Processors. Processors Maximum Clock Speed ​​with Turbo Boost Technology

Intel's tick-tock principle, which describes the ideology of alternately introducing new microarchitectures and introducing more subtle technical processes, continues to operate. Initially, the company promised to release new products every year, and I must say, in general, it adheres to this plan. Last year we were presented with microarchitecture Sandy bridge, which significantly increased the performance modern computers And now Intel is launching the Ivy Bridge project, an advanced processor design that uses new manufacturing technology with 22nm norms and innovative 3D transistors.

However, the weakening of competition in the market of high-performance processors still cannot but affect the pace of progress. The pendulum of Intel's concept is gradually slowing down, and if Sandy Bridge was presented at the very beginning of 2011, then we had to wait until the end of April for the announcement of Ivy Bridge. However, Intel has a good excuse: the new generation of processors is not a simple cosmetic rework of the old core, taking into account new technological standards. Engineers have made a number of significant changes to the microarchitecture, so Ivy Bridge is proposed to be counted not as one "tick", but as a "tick" and another "half a tak" in the bargain.

Can such an explanation for the delay that has arisen be accepted? It all depends on the position from which to evaluate modern processors in general. Most of the changes that have occurred in the design of Ivy Bridge are not about the computing cores, but the graphics core. Therefore, for traditional CPUs, this is a clear "tick". However, if we consider that the paradigm of heterogeneous processors proposed by AMD turned out to be yet another prophecy (unlike microarchitectures, AMD clearly succeed), then Ivy Bridge can pull on a full-fledged "so".

So it turns out that the new Intel product is a very multifaceted and contradictory thing. Desktop enthusiasts who see Ivy Bridge as a possible incentive to upgrade their systems will likely be disappointed. For them, there is nothing particularly attractive about it, since a simple transition to new technology production itself does not add anything special. Moreover, the "thinning" of the technical process has long resulted not in an increase in CPU clock frequencies, but in a decrease in their heat release.

But for users of various kinds of mobile or compact systems, Ivy Bridge promises a very good deal. Finally, about the representatives of the series Intel Core you can think of it as full-fledged hybrid processors - APUs that provide good 3D performance, are compatible with DirectX 11 and are capable of performing GPGPU computing. No wonder it is with the release of Ivy Bridge that Intel directly connects the flourishing of ultrabooks - new items fit into this class of computers almost perfectly.

However, in this article we will position ourselves as enthusiasts of the old school. Any ultra-compact computers are children's toys, give us traditional computing systems that inspire respect for both their appearance and the level of performance. Can Ivy Bridge fit organically into such an ecosystem? Let's try to answer this question.

⇡ Ivy Bridge Microarchitecture: An Overview

Although we said that the Ivy Bridge microarchitecture has significant differences from its predecessor, Sandy Bridge, it is easy to see the close relationship between them. At the highest level, nothing has changed in the general structure of the new processors, all the improvements made are in the details. Detailed description innovations can be found in a special material, here we will give short review key points.

Perhaps to start with, the emergence of new Ivy Bridge processors does not mean a platform change. These CPUs use the same LGA1155 processor socket as their predecessors and are fully compatible with the existing motherboard fleet. Intel has timed the release of Ivy Bridge with the appearance of the seventh series chipset family led by the Z77, but the use of motherboards based on it together with new processors is not necessary. To connect the Ivy Bridge to the system logic set, the same DMI 2.0 bus with a bandwidth of 20 Gb / s is used as in the case of Sandy Bridge. Therefore, the new processors work perfectly in any motherboard with the LGA1155 socket.

Like Sandy Bridge, the Ivy Bridge processor family consists of the same set functional units... They contain two or four computing cores equipped with an individual 256 KB L2 cache; graphics core; shared cache memory of the third level up to 8 MB; dual-channel memory controller with DDR3 SDRAM support; graphics bus controller PCI Express; as well as a system agent responsible for the Turbo technology and implementing auxiliary interfaces. All the components of the Ivy Bridge are connected via the Ring Bus - nothing new here either.

If we talk about the differences between the Ivy Bridge and its predecessors, then this is primarily a new 22-nm production technology used by the manufacturer for the manufacture of semiconductor crystals. Moreover, the novelty in this case lies not only in the "refined" norms, but also in a fundamental change in the internal design of transistors. Intel characterizes the new transistors as having a three-dimensional design (Tri-Gate), which in practice results in installing a high high-K dielectric coated vertical rib on a silicon substrate, cutting into the gate.

Considering that one of the main goals of the Ivy Bridge release is their massive penetration into ultra-mobile computers, this improvement in economy is by no means superfluous. In addition, Intel developers have strengthened the achieved effect by introducing new energy-saving technologies: deeper sleep states, the ability to disconnect from the memory controller power lines and support for DDR3L SDRAM with reduced voltage. There was also such a concept as a configurable TDP. As a result, among the various modifications of Ivy Bridge, there is a whole class of ULV products with a 17-W thermal package, which is reduced if necessary to 14 W.

The commissioning of a fresh production technology automatically means a reduction in the size of semiconductor crystals. So, the crystal of the quad-core Ivy Bridge has an area of ​​160 sq. mm is 35% less than the Sandy Bridge area.

At the same time, the complexity of the new processor has grown significantly, it consists of 1.4 billion transistors, while the number of transistors in the predecessor processors of the same class was 995 million pieces.

The most common way to use an additional transistor budget is to increase the amount of cache memory. However, there is nothing like that in Ivy Bridge, these processors have exactly the same capacity and L1-, L2- and L3-caches as in Sandy Bridge. The additional transistors, for the most part, went into the integrated graphics core - in Ivy Bridge it differs from the graphics of the previous generation, Intel HD Graphics 3000/2000, a little less than completely.

The new video core, dubbed HD Graphics 4000, can finally be called modern in every sense of the word. The main achievement of the developers is that with new version graphics they were able to meet the requirements of DirectX 11 with DirectCompute and Shader Model 5.0, and also opened the possibility of GPGPU computing through the OpenCL 1.1 interface. In addition to this, HD Graphics 4000 has support for three independent monitors, and the level of performance has significantly increased due to the addition of additional executive devices: now there are 16 instead of 12. Therefore, Intel believes that the number of systems using the company's processors without an external video card will increase significantly, however this will happen mainly in the mobile market segment.

But for desktop users, the graphics core is not very interesting. Much more they expect improvements in the microarchitecture of the computing part, which can affect performance. And here the new Ivy Bridge generation processors have nothing to boast of. Possible performance gains when Ivy Bridge and Sandy Bridge operate at the same clock frequency, even according to the most optimistic official data, does not exceed 5%. The fact is that the computational cores in the new processors have not been reworked, and there are only minor cosmetic improvements. So, in Ivy Bridge, the operation of integer and real division instructions is accelerated, taking into account the use of a register file, the execution of instructions for transferring data between registers is optimized, in addition, a dynamic, rather than static, distribution of internal buffer resources between threads is implemented when using Hyper-Threading technology.

To evaluate the practical effect of these changes, we used synthetic benchmarks from SiSoft Sandra, which implement simple algorithms that allow us to assess the performance of processors when performing a variety of operations. As part of this preliminary test, we compared the performance of the four-core Sandy Bridge and Ivy Bridge, operating at the same 4.0 GHz frequency without using Hyper-Threading technology.

The results are really not very encouraging. Improvements to the microarchitecture of computing cores in Ivy Bridge translate into almost elusive performance gains.

Therefore, the changes that affected the operation of adjacent internal processor interfaces - memory and PCI bus Express. Thus, the PCI Express controller built into the Ivy Bridge received support for the third version of this specification, which automatically (provided that compatible end devices are used) means an increase in the bus bandwidth in comparison with PCI Express 2.0 almost twice - up to 8 Gtransactions per second.

At the same time, the sixteen PCI Express lanes supported by Ivy Bridge can be split into two or three parts - according to the 8x + 8x or 8x + 4x + 4x scheme. The latter option may be interesting for systems with three video cards, especially since PCI Express 3.0 is quite capable of providing acceptable bandwidth for video cards even if only four lanes are used.

As for the Ivy Bridge memory controller, its baseline specs have not changed from what we saw in Sandy Bridge. It can also work with dual channel DDR3 SDRAM. But at the same time, Intel engineers took certain steps towards overclocking memory manufacturers and added the ability to more flexible frequency settings to the processor. First, the maximum supported frequency is now DDR3-2800 SDRAM. Secondly, to change the memory frequency, you can now use two clock modes - with a step of 200 or 266 MHz.

The practical speed of the memory controller has also changed slightly. This is also confirmed by benchmarks. For example, below we show the AIDA64 Cache & Memory Benchmark measured on a system with Sandy Bridge and Ivy Bridge processors clocked at 4.0 GHz.

Sandy Bridge 4.0 GHz, DDR3-1867 (9-11-9-30-1T)

Ivy Bridge 4.0 GHz, DDR3-1867 (9-11-9-30-1T)

The Ivy Bridge generation processor provides a slightly lower practical memory subsystem latency, but this advantage is minimal. At the same time, the test reveals another interesting detail: the L3 cache of the new processors allegedly became noticeably faster. However, we are forced to disappoint - in this case, the difference in AIDA64 Cache & Memory Benchmark is caused not by an improvement in the speed characteristics of the L3 cache, but by changes in the speed of execution of instructions appearing in the test algorithm. In fact, the latency of the Ivy Bridge L3 cache is 24 cycles - and this is one cycle more than the L3 cache latency of the Sandy Bridge processors. In other words, the cache in the new processors began to work even a little slower than before, but this is not noticeable in practical tasks.

⇡ Processors Ivy Bridge for desktops, first run

Manufacturing problems that arise almost every time it comes to introducing any fundamental innovations have not yet allowed Intel to flood the market with variegated modifications of Ivy Bridge. Therefore, the introduction of the new design is taking place in stages: today only quad-core modifications of new processors are announced, belonging to the Core i7 and Core i5 families.

There are only five desktop models, the following table reveals their specifications.

To be honest, familiarity with the above characteristics does not add much optimism about the new processors. Compared to Sandy Bridge, we see no progress either in the number of cores, or in clock speeds, or in the size of the cache memory. And since the new microarchitecture practically does not increase the number of instructions processed per clock cycle, it becomes clear: according to traditional processor concepts the lineup Ivy Bridge is an evolutionary update to Sandy Bridge. There are only two positive points: an attractive graphics core for certain categories of users and reduced heat emission.

By the way, a very funny incident is connected with the TDP characteristic. Although in the official documentation the typical heat dissipation of new processors is indicated as 77 W, on boxes with real products Intel writes "95 W". This inconsistency has already generated a lot of ridiculous judgments, but in fact the explanation is very simple. In reality, the observed heat dissipation does not go beyond the 77-watt limit, but such a TDP value has not been used before, so Intel decided not to complicate the life of users, component manufacturers and system builders and will indicate a well-known number on the boxes. In addition, as we managed to find out from the company representatives, in the future, it is possible to release more high-speed Ivy Bridge models, which will bring the real and formal TDP to a single denominator.

There are no fundamental changes in the general structure of the proposals. Older LGA1155 processors of the new formation are aimed at advanced users and have the letter "K" in their index. Such offers have a free multiplier and are open to overclocking experiments. Other models Core i7 and Core i5, as before, do not allow increasing the multiplication factor by more than four units.

The lack of striking revolutionary changes in the computing performance of the new processors did not deter Intel from assigning them numbers from the three thousandth series. Thus, in the structure of Intel's Ivy Bridge proposals for LGA1155 systems, they become Sandy Bridge-E processors for LGA 2011 and replace the 2000th Sandy Bridge. This is indicated by the prices. New items are not more expensive than Core a year ago, so the usual course of processor life, when generations of Intel CPUs successively replace each other, will not be broken this time either.

For testing, Intel provided us with samples of senior processors in updated lines. Core third generations: Core i7-3770K and Core i5-3570K.

Please note that 22nm manufacturing technology is clearly visible through practical aspects operation of new products. Their operating voltage has dropped from Sandy Bridge by about 15-20 percent and is now around 1.0 V. This is one of the main reasons for the lower heat generation.

Thanks to the Enhanced Intel SpeedStep and C1E power saving technologies, the Ivy Bridge voltage drops to about 0.9V while idle, and the frequency drops to 1.6GHz.

Everything has remained as before.

Ivy Bridge Review | Introduction

Ivy bridge- this is a new "teak" - already known architecture Sandy bridge on a reduced 22 nanometer crystal. but Intel calls this cycle "tick-plus" because there are some internal improvements in the logic.

Unfortunately for desktop enthusiasts, most of the changes are about the integrated graphics engine, which most simply don't use.

Naturally in the sphere mobile devices the situation is quite different. Here, lower power consumption and "reasonably fast" graphics provide longer battery life and acceptable performance levels. However, in today's review, we will not touch on mobile versions processors. We present to your attention a model Core i7-3770K with an unlocked multiplier to replace existing Core i7-2700K processors and Core i7-2600K .

Ivy Bridge Review | Updated architecture

Processor based on Ivy Bridge microarchitecture

Everything indicates that Ivy bridge Is another highly integrated architecture from Intel... Independent teams from all over the world worked on its components: engineers from Israel developed the IA kernels, a team from Folsom (California) created a graphics engine, a second team in Folsom implemented connections, cache and a system agent. And of course the development team at Origon made sure it was assembled and run on a 22nm core.

What is the new architecture capable of? Let's get to know the architecture step by step Ivy bridge and we will try to identify its pros and cons.

Ivy Bridge Review | Familiar core

The company needed to rethink its approach to implementing integrated graphics, which allows Intel not only stick to a more aggressive plan to improve graphics in the future, but also fix some of the flaws seen in the architecture Sandy bridge... As a result, the architecture was divided into five areas.

1. The first area includes global tools such as channel geometry. The hull (HS) and domain shader (DS) programmable components complement the fixed function tessellation block required to support DirectX 11.
2. Second area Intel calls Slice Common. It contains rasterization units, pixel pipelines (pixel back-ends) and a third-level cache. V Sandy bridge there was no separate L3 cache for graphics because Intel could not get a significant level of performance from him. The processor's ring bus provides enough bandwidth that L3 cache can handle quite well. But since in Ivy bridge more emphasis on graphics, a separate L3 cache complements the bandwidth requirements while lowering power consumption when the engine runs on its own storage rather than across the entire bus.
3. The third area is called Slice. It includes shaders, texture units, texture samplers, L1 cache for instructions and a media sampler that uses the technology Quick Sync... This set Intel plans to use to increase performance in the future. It can also work with the Slice Common add-on to increase bandwidth.
4. The fourth area consists of fixed function media components. It can also be scaled up depending on how deep Intel wants to work with the performance of a media resource.
5. The display outputs make up the last area. There are three digital outputs available on the desktop platform (which must be provided by the vendor), two of which must be DisplayPort connectors, one supporting 2560x1600, the other 1920x1200. The third screen can be connected via HDMI (up to 1080p), DVI, VGA or DisplayPort with a maximum resolution of 1920x1200.

It's no secret that many overclockers have been looking forward to the release of the latest 22nm Intel Ivy Bridge processors. There are several reasons for this.

Few would argue with the fact that Intel in recent years has managed to provide a very noticeable lead over its eternal rival - AMD, in terms of pure performance. specific models processors, and in terms of absolute "performance per clock". In the lower and middle price ranges, there is still a real struggle (mainly due to AMD's aggressive pricing policy), but there is no trace of competition in the top segment: apart from Sandy Bridge and Sandy Bridge-E, there is essentially nothing to buy.

The past generation of Intel processors has been particularly successful. 32-nm Sandy Bridge deservedly received a solid "residence" in the system units of most enthusiasts. What was the reason for this?

First, the new architecture, thanks to numerous optimizations, turned out to be very successful. The old 45nm Bloomfield (remember the widespread Core i7-920?) Were pretty good too. So much so that to this day they are suitable for solving the vast majority of tasks and can work even in very powerful gaming computers... However, Sandy in many tests showed a noticeable advantage over equal-frequency processors with the Nehalem architecture.

Secondly, there was no talk of "equal frequencies". The new CPUs made it possible to achieve unprecedented frequencies "in the air": the result of 4500 MHz, which is hardly attainable for the best Bloomfield and Lynnfield, was considered mediocre; many overclockers have successfully overclocked processors up to 5 GHz, and with an eye to everyday use! The combination of advanced architecture and outstanding frequency potential has allowed them to become the performance benchmark for all gaming-grade systems.

That is why the very first rumors about the imminent release of the latest 22nm processors became a real sensation. The most optimistic readers of our site, having heard about unknown transistors of a new design, low leakage currents and a small core area, expressed bold judgments like “well, it will take 5.5 GHz in the air, don’t go to the grandmother, or maybe all 6 GHz!”. This is not surprising - such a conclusion is easy to draw, taking into account the significant improvement in overclocking potential during the previous changes of the Intel CPU technical process.

In general, advances like "how it comes out - I'll take it right away" and "I've already specially bought a motherboard for Intel Z77" for the new processor was distributed a lot. How it all ended, I believe, is known to almost all readers. The 22nm Ivy Bridge, due to high operating temperatures and difficult overclocking, did not meet the expectations of many enthusiasts. So the "popular voice" instantly changed its tone - nowadays it is fashionable to scold Ivy. It comes to the point that some in all seriousness consider the new CPUs "overclockable" and incredibly hot, so much so that it is unrealistic to operate them at increased voltage without removing the heat distribution cover or, at least, using the CBO. But is it really so?

There is no doubt that overclockers who are closely monitoring the release of new hardware already know a lot about Ivy Bridge. Therefore, I propose not to go into the jungle of architecture (although there is certainly such a section in the article) and not spend time researching a huge number of accompanying parameters, but just check in practice - do you need new processor as part of a typical performance system, "sharpened" for overclocking.

Architecture and lineup

The new processors use the same architecture as previously released Sandy Bridge. As part of the proprietary "tick-tock" strategy (or "tick-tock" in the English version), which provides for the alternate update of technological processes and microarchitectures with the release of new products once a year, the Ivy Bridge release is a "Tick":

In the next season, fundamentally new processors using the same technical process should be presented - this will be "So".

In the meantime, we can conclude that Ivy Bridge should not differ from its predecessors in terms of the general layout and applied architectural solutions (Intel experts speak only of minor improvements that provide a performance advantage of 5%). The main innovation was the transfer of the core to a 22-nm process technology. Compared to the previously used 32nm, this was supposed to provide a significant reduction in core area, power consumption and heat dissipation.

So, the crystal of the new processor became immediately smaller by 35%. In comparison with the Sandy Bridge, which is very similar in design, its area has been reduced from 216 to 160 sq. mm. This is especially impressive given the fact that Intel specialists used a much more complex graphics core (the total number of transistors increased from 995 million to 1.4 billion, mainly due to the iGPU). If Ivy Bridge were just "22nm Sandy," the core footprint could be even smaller. But this is already a record of recent years - for comparison, you can cite a pair of CPUs made using a 32-nm process and containing a similar number of transistors. The eight-core AMD Bulldozer core is 325 square meters. mm at 1.2 billion transistors, the area of ​​the "stripped-down" quad-core Sandy Bridge-E is 294 sq. mm at 1.27 billion transistors.

Progress is clear. By the way, partly such a reduction in area became possible not only due to the new technical process, but also due to the use of original "three-dimensional" Tri-Gate transistors, instead of conventional planar ones.

The addition of an additional silicon "rib" allows for reduced leakage currents and a smaller overall structure. Also among the advantages of this model is the increased switching speed, although in practice many overclockers have already seen the opposite. However, overclocking problems can be caused by a dozen other reasons, it is likely that the three-dimensional structure will still reveal its potential on other processors from the company.

The TDP level announced for the new processors is 77 W. Although everything is not so simple here. The specifications presented to the sellers, as well as on the boxes, indicate 95 watts. Let me remind you that this value is typical for most quad-core Sandy Bridge, except for special "energy-saving" models. No matter how beautifully the representatives of the company explain this situation, it seems to me the most likely common "conspiracy" version, according to which the TDP had to be increased due to the strong heating of serial CPU samples. The situation when the novelty heats up more than its predecessor, despite the fact that according to the declared data, everything should be the other way around, would be utterly ridiculous.

However, this is what appears on the slide in the official press release:

So far, five models of the line have been presented, priced from $ 174 to $ 313. The maximum amount is asked for an unlocked Intel Core i7-3770K multiplier, which should replace the i7-2700K and i7-2600K, which are common among overclockers. Its analogue of the "budget" model i5-2500K, characterized by a free multiplier and the absence Hyper threading, this list also has - Intel Core i5-3570K. Let me remind you that at the time of release, they asked for $ 317 for the i7-2600K, and $ 216 for the i5-2500K, so the new items turned out to be even a little cheaper, although the difference is quite insignificant.

The cheapest 22nm CPU model is priced at $ 174, it is significantly cut in frequency and lacks Hyper Threading. The latest graphics core HD Graphics 4000 received all processors in the lineup, except for the two cheapest. All processors of the 37xx family are characterized by the maximum amount of L3 cache-memory (8 MB), and for the 35xx this figure has been reduced to 6 MB.

In general, everything is very similar to the Sandy Bridge line. By the way, like last time, Intel introduced several models with S and T indices, which differ in lower TDP. In general, the prices look quite reasonable, however, given weak competition from AMD in this segment, Intel there is no need to reduce them over time - so these processors can cost the same even before the release of the next generation 22nm CPU.

One of the significant advantages of Ivy Bridge is full (except for PCI-e 3.0 support) compatibility with previous generation motherboards based on Intel sixtieth series system logic.

Since the computing cores, in fact, have changed very little, Intel pays increased attention to the graphics subsystem:

The main pride of the company is the introduction of support for DirectX 11. Based on our own experience in testing video cards entry level I cannot help but note that this is pure profanation, it will be possible to use the advanced API in real games only at extremely low settings and far from FullHD resolution. In addition, support for OpenGL 3.1, OpenCL 1.1, Direct Compute and Shader Model 5.0 is announced. The most interesting thing is the possibility of simultaneous use of three monitors - the new CPU can be easily imagined as the basis for a working computer with three screens.

In terms of performance, the new graphics core boasts 16 universal execution units instead of 12 v previous version HD 3000.

Of course, this is all great, but I am still skeptical about the presence of "built-in" on the older models of the series and the need to necessarily acquire the increasingly complex video core along with the processor. Just think, iGPU eats up about a third of the transistor budget and core area, how much cheaper could a CPU be made without it? Although for the mobile segment, upgrading the graphics component of new CPUs can be extremely beneficial.

Overclockers are not forgotten either.

Of all the features of the new processors, the most interesting is the increased multiplier (from 59 to 63 units for "unlocked" models). This has already allowed several enthusiasts to conquer record frequencies using extreme cooling, not so long ago the 7 GHz mark was passed. It should also be noted the introduction of new increasing multipliers for RAM and improved overclocking capabilities of the video core.

A serial sample of the Intel Core i7-3770K processor was presented to the laboratory for testing. At the moment, this is the top model in the series, featuring an unlocked multiplier, Hyper Threading and 8 MB Cache L3. The maximum frequency with Turbo Boost is 3900 MHz, the base frequency is 3500 MHz.

In appearance, it practically does not differ from the “stones” of Sandy Bridge already familiar to all overclockers. Of course, the new CPU is easy to identify by the markings, you can also pay attention to the pinout of the capacitors on the reverse side.

Test stand

• Motherboards:
• ASUS P8Z77 DeLuxe (BIOS v 0603) for LGA 1155 processors;
• Sapphire Pure Black X79N (BIOS v 4.6.1) for LGA 2011 processor;
• Rivals (processors provided by Regard):
• Intel Core i7-2700K;
• Intel Core i7-3930K;
• CPU cooling system: Noctua NH-D14 (standard fans);
• RAM: Corsair TR3X6G1600C7 DDR3-1600, 7-7-7-20, 2 GB, dual-channel mode / four-channel mode;
• Video card: AMD Radeon HD 6970 (ref);
• Hard drive: Western Digital WD10EALX, 1000 GB;
• Power supply unit: Hiper K1000, 1 kW;
• Case: open stand.

Software

• Operating system: Windows 7 x64 Ultimate (no SP1);
• Video card drivers: AMD Catalyst 12.4 for Radeon HD 6970;
• Supporting utilities: SpeedFan 4.44, Real Temp 3.60, CPU-z 1.60, LinX 0.6.4, Prime 26.5 build 5 (In-Place Large FTTs).

Testing tools and methodology

Processors were overclocked without using third party utilities by directly changing the parameters in BIOS Setup. The Real Temp 3.60 utility was used to monitor the temperature of the cores, and the Linpack test in the Linx shell was used to create a load in the study of the temperature regime. The room temperature at the time of testing was ~ 26 degrees.

To test the performance of processors, we used following applications and synthetic tests:

• SuperPi Mod 1.5 (XS) - the time required to calculate 1 million digits of Pi after the decimal point (Super Pi 1M) was taken into account. Single-threaded test.
• Fritz Chess Benchmark - the number of operations per second (kilo Nods). All processors performed the test in eight threads.
• WPrime Benchmark v. 2.09 - the time required to complete the test in 32M mode was taken into account. The algorithm was executed in four / six threads according to the recommendations of the test developers, although modern Intel processors can get an advantage using Hyper Threading, but for this comparison, the absolute result is insignificant.
• 3DMark Vantage 1.0.1 - Performance preset, CPU Score was taken into account.
• SiSoft Sandra Professional 2010 - Based on the following benchmarks: CPU arithmetic performance (overall performance), overall cryptography speed.
• True Crypt 7.1a - built-in benchmark, AES-Twofish-Serpent encoding rate was taken into account. Quad-core processors executed the algorithm in eight threads, six-core in twelve.
• Cinebench 11.5 x64 - Scene rendering, taking into account the overall CPU rating in points. Quad-core processors executed the algorithm in eight threads, six-core in twelve.
• PovRay 3.7 - built-in benchmark, All CPU's mode, the time required to render the scene was taken into account.
• WinRar 4.20 beta 2 (x64) - built-in performance test. The multithreading mode was activated in the program settings.
• x264 HD Benchmark v4.0 is a standard video conversion algorithm. The graphs show the minimum and maximum FPS values ​​obtained in two test passes. Quad-core processors executed the algorithm in eight threads, six-core - in twelve threads.
• Adobe Photoshop CS5 - the time taken to apply a sequence of filters to a reference image was measured.

In addition, we tested the system performance in several games.

• Hard Reset - built-in test;
• F1 2011 - built-in test;
• Batman: Arkham City - built-in test;
• Сrysis 2 - Adrenaline Crysis 2 Benchmark Tool;
• Metro 2033 is a proprietary performance testing utility that comes with the game.

VSync was disabled during all tests. The list of game settings will be provided in each case separately for the sake of convenience.

Product release date.

Lithography

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

Number of Cores

Core count is a hardware term that describes the number of independent central processing units in a single computing component (die).

Number of threads

Thread or thread of execution is the term software, denoting a basic ordered sequence of instructions that can be transmitted or processed by a single CPU core.

CPU base clock speed

The base frequency of the processor is the rate at which the processor transistors open / close. The processor base frequency is the operating point where the TDP is set. Frequency is measured in gigahertz (GHz) or billions of computational cycles per second.

Maximum clock speed with Turbo Boost technology

Turbo Maximum Clock Speed ​​is the maximum clock speed of a single core processor that can be achieved using Intel® Turbo Boost and Intel® Thermal Velocity Boost Technologies. Frequency is measured in gigahertz (GHz) or billions of computational cycles per second.

Cache memory

Processor cache is an area of ​​high-speed memory located within the processor. Intel® Smart Cache refers to the architecture that allows all cores to dynamically share last-level cache access.

System bus frequency

A bus is a subsystem that transfers data between computer components or between computers. An example is the system bus (FSB), through which data is exchanged between the processor and the memory controller unit; DMI, which is a point-to-point connection between the Intel Integrated Memory Controller and the Intel I / O Controller Hub on motherboard; and a Quick Path Interconnect (QPI) interface between the processor and the integrated memory controller.

Design power

Thermal Design Power (TDP) refers to the average performance in watts when the processor is dissipating power (at base clock with all cores active) under a complex load as defined by Intel. Read the requirements for thermoregulation systems presented in the datasheet.

Embedded Options Available

Embedded Options Available indicates products that provide an extended purchase option for smart systems and embedded solutions. Product specifications and conditions of use are presented in the Production Release Qualification (PRQ) report. Please contact your Intel representative for details.

Max. memory size (depends on the type of memory)

Max. memory size refers to the maximum amount of memory supported by the processor.

Memory types

Intel® processors support four different types memory: single-channel, dual-channel, three-channel and Flex.

Max. number of memory channels

Application bandwidth depends on the number of memory channels.

Max. memory bandwidth

Max. memory bandwidth refers to the maximum rate at which data can be read from memory or stored in memory by the processor (in GB / s).

ECC memory support ‡

ECC memory support indicates the processor is supporting ECC memory. ECC memory is a type of memory that supports the identification and repair of common types of internal memory corruption. Note that ECC memory support requires both processor and chipset support.

Processor Graphics ‡

The graphics system of the processor is a graphics processing circuit integrated into the processor, which forms the operation of the functions of the video system, computing processes, multimedia and information display. Intel® HD Graphics, Iris ™ Graphics, Iris Plus Graphics, and Iris Pro Graphics provide advanced media conversion, high frame rates, and 4K Ultra HD (UHD) video display capability. See the Intel® Graphics Technology page for more information.

Graphics Base Frequency

The graphics base clock is the rated / guaranteed graphics rendering clock (MHz).

Max. dynamic frequency of the graphics system

Max. Graphics Dynamic Frequency is the maximum conditional rendering frequency (MHz) supported by Intel® HD Graphics with Dynamic Frequency.

Intel® Quick Sync Video

Intel® Quick Sync Video Technology provides fast video conversion for portable media players, web hosting, and video editing and creation.

InTru ™ 3D technology

Intel® InTRU ™ 3D technology plays back 3D stereoscopic Blu-ray * video at 1080p using HDMI interface* 1.4 and high quality sound.

Intel® Flexible Display Interface (Intel® FDI)

Intel® Flexible Display is an innovative interface that allows you to display independent images on two channels using the integrated graphics.

Intel® Clear Video HD Technology

Intel® Clear Video HD Technology, like its predecessor Intel® Clear Video Technology, is a collection of video encoding and processing technologies built into the processor's integrated graphics. These technologies make video playback more stable and graphics clearer, more vibrant, and more lifelike. Intel® Clear Video HD Technology delivers brighter colors and more lifelike skin through video enhancements.

PCI Express Revision

PCI Express Revision is the version supported by the processor. PCIe (Peripheral Component Interconnect Express) is a high-speed serial expansion bus standard for computers to connect to hardware devices. Different versions PCI Express supports various data transfer rates.

PCI Express Configurations ‡

PCI Express (PCIe) Configurations describe the available PCIe lane configurations that can be used to map PCH PCIe lanes to PCIe devices.

Supported connectors

A connector is a component that provides mechanical and electrical connections between the processor and the motherboard.

Cooling system specifications

Intel thermal reference specifications for the proper operation of this heading.

T CASE

The critical temperature is the maximum temperature allowed in the integrated heat spreader (IHS) of the processor.

Intel® Turbo Boost Technology ‡

Intel® Turbo Boost Technology dynamically increases the processor frequency to the required level by taking advantage of the difference between the nominal and maximum values ​​of the temperature and power consumption, which can be used to increase energy efficiency or "overclock" the processor when needed.

Intel® vPro ™ Platform Compliant ‡

Intel® vPro ™ Technology is a processor-based management and security suite that addresses four key areas information security: 1) Threat management, including protection against rootkits, viruses and other malware 2) Identity protection and point-to-point protection of website access 3) Protection of confidential personal and business information 4) Remote and local monitoring, patching, PC repair and workstations.

Intel® Hyper-Threading Technology ‡

Intel® Hyper-Threading Technology (Intel® HT Technology) provides two processing threads for each physical core. Multi-threaded applications can perform more tasks in parallel, which greatly speeds up work.

Intel® Virtualization Technology (VT-x) ‡

Intel® Virtualization Technology for Directed I / O (VT-x) allows a single hardware platform to function as multiple “virtual” platforms. The technology improves management capabilities by reducing downtime and maintaining productivity by allocating separate partitions for compute operations.

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

Intel® technology Virtualization Technology for directional I / O, supplements virtualization support on IA-32 (VT-x) and Itanium® (VT-i) processors with I / O virtualization. Intel® Virtualization Technology for Directed I / O helps users increase the security and reliability of systems and improve I / O performance in virtualized environments.

Intel® VT-x with Extended Page Tables (EPT) ‡

Intel® VT-x with Extended Page Tables, also known as Second Level Address Translation (SLAT) technology, accelerates memory-intensive virtualized applications. Extended Page Tables on Intel® Virtualization Technology-enabled platforms reduce memory and power overhead and increase battery life by hardware optimized page table management.

Intel® TSX-NI

Intel® Transactional Synchronization Extensions New Instructions (Intel® TSX-NI) are a set of instructions designed to scale performance in multi-threaded environments. This technology helps to more efficiently perform concurrent operations through improved control over software locking.

Intel® 64 architecture ‡

Intel® 64 architecture, when combined with the associated software, supports 64-bit applications on servers, workstations, desktops, and laptops ¹ Intel® 64 architecture provides performance enhancements that allow computing systems to consume more than 4 GB of virtual and physical memory ...

Command set

An instruction set contains basic commands and instructions that the microprocessor understands and can execute. The value shown indicates which Intel instruction set the processor is compatible with.

Instruction set extensions

Instruction set extensions are additional instructions that you can use to improve performance when performing operations on multiple data objects. These include SSE (Support for SIMD Extensions) and AVX (Vector Extensions).

Idle states

Idle state (or C-state) mode is used to conserve power when the processor is idle. C0 means an operational state, that is, the CPU is currently doing useful work. C1 is the first dormant state, C2 is the second dormant state, and so on. The higher the numerical indicator of the C-state, the more energy-saving actions the program performs.

Enhanced Intel SpeedStep® Technology

Enhanced Intel SpeedStep® Technology helps ensure high performance while meeting the power-saving requirements of mobile systems. Standard Intel SpeedStep® Technology allows voltage and frequency switching based on processor load. Enhanced Intel SpeedStep® Technology is built on the same architecture and uses design strategies such as decoupling voltage and frequency changes, and clock distribution and recovery.

Thermal control technologies

Thermal management technologies protect the processor case and system from overheating failure with multiple thermal management features. The Digital Thermal Sensor (DTS) detects core temperature and thermal management functions reduce the power consumption of the processor chassis as needed, thereby lowering temperatures to ensure operation within normal operating specifications.

Intel® Privacy Shield Technology ‡

Intel® Privacy Shield Technology is built-in token-based security technology. This technology provides simple, reliable controls for online access to business and business data, protecting against security threats and fraud. Intel® Privacy Shield Technology uses hardware-based PC authentication mechanisms for websites, banking systems, and network services confirming the uniqueness of this PC, protects against unauthorized access and prevents attacks using malware. Intel® Privacy Shield Technology can be used as a key component of two-factor authentication solutions designed to protect information on websites and control access to business applications.

Intel® Trusted Execution Technology ‡

Intel® Trusted Execution Technology enhances secure command execution capabilities by hardware expansion of Intel® processors and chipsets. This technology provides security features such as measurable application launch and secure command execution for digital office platforms. It does this by creating an environment where applications run in isolation from other applications in the system.

Function Cancel Execute Bit ‡

The cancel execute bit is a hardware security feature that can reduce vulnerability to viruses and malicious code and prevent malware from running and spreading on a server or network.

Anti-Theft technology

Intel® Anti-Theft Technology helps keep your laptop secure if it is lost or stolen. To use Intel® Anti-Theft Technology, you must subscribe to an Intel® Anti-Theft Technology Service Provider.

Each of the processor manufacturers has “top-end” models, in which all the available developments are implemented, capable of fully unleashing the full potential of the development team. So our laboratory got a "top" model of a new generation CPU, which includes almost all the "chips" of Intel. First of all, it is worth recalling that the novelty is made in accordance with the 22 nm technical process. The announcement of this model took place on the day when the public presentation of the Ivy Bridge architecture took place. Like all flagship solutions related to architectures of previous years, this CPU is positioned as a solution for the most demanding users. This category includes connoisseurs of computing power to perform complex engineering tasks and, of course, run modern computer games. In addition, this CPU is able to interest overlockers, because the model has an unlocked multiplier. Now let's move on to the direct study of the processor.

Appearance and packaging

We are testing a full-fledged retail copy, "boxed version", of an Intel Core i7-3770K processor. The external design does not fundamentally differ from the packaging of younger models. The text part informs about the presence of an unlocked multiplier, and also high level productivity.

The package has a plastic window through which you can see the markings on the processor.

On the side of the box, the features of the model are shown. The manufacturer highlights a number of features inherent in almost all models of the Intel Core i7 family with an unlocked multiplier:

naturally unlocked multiplier;

the presence of four cores and processing of eight data streams;

the presence of a built-in dual-channel memory controller that supports DDR3 memory;

support for the most significant technologies from Intel (Intel Turbo Boost Technology 2.0, Intel Hyper-Threading Technology, Intel Smart Cache);

the presence of an integrated graphics core, in this case it is Intel HD Graphics 4000.

The traditional white label contains key data on the characteristics of the CPU: CPU clock speed (3.50 GHz); the amount of cache memory (8 MB); processor socket (LGA 1155); TDP (77W), serial number and product code. You can see that the main characteristics this model The CPU is very similar to the Intel Core i7-2700K. The only difference is the thermal package, which was 95W in the previous generation model.

The processor package is quite standard. The package contains the cooling system, a sticker on the system unit and instructions that will help the owner in installing the CPU.

We've already seen the E97378-001 complete cooling system from a number of high-performance CPU reviews. In general, its design is absolutely standard and consists of a fan and a radiator. The performance of this cooler is sufficient to ensure a normal temperature regime at the nominal parameters of the CPU, but remember that in order to overclock it is necessary to purchase a more efficient cooling system. As for the manufacturer, in this case it is DELTA. With regard to the background noise, we can say that it is quite low, therefore, when placing system unit on the desktop, you will not feel any discomfort.

The processor cover shows the processor model (Intel Core i7-3770K), clock speed (3.5 GHz) and place of manufacture (Costa Rica).

On the back side there are contacts for the Socket LGA1155. This connector is native to all mainstream CPU models of the second and third generations, while motherboards with new sets of system logic support the operation of the previous generation CPUs.

Specification

According to the specification data, we see that apart from the change in the technical process, which in this case is 22 nm, there are not so many visible changes. Majority technical characteristics practically did not undergo any changes in comparison with It is also worth mentioning that in addition to general changes in the transistor architecture, the new generation of CPUs has received a full-fledged PCI Express 3.0 bus controller for general use platforms. Special attention should be paid to the memory controller, which in this processor is capable of working with overclocking modules at frequencies up to DDR3-2133 and more, and if we take into account that the maximum value of the processor multiplier is increased to x63, then we can hope for a good overclocking potential.

The helper utility confirms that the Intel Core i7-3770K is made according to the 22nm process technology. With nominal parameters and close to the maximum load level, the voltage on the core was 1.080 V, while the clock frequency at the time of reading was 3511 MHz.

As you remember, all modern high-performance processor models belonging to Intel series Core i7, equipped with Intel Turbo Boost Technology 2.0. Thus, if necessary, the CPU frequency can be increased to 3.9 GHz, but this is possible for a while while there is a need, or the power consumption of the core is within the declared thermal package.

The cache memory allocation is identical to the previous generation of the tested model. The cache memory of the first level is 64 KB per core, of which 32 KB is used for caching data and the same amount for instructions (8 association lines are used). Additionally, in the second block, space is allocated for the decoded micro-op cache. The L2 cache is 256 KB per core (8 association lines are also used). The L3 cache is shared by the entire processor and is 8 MB in size (16 association lines).

The dual channel memory controller is capable of supporting both DDR3-1333 and DDR3-1600 in nominal mode. At the same time, fans of overclocking can easily see it work with DDR3 modules at frequencies up to 2133 MHz and even more. But, as you understand, in order to fully reveal the overclocking potential of the CPU, you will have to purchase a motherboard with the appropriate set of system logic, which allows you to perform such operations with a high-quality CPU power system.

During testing, we used Processor Test Bench # 1

The test results for the Intel Core i7-3770K processor are generally impressive. The new generation brought an additional 5-10% performance, although you will have to pay a little more for them, but in games the situation has not undergone any fundamental changes. Therefore, at the moment, for gaming systems, the "top-end" processor is not so relevant, because with lower financial costs for a little less fast model it will be possible to invest the difference in a more efficient video accelerator, which will ultimately make the system more suitable for modern games.

As for Intel Turbo Boost 2.0 technology, once again we see the familiar 3-5% performance gain. Therefore, disabling it will not be the most reasonable step, because it will not require additional costs, everything has already been paid for.

The most interesting situation arises when comparing performance with Intel Core i7-3820, a "budget" representative of the Sandy Bridge-E architecture. If earlier there were doubts about the advisability of acquiring it for systems with a wide range of tasks to be solved, now these doubts are even stronger, because the only element that plays in its favor is a four-channel memory controller. It is thanks to him that the advantage is observed in tests that assess the efficiency of RAM. Otherwise, the novelty wins this "race", especially if we consider the model without the unlocked multiplier, which is almost identical in cost to Intel Core i7-3820.

As for the "top" CPU model from AMD, which we tested, the only advantage is this decision is the cost, which is more than $ 100 lower than that of Intel Core i7-3770K. An additional positive point is the support for an extended set of instructions, which with the availability of the appropriate software that would make full use of them, is able to reveal its capabilities to the maximum. But, unfortunately, such software products extremely few.

Integrated graphics performance

As you already know, the Intel Core i7-3770K model is equipped with the new Intel HD Graphics 4000 graphics core. Equipping it with 16 compute units instead of 12 for Intel HD Graphics 3000 has become extremely important from the point of view of improving performance.

The implementation of a kind of Turbo Boost mode has become a traditional element of the video core. Ultimately, this led to the fact that in idle mode the GPU frequency drops to 650 MHz, while the power consumption is only 0.1 W. At maximum load, the frequency reaches 1150 MHz (which is slightly different from the materials of the presentation, in which the maximum frequency was declared within 1350 MHz). Power consumption in this mode is 4.5 W, which is comparable to low-power GPUs installed in mobile devices.

A series of tests were performed to determine the performance of the Intel HD Graphics 4000 graphics core.

According to the test results, it is noticeable that such a reworking of the graphics core did not pass without leaving a trace. The novelty quite successfully competes with top AMD Previous generation APU. With some lag, Intel HD Graphics 4000 still demonstrates good performance even in gaming applications with DirectX 11 instructions, but you should count on a comfortable level of play in modern games only at low screen resolutions, and even then not always. Otherwise, the novelty copes well with the traditional tasks faced by a multimedia PC, especially when you consider that with the appropriate motherboard up to 3 monitors can be connected. In general, the new graphics core will be an excellent assistant when assembling high-performance computing systems or servers of large corporate networks. For other tasks, it is worth completing the system with a modern discrete video card with performance that meets your requirements.

Overclocking

Thanks to the unlocked multiplier, we managed to achieve stable operation of the system at a frequency of 4.7 GHz with a multiplier of x47 without any problems. The core voltage has been increased to 1.296V to ensure system stability. Recall that the ceiling in x63 declared by the manufacturer is potentially achievable only under extreme overclocking conditions with the use of liquid nitrogen. We use traditional air cooling, which explains such a modest result. It is also worth remembering that overclocking is a kind of lottery, so the results may vary from case to case.

The table below shows the performance indicators in the nominal mode and after overclocking the processor.

As a result of the increase in clock speed to 4.7 GHz, the average performance gain was almost 20%. The result, of course, is slightly worse when compared with the Intel Core i7-2700K, although not bad. The most striking performance increase (about 30-40%) occurred in the computing component of the processor, which is quite expected. Also, a good response has occurred with the memory, because inevitably during overclocking, the frequency of the modules is also affected. Thus, if you are in the process work activities often having to solve resource-intensive tasks related to computing, image rendering, or archiving, overclocking can make your life easier.

Analysis of the effectiveness of Hyper-Threading technology.

A traditional element of reviews of "top" solutions is the analysis of the effectiveness of the technologies involved. In this case, too, we will consider the efficiency of Hyper-Threading.

To analyze the effectiveness of this technology, a series of tests was carried out on the same system with Hyper-Threading enabled and disabled, which is selected in the corresponding BIOS menu.

Well, you can see that the average performance gain as a result of using Hyper-Threading technology was only 3.84%. In general, the result is not very impressive, but for computational tasks and operations related to archiving, this figure exceeds 20%. Thus, it is for these purposes that the use of such a CPU will be relevant. In other tasks, most likely, you can do with cheaper models from the Intel Core i5 line, in which work is performed without emulating additional cores.

Performance differences resulting from the use of DDR3-1333 and DDR3-1600 RAM

Quite often, a potential buyer is faced with the question of what RAM choose for your system, because many CPU models are capable of supporting several types of RAM. So we ran a series of tests to determine the optimal memory type for systems equipped with an Intel Core i7-3770K CPU.

The difference in performance averaged about 3%. Naturally, the most sensitive tests were those associated with frequent data exchange with memory, for example, the archiving operation, for which the increase was 15%. Otherwise, the differences are insignificant. Thus, when assembling a system with the maximum performance level, you should pay attention to DDR3-1600 or faster memory modules with not much increased latency.

Energy consumption

As you can see, in terms of power consumption among high-performance solutions, the competition is exclusively between Intel solutions. The new generation of CPUs provided 20-30 W more efficient power consumption in comparison with similarly positioned models of the outgoing generation. The gap with the "top" CPUs of various AMD architectures is only widening. Thus, the tested CPU, given its level of performance, can be used in systems designed for long computations. The only "serious" drawback of this solution seems to be only the cost, therefore, before setting the goal of purchasing this model, it is worth assessing its necessity.

conclusions

As a result of testing the processor, we saw a record at the moment performance indicators among systems general purpose... Of course, there is a specific platform with Socket LGA2011 with "exclusive" CPU models, but they have a different price order and the target orientation is somewhat different. A system based on this CPU will be an indispensable assistant in solving serious computing tasks, but you should immediately assess the need for overclocking, because if you do not plan to perform it, then an Intel Core i7-3770 with a locked multiplier would be a more reasonable choice. As a result of the research, it was found that the most harmonious memory modules used with this processor are DDR3-1600 or faster, which provide an additional average increase in power, while their choice will not have a significant impact on the final cost of the system.

The overclocking potential of the tested CPU model is slightly lower than that of the similarly positioned model of the previous generation, although a 20% increase in power will be clearly noticed by the user. The most pronounced advantage of the overclocked version of the CPU is manifested precisely when solving computational problems, converting video, and processing images.

Special attention should be paid to the significantly redesigned graphics core Intel HD Graphics 4000 thanks to the use of as many as 16 computing units, as well as some optimization of their work (as the manufacturer claims). The new graphics core can quite confidently compete with the previous generation APU video cores from AMD. Of course, you shouldn't expect it to fully replace a good discrete video card, but it is quite enough to solve standard tasks that multimedia systems face. In this case, you can run not too demanding modern games, although a more or less comfortable gameplay is possible at low monitor resolutions and / or low image quality settings.

We express our gratitude to the company LLC PF Service (Dnepropetrovsk) for the processor provided for testing.

We express our gratitude to the companies ASUS , Kingston and Sea sonic for the equipment provided for the test stand.

Article read 91019 times

Subscribe to our channels