may be aimed at maximum speed, others at maximum silence. The cooling system plays a role, and the size of the case depends on it.
This case compares the MSI GS65 Stealth Thin laptop with a 6-core processor with a 17-inch Lenovo Legion Y920. The latter runs on a 4-core Core i7-7820HK, which is an unlocked chip with overclocking capabilities.
The past generation represents Asus ROG Zephyrus GX501. This is a 17-inch laptop, very thin and powered by a 4-core Core i7-7700HQ processor.
6-core Core i7-8750H in MSI GS65 Stealth Thin
Performance
All three laptops use different GPUs. The Lenovo Legion Y920 has a GeForce GTX 1070, the Asus ROG Zephyrus GX501 has a GeForce GTX 1080 Max-Q, and the MSI GS65 Stealth Thin uses a GeForce GTX 1060.
Because of this disparity, graphics performance receives little attention. In this case, the emphasis is on central processors.
This benchmark is built on the Maxon Cinema4D engine and prefers more cores. As a result, the transition from 4 to 6 cores provides a fairly large performance increase. Similar results can be expected in all applications using 6 cores or 12 instruction threads of the Core i7-8750H.
Overclocked Core i7-7820HK lags behind Core i7-8750H
True, not all applications support multithreading. Of these, few are effective enough to show the results shown in the graph above. Without 3D graphics, video editing and other demanding tasks, it's better to look at the single-threaded performance of laptop processors.
That's exactly what was done, reviewers tested Cinebench R15 using a single command stream. The results have leveled off, but new processor still in the lead. Even against the overclocked Core i7-7820HK it has a 7% advantage. Compared to the Core i7-7700HQ in the Asus ROG Zephyrus GX501, the difference is 13%.
Leadership through higher frequency
Benchmark based on the Corona Photorealistic renderer for Autodesk 3ds Max. Like Cinebench and most rendering applications, it loves a lot of cores. As a result, 6 cores are again better than 4.
The latest rendering benchmark measures the processing time per frame. Here the difference is not so significant. Perhaps it's the length of the tests. Cinebench and Corona last a couple of minutes, Blender about 10 minutes.
When the processor in a laptop heats up, the clock speed begins to decrease. The Core i7-8750H has an advantage in the number of cores and clock speed. With continued use, this benefit begins to diminish. For the same reason, the nominal frequencies on the Core i7-7820HK are not impressive, while when overclocked the processor is much closer to the Core i7-8750H.
Encoding speed
Used MKV file 30 GB 1080p, HandBrake 9.9 and Android Tablet profile. Here the process took about 45 minutes on a 4-core laptop, because of this the difference in frequency is minimized. Under long-term workloads, you can see the value of the extra cores: the new processor completed encoding in about 33 minutes versus 46 minutes on the Core i7-7700HQ.
Compression speed
The internal WinRAR benchmark is used. The first results are single-threaded, so the higher frequency of the Core i7-8750H gave it an advantage. True, the advantage is small.
Single thread performance
The Core i7-7700HQ in the Asus ROG Zephyrus GX501 performed poorly, despite several attempts. Since its performance in the remaining tests was at the expected level, memory may be to blame. Asus uses 16GB in one slot and 8GB in the other, so dual-channel mode may not always be enabled. In WinRAR throughput memory plays an important role.
Multi-threaded performance
Multithreaded mode showed the expected results. The advantage of the new processor immediately became overwhelming, and the Core i7-7700HQ showed normal results.
Performance Analysis
So, the Core i7-8750H has more cores and a higher clock speed. Repeated testing of Cinebench R15 was performed with the number of threads from 1 to 12 on the Core i7-8750H and from 1 to 8 on the Core i7-7700HQ.
The results are not very consistent with the actual performance differences. The graph below shows this difference more clearly. As you can see, the more threads, the higher the difference, which ultimately reaches 50%.
Coffee Lake H has the same architecture as Kaby Lake H, so the only difference is the increased clock speeds. For a more detailed analysis, Cinebench R15 was launched again and the number of threads was increased. Clock speed has been analyzed for some time.
Core i7-8750H runs at more high frequencies under light loads compared to the Core i7-7700HQ. The further to the right, the more the processors heat up, the difference is minimized.
Conclusion
In recent years, there has been no reason to change processors and laptops. For example, if you had a 5th generation Core i7, there was no point in upgrading to the 6th generation. The performance difference was only 6%-7%. This is no longer the case.
When upgrading from a 7th-gen Core i7 laptop to an 8th-gen Core i7, there's a more substantial jump in performance for video editing, graphics processing, and other heavy-duty tasks. This is visible even under low load, but is especially noticeable under high load.
Of course, for many users, what they have is enough. You don't need much for Word and the browser, so you need to understand whether you need increased performance or not.
Loading...
