Intel hd 4000 year of manufacture. Graphics: fast, slow and integrated. Test bench configuration

The emergence of the Windows 8 operating system has become a kind of “engine of progress” for a huge number of manufacturers computer equipment. The new OS, which has two types of control (touch and classic), gave an additional impetus to the creation of devices of a new form factor that combine a tablet and a laptop. We have already introduced you to one of the representatives of this class, namely a laptop. IN this material We will look in as much detail as possible at the next new product in the “transformers” line, which you have probably already heard about.

Review of the Impression X70.02 ultrabook

August of this year was marked by the fact that the company “ Navigator"presented to the public its first ultrabook, which will be produced under its own brand Impression Computer, and this, one might say, is a rather significant event for the domestic IT market. After all, it is known that the production of new items, like all other devices of this trademark, is carried out on the territory of our country.

Model ImpressionX70 is positioned by the manufacturer as a solution for the corporate segment of users, which is emphasized by an extended warranty and support of up to 24 or 36 months Intel technologies Anti-Theft with McAfee Anti-Theft software to remotely lock a stolen device and protect information stored on the drive. At the same time, almost the main feature of the ultrabook, in addition to the compact dimensions inherent this class solutions is to use a high-capacity battery - 7800 mAh.

GIGABYTE BRIX GB-XM12-3227 mini computer review

Thanks to the active development of the computer sector and constant transitions to more technologically advanced and much more energy-efficient processes for creating components, among which the largest and the most important role performed by processors, equipment manufacturers are given the opportunity to implement rather unusual devices into reality, characterized by the most compact dimensions while maintaining maximum quantity opportunities. This was precisely the decisive factor in the emergence of such a class of desktop solutions as mini-computers, which are now actively promoted not only by manufacturing companies, for example, ZOTAC with its ZBOX nano XS model, but also by Intel itself in the form of a conceptual device NUC (Next Unit of Computer), equipped with “full-fledged” processors of the line Intel Core.

Not long ago, the Taiwanese GIGABYTE joined these companies, which brought to the market a series of very compact mini-computers under the laconic name GIGABYTE BRIX, and is now actively expanding the model range of this line. At the moment, the “bricks” are available both in the basic version and in a very unique version with a built-in mini-projector with a brightness of 75 lumens, capable of displaying an image measuring from 7 to 85 inches diagonally at a resolution of 864 by 480 pixels. The gaming GIGABYTE BRIX II, which is said to be capable of playing games at the level of Crysis 3, should also go on sale soon. It is important to note that the manufacturer gives preference not only to solutions from Intel, but also to accelerated processors from AMD.

In this material we will dwell in more detail on one of the models of the starting line, namely GIGABYTE BRIX (GB-XM12-3227). Its main feature, undoubtedly, is the incredibly compact body, in which the manufacturer managed to fit an energy-efficient dual-core Intel Core i3-3227U processor with integrated Intel graphics HD Graphics 4000. In this case, selection and installation RAM and the drive is left to the discretion of the consumer, which expands the configuration options. However, not everything is so happy in the mini-computer and already at the first acquaintance a number of complaints are revealed.

Ultrabook review and testing Lenovo ThinkPad T431s

A bright representative of Lenovo T-series ultrabooks, in model range which includes only premium devices. And this means that this model, according to the company, is the embodiment of functionality, the highest quality workmanship and stylish design.

At first glance, it is clear that the Lenovo ThinkPad T431s was developed not just as another “laptop” squeezed into an ultrabook form factor, but as a device with its own, unique look, as evidenced by its appearance and software and hardware capabilities. A reinforced carbon body, a spill-resistant keyboard, and advanced security features - that's not all. full list distinctive features this ultrabook. Lenovo ThinkPad T431s is produced in various configurations, differing primarily in processor models, as well as the volume and type of drives. We received a sample for testing based on the Intel Core i5-3337U.

Intel Core i3/Core i5 (Haswell) processors for embedded systems coming in Q4 2013

Review and testing of the Dell XPS 12 ultrabook

Thanks to the release of the latest operating system from Microsoft, namely Windows 8, which is quite strongly focused on touch control, almost each of the manufacturers presented their vision of new devices that would simultaneously offer an equally convenient way to use both in classic mode and in tablet mode. Some of them began to develop completely new form factors for devices. For example, Lenovo introduced an ultrabook Lenovo Yoga with an innovative display unit hinge design that opens 360°, thereby transforming the laptop into a tablet. Other companies decided to go the proven route and use the concept of a laptop with a detachable display, which was originally developed by ASUS and initially used for its Android tablets.

Dell, keeping up with its competitors, decided to use its early developments, especially since one of these developments had already been used to produce the first of its kind flip-flopping laptop, the Dell Inspiron Duo, with a 10" display rotating around its axis. An original and very reliable design aroused quite a lot of interest in the device, but it did not become particularly popular due to its small diagonal and not very convenient for touch Windows mode 7.

The second changeover was the ultrabook, which should attract much more public attention, because the new product is not only made in the same unique premium style as the Dell XPS 13, but is also equipped with an excellent Full HD display with a diagonal of 12.5", perfect for touchscreen Windows interface 8. However, no matter how bitter it may sound, there was a fly in the ointment. We'll find out which one later.

Fujitsu LIFEBOOK E743 - a reliable and productive business class laptop

It is noted that this generation GPUs Intel will support a number of new APIs (DirectX 11.1, OpenCL 1.2, OpenGL 3.2), will provide improved content experience, allow for multi-screen configurations and will guarantee support for the DisplayPort 1.2 interface.

As for the performance level of the Intel HD Graphics 4600 GPU, Intel claims that it is in the class server solutions This GPU can replace discrete video cards costing up to $150. The basis for such conclusions was comparative testing processor Intel Xeon E3-1275 v3 (Intel HD Graphics 4600) with its predecessor Intel Xeon E3-1275 v2 (Intel HD Graphics 4000) and two entry-level discrete graphics cards in the SPECaps PTC Creo 2.0 benchmark. Increasing the number of computing units in the Intel HD Graphics 4600 model and optimizing its driver allowed the new product to demonstrate better results than unnamed budget discrete video cards in three out of five test sets. And the lag graphics core the previous generation compared to the new product based on testing results averaged 26%.

Ultrabook Samsung Series 9 Premium Ultrabook is cheaper

Good news for everyone who was planning to purchase an ultrabook SamsungSeries 9 Premium Ultrabook, but was stopped by its original recommended price of $1900, which announced late last month. Today, some online stores are accepting pre-orders for the new product at prices starting from $1,350 per model with solid state drive 128 GB capacity

Despite the considerable cost SamsungSeries 9 Premium Ultrabook looks like a very attractive purchase. The ultrabook is equipped with a 13.3-inch display with a resolution of 1920 x 1080 pixels, protective glass Gorilla Glass and SuperBright backlighting, Intel Core i7-3517U processor, 4 GB of RAM, card reader, SoundAlive HD Audio stereo speakers, module wireless communication Wi-Fi and a wide range of connection interfaces. Stated time battery life- about 8 hours.

The ultrabook body is made of aluminum, and its total weight is 1150 g.

Specifications:

Ultrabook ASUS ZENBOOK U500VZ-CN097H with 15.6-inch touch display

For everyone who wants to purchase a high-performance and elegant ultrabook, ASUS has developed and introduced the ASUS ZENBOOK U500VZ-CN097H model. This 15.6-inch new product is equipped with a quad-core standard mobile Intel Core i7-3632QM processor, six gigabytes of DDR3-1600 RAM and hybrid disk subsystem. The latter consists of a 128 GB SATA SSD drive and a 500 GB HDD drive.

ASUS specialists also worried about the high quality of multimedia content playback by equipping the mobile ASUS computer ZENBOOK U500VZ-CN097H with Full HD IPS touch display, NVIDIA GeForce GT 650M mobile video card and 2.1-channel Bang & Olufsen IcePower audio subsystem with support for Sonic Master technology. And for video communication, the new product features an HD (720p) webcam with an integrated microphone.

The new product went on sale with an 8-cell battery and an installed operating system Windows 8. Its estimated price is €1699. Summary technical specification ASUS ultrabook ZENBOOK U500VZ-CN097H is presented in the following table:

In the previous article we told you about the new processors from the line Ivy Bridge, today we will touch on one of the components of these processors - the built-in Intel graphics HD 4000, codenamed Carlow.

The graphics, like its previous version, Intel HD 3000, has four processor cores, but the new version also has support for DirectX 11. However, it’s too early to rejoice. DirectX 11 can only be found in the latest games, which are so demanding on system resources that our built-in video card will probably be left behind system requirements. And this is even despite the fact that compared to the graphics in Sandy Bridge, our 4000 has tripled its performance (at least, that’s what Intel claims). And in general, there are so many changes in the graphics core that this is a clear big step forward compared to the previous options.

It is now possible to connect as many as three monitors to the graphics at the same time (although this may require DisplayPort). If you need to open many windows for work, and they all need to be in front of your eyes, then this function will certainly be useful to you. In addition, a powerful processor will make it possible to run demanding graphics programs if you are a designer. In general, a rather bright prospect emerges here in terms of using a laptop or ultrabook on Ivy Bridge. When you need mobility, you take it and go where you need to go. When you need to work at a stationary place, you connect a large monitor (or even several) to your mobile computer and work.

The base clock speed of this graphics can be increased because Turbo Boost technology is built into the processor chip. Depending on the processor model base frequency and overclocking frequency may vary. For example, its performance on low-power processors will be 30% below average. In general, it can operate at clock frequencies from 350 to 1350 MHz.

The clock frequency here is lower than in previous versions, which makes it possible to reduce power consumption. Since the microarchitecture of the graphics core was changed for the better, Intel felt that this would not reduce its performance, which was already quite sufficient.

Intel HD 4000 graphics includes 16 execution units, or unified shaders, while Intel HD 3000 could boast only 12. In addition, there is support for OpenGL 3.1 and OpenCL 1.1 (the latter using shader processors). The totality of the characteristics of the new graphics is such that it is almost equal to a very productive development from AMD - Llano. In terms of performance, HD 4000 is on par with discrete Nvidia GeForce GT 330M and exceeds the performance of the integrated Radeon HD 6620G (though only when paired with a quad-core processor).

The encoding quality has also improved, and the video encoding speed has doubled. By the way, the hardware video encoder can play back at least 16 video streams, all in high definition. It can also play ultra-high resolution content of 4096x2304.

However, even though we wrote that in latest games It is unlikely that you will be able to play on this graphics, but some will still run on it - unless, of course, they are too demanding of graphic resources. The gaming performance of the Intel HD 4000 is 50% higher than that of the 3000. Among the games you can play on it are Left 4 Dead 2, DiRT 3, Street Fighter 4 and others. If you have run games on the Intel HD 4000, write in the comments what works on it and what doesn’t. We will make an update later.

Here is a short table for now (click on the picture to enlarge):

Also playable:
FIFA 11 (2010)
Battlefield: Bad Company 2 (2010)
F.E.A.R. 2 (2009)
Counter-Strike Source (2004)

Just a few years ago, talking about the performance of integrated graphics cores made virtually no sense. It was possible to rely on such solutions only in cases where working with three-dimensional graphics was not among the possible uses of the computer, because the built-in graphics cores, compared to discrete video accelerators, had minimalistic functionality in 3D modes. However, today this situation has changed radically. Since 2007, the instigator of the bulk of changes in the computer market, Intel considers increasing the capabilities and performance of its own integrated graphics one of the most important tasks. And its successes are impressive: the built-in graphics cores have not only increased their performance by more than an order of magnitude, but have also become an integral integral part modern processors. Moreover, the company clearly does not intend to stop there and has ambitious plans to increase the speed of embedded graphics by another order of magnitude by 2015.

The sudden interest among processor developers in improving graphics cores became a reflection of the desire of users to have at their disposal fairly compact, but at the same time quite productive computing systems. It would seem that quite recently the term “ mobile computer“was associated with a system that can simply be moved from place to place with one hand, and few people were concerned about the issue of its size and weight. Today, even looking at fairly small two-kilogram laptops, many consumers wrinkle their noses with dissatisfaction. The trend has turned tablet computers and ultra-compact solutions that Intel calls ultrabooks. And it was precisely this desire for lightness and miniature that became the main driving force in the integration of graphics into central processing units and in increasing its productivity. One chip that fully replaces both the CPU and the GPU and has low heat dissipation is exactly the basis that is needed to create mobile solutions that entice modern users. That is why we are seeing the rapid development of hybrid processors, the existence of which even adherents have to put up with desktop systems. The latter, it must be said, also receive certain dividends from such progress.

Ivy Bridge processors are the second version of Intel's microarchitecture, characterized by a hybrid design that combines computing cores with graphics in one semiconductor chip. Compared to previous version microarchitecture, Sandy Bridge, dramatic changes have occurred, and they primarily affect the graphics core. Intel even had to give special explanations regarding the violation of the “tick-tock” principle: Ivy Bridge was supposed to be the result of a transfer of the previous design to a new, 22-nm process technology, but, in fact, in terms of graphics capabilities there was a very significant step forward. That is why we reviewed the new video core included in Ivy Bridge in the form of a separate material - the number of various innovations is extremely large, and the improvement in 3D performance is quite serious.

A great idea of ​​how significant the changes have been can be obtained by simply comparing Ivy Bridge and Sandy Bridge semiconductor crystals.

Sandy Bridge - area 216 sq.mm; Ivy Bridge - area 160 sq.mm

Both of them are made according to different technological processes and have different areas. But note that while the Sandy Bridge design allocated approximately 19 percent of the die area to the graphics core, the Ivy Bridge design increased that share to 28 percent. This means that the complexity of the graphics included in the processor has more than doubled: from 189 to 392 million transistors. It is quite obvious that such a noticeable increase in the transistor budget could not be wasted.

It must be emphasized that Intel's policy regarding combining computing and graphics cores and increasing the power of the latter is somewhat at odds with the APU concept proposed by AMD. Intel's competitor is considering the on-chip graphics core as a complement to the computing core, hoping that flexible programmable shader processors can help increase the overall performance of the solution. Intel does not take into account the possibility of widespread use of graphics for computing: with traditional processor speed Ivu Bridge is fine as is. At the same time, the primary role of the graphics core is completely traditional, and the struggle of developers to increase its power is due to the desire to minimize the number of cases when a discrete video card acts as a necessary system component, especially in mobile computers.

However, whether AMD’s approach or Intel’s, the result turns out to be the same. The market share of discrete graphics is steadily declining, giving way to new generations of integrated graphics, which have now acquired support for DirectX 11 and have received performance higher than that of a number of budget video cards. In this material, we will look at the Intel HD Graphics 4000 and Intel HD Graphics 2500 graphics accelerators implemented in Ivy Bridge and try to evaluate which discrete video cards have lost their meaning with the advent of the new generation of Intel graphics.

⇡ Graphics architecture Intel HD Graphics 4000/2500: what's new

The increase in performance of integrated graphics cores is far from the same simple task. And the fact that Intel was able to raise it by more than an order of magnitude in just a few years is actually the result of a serious engineering work. The main problem here is that integrated graphics accelerators cannot take advantage of dedicated high-speed video memory, but share regular video memory with the computing cores. system memory with a fairly low bandwidth by the standards of modern 3D applications. Therefore, optimizing memory is the very first step that must be taken when designing high-speed embedded graphics.

And Intel took this important step in the previous version of the microarchitecture - Sandy Bridge. The introduction of a ring intraprocessor bus that links together all CPU components (computational cores, third-level cache, graphics, system agent with a memory controller) opened up a short and progressive route for memory access for the built-in video core - through a high-speed third-level cache. In other words, the integrated graphics core, along with the computing processor cores, became an equal user of the L3 cache and memory controller, which significantly reduced downtime caused by waiting for graphics data to be processed. The ring bus turned out to be such a successful find from the previous design that it migrated to the new Ivy Bridge microarchitecture without any changes.

As for the internal structure of the Ivy Bridge graphics core, in general it can be considered a further development of the ideas inherent in the HD Graphics accelerators of previous generations. The architecture of the current Intel graphics core has its roots in the Clarkdale and Arrandale processors introduced in 2010, but each new reincarnation of it is not a simple copy of the previous design, but its improvement.

Ivy Bridge Generation HD Graphics Core Architecture

Thus, when moving from the Sandy Bridge microarchitecture to the Ivy Bridge, an increase in graphics performance is achieved primarily due to an increase in the number of execution units, especially since the internal structure of HD Graphics initially implied technical feasibility their simplest addition. While the older version of graphics from Sandy Bridge, HD Graphics 3000, had 12 devices, the most productive modification of the video core built into Ivy Bridge, HD Graphics 4000, received 16 actuators. However, the matter was not limited to this; the devices themselves were also improved. They added a second texture sampler, and throughput increased to three instructions per clock cycle.

The increase in the speed of data processing by the graphics core required developers to think again about their timely delivery. Therefore, the Ivy Bridge graphics core now has its own cache memory. Its volume has not been disclosed, however, apparently, we are talking about a small but high-speed internal buffer.

Although the innovations in the microarchitecture of the graphics core do not seem too significant at first glance, in total they result in a clearly noticeable naked eye 3D performance gains measured by the most by Intel like double. By the way, the next generation of HD Graphics accelerators, which will be built into processors of the Haswell family, should offer approximately the same increase. In them, the number of execution units will increase to 20, and the fourth level cache will be included in the fight to reduce latencies when the graphics core works with memory.

As for Ivy Bridge graphics, increasing its performance was not the only goal of the engineers. In parallel, the formal specifications of the new graphics core have been brought into line with modern requirements. This means that HD Graphics 4000 finally has full support for Shader Model 5.0 and hardware tessellation. That is, now Intel graphics are fully compatible “in hardware” with DirectX 11 and OpenGL 3.1 software interfaces. And of course, the work of HD Graphics 4000 in the upcoming operating room will not be a problem Windows system 8 - the necessary drivers are already available on the Intel website.

Intel also added to the new graphics core the ability to perform computational work using it; for this purpose, the new generation of HD Graphics added support for DirectCompute 5.0 and OpenCL. In Sandy Bridge processors these software interfaces were also supported, but at the driver level, which redirected the corresponding load to the computing cores. With the release of Ivy Bridge, full-fledged GPU computing became available on systems with Intel graphics.

In light of modern realities, Intel engineers paid attention to supporting multi-monitor configurations that are becoming increasingly popular. The HD Graphics 4000 graphics core was Intel's first integrated solution capable of running three independent displays. But keep in mind that to implement this function it was necessary to increase the width of the FDI bus, through which the image is transferred from the processor to the set system logic. So support for three monitors is only possible with new motherboards using seventh series chipsets.

In addition, there are some restrictions in resolutions and methods of connecting monitors. In a desktop platform based on processors of the Ivy Bridge family, theoretically you can get three outputs: the first is universal (HDMI, DVI, VGA or DisplayPort) with maximum resolution 1920x1200, second - DisplayPort, HDMI or DVI with resolution up to 1920x1200 and third - DisplayPort with support high resolutions up to 2560x1600. That is, the popular option of connecting WQXGA monitors via Dual-Link DVI with Intel HD Graphics 4000 is still impossible to implement. But the version of the HDMI protocol has been brought to 1.4a, and the DisplayPort protocol to 1.1a, which in the first case means support for 3D, and in the second - the ability of the interface to transmit an audio stream.

Innovations have also affected other components of the graphics core of Ivy Bridge processors, including their multimedia capabilities. High-quality hardware decoding of AVC/H.264, VC-1 and MPEG-2 formats was successfully implemented in the last generation of HD Graphics, but in Ivy Bridge graphics the AVC decoding algorithms have been adjusted. Due to the new design of the module responsible for context-adaptive encoding, the performance of the hardware decoder has increased, which has resulted in a theoretical possibility simultaneous playback multiple streams with high resolution, up to 4096x4096.

Considerable progress has also been made to Quick Sync technology, designed for fast hardware encoding video in AVC/H.264 format. Commissioned at Sandy Bridge, it was recognized as a colossal breakthrough a year and a half ago. Thanks to her Intel processors have moved into first place in the speed of transcoding high-resolution video, for which a separate hardware unit is now allocated, which is part of the graphics core. As part of HD Graphics 4000, Quick Sync technology has become even better and has an improved media sampler. As a result, the updated Quick Sync engine provides approximately a twofold advantage in transcoding speed to the H.264 format compared to its previous Sandy Bridge version. At the same time, as part of the technology, the quality of the video produced by the codec has also improved, and ultra-high resolution video content, up to 4096x4096, has become supported.

However, Quick Sync still has weaknesses. At the moment, this technology is used only in commercial video transcoding applications. There are no popular freely available utilities that work with this technology on the horizon. Another disadvantage of the technology is its close combination with the graphics core. If your system uses an external graphics card that disables general case integrated graphics, Quick Sync cannot be used. True, a solution to this problem can be offered by a third-party company, LucidLogix, which has developed the Virtu graphic virtualization technology.

Nevertheless, Quick Sync remains a unique technology for the market. A highly specialized hardware codec implemented within its framework turns out to be significantly better in all respects than encoding using the power of shader processors of modern video cards. Following Intel, only NVIDIA was able to implement a similar utilitarian hardware solution for encoding. And that company’s specialized tool, NVEnc, appeared only very recently - in Kepler generation accelerators.

⇡ Intel HD Graphics 4000 vs Intel HD Graphics 2500: what's the difference?

As before, Intel is integrating two graphics core options into Ivy Bridge. This time these are HD Graphics 4000 and HD Graphics 2500. The older and high-performance modification, which was primarily discussed in the previous section, has absorbed all the improvements inherent in the microarchitecture. The junior version of graphics is not aimed at establishing new performance standards for integrated solutions, but at simply providing modern processors with the minimum required level of graphics functionality.

The difference between HD Graphics 4000 and HD Graphics 2500 is dramatic. The fast version of the video core has sixteen actuators, while in the younger version their number is reduced to six. As a result, while HD Graphics 4000 delivers roughly 2x the theoretical 3D performance over the previous-generation HD Graphics 3000, HD Graphics 2500's performance advantage over HD Graphics 2000 is projected to be 10 to 20 percent. The same applies to the speed of Quick Sync - a twofold increase in speed compared to its predecessors is promised only in relation to older versions of the video core.

At the same time, the “full-fledged” HD Graphics 4000 core can be found not in all representatives of the Ivy Bridge generation, but mainly only in mobile devices, where graphics integrated into the CPU are most in demand. In desktop models, HD Graphics 4000 is present either in Core i7 series processors or in overclocking Core i5 series processors (with the K suffix in the model number) with the only exception to this rule - the Core i5-3475S processor. In all other cases, desktop system users have to either deal with HD Graphics 2500 or resort to the services of external graphics accelerators.

Fortunately, the widening gap between the older and younger modifications of Intel graphics occurred solely in performance. The functionality of HD Graphics 2500 was not affected at all. Just like HD Graphics 4000, the younger version has support for DirectX 11 and three-monitor configurations.

It should be noted that, as before, in different Core processors The third generation graphics core can operate at different frequencies. For example, Intel is more concerned about integrated graphics performance when it comes to mobile solutions, and this is reflected in frequencies. In general, Ivy Bridge mobile processors have an HD Graphics 4000 core running at slightly more high frequency than in the case of their desktop modifications. In addition, the difference in the frequency of integrated graphics may also be due to limitations in heat dissipation different models CPU.

In addition, the frequency of graphics operation is variable. IN Ivy processors Bridge implements a special Intel HD Graphics Dynamic Frequency technology, which interactively controls the frequency of the video core depending on the load on the processor cores and their current power consumption and heat dissipation.

Therefore, among the characteristics of specific HD Graphics implementations, two frequencies are indicated: minimum and maximum. The first is typical for the idle state, the second is the target frequency to which the graphics core seeks to accelerate, if current power consumption and heat dissipation allows, under load.

Part 18: Intel HD Graphics 4000 in different environments and the impact of the latter on the performance of the former

Processors based on the Ivy Bridge microarchitecture appeared a year ago, so everyone who follows this topic even a little knows the name of the older video core built into desktop Core i7s. That's right - Intel HD Graphics 4000. And if we go down a little lower in the ranking table, somewhere like the Core i3 level, then what will we find there? Most models have Intel HD Graphics 2500, but the i3-3225 and the recently announced 3245 still have the same HDG 4000. Laptop models have it too, and in all of them (with the exception of Celeron and Pentium, which are considered separately from the Core categories) : from the extreme i7-3940XM (four cores with a frequency of up to 3.9 GHz, TDP 55 W), to the tablet i3-3229Y (two cores with a frequency of 1.4 GHz, TDP 13 W). But is this video core the same? In case discrete video cards the question would be meaningless: it can be installed in a computer with any processor (at least theoretically). With an integrated solution, everything is more complicated. Firstly, even at a quick glance, the difference in the maximum operating frequency of the GPU is noticeable, and the range is extremely wide - from 850 MHz (just i3-3229Y) to 1.35 GHz (i7-3940XM), i.e. it differs by more than one and a half times. Secondly, we are not talking about some fixed frequencies - back in the first Generation Core GPU mobile processors started using Turbo technology Boost, and it is also used for processor cores. What does this lead to? The frequency of both changes dynamically, and depends both on the load on the CPU and GPU, and on which heat package ultimately needs to be “fitted”. In general, everything is unpredictable in advance, but there is an assumption that mobile graphics, although they have the same name as desktop graphics, work slower.

The discrepancy in end systems is not limited to GPU frequency alone. Even in the discrete graphics card market entry level their final characteristics are left to the manufacturers, and are not controlled in any way by the developer of the video processor itself. The discrepancy with the official performance characteristics can be significant, as we recently observed: four(!) out of five Palit video cards somewhat (to put it mildly) different from what NVIDIA intended. Moreover, it is easy to notice that the main differences did not even concern the frequencies of the chip, but the memory system. However, this is quite possible in the case of integrated graphics, especially since in this case Memory is rarely soldered on the board. Accordingly, options are possible. For example, the “official” DDR3-1600 or the slower DDR-1333 - whichever modules the manufacturer (or user) decides to use will be the same. But at least it's somehow manageable manual setting, but if the manufacturer decides to install only one SO-DIMM slot (inexpensive ultrabook models most often suffer from this, but not only them), we will get a completely different level of graphics core performance, despite the fact that the computer’s specifications will still indicate "Intel HD Graphics 4000".

Is it possible to test all the options and give a clear answer: what does each of them represent? It is possible, but difficult - the number of possible configurations is finite, but large. And it’s not very interesting to do this: it has long been known that HDG 4000 even in “ at its best"is not a full-fledged gaming solution, but to solve most other problems, as a rule, older and weaker GPUs are sufficient - up to HD Graphics Celeron processors on the Sandy Bridge core. On the other hand, you can try to estimate the approximate range where most solutions should fall - this is not so difficult. And in the process of a variety of testing, we have a certain set useful information accumulated In any case, it turned out that lately Using the same driver version (which is relevant in this case), we tested five different computer configurations for different purposes, having exactly the desired graphics subsystem. Thus, in this article we will simply put the results together and try to evaluate the influence of various factors on the performance of the Intel HD Graphics 4000 graphics core.

Test bench configuration

We have already indicated the range of potential clock frequencies above - from 850 MHz in Y-series processors to 1350 MHz in Core i7 Extreme Mobile. Thus, the most correct approach from a theoretical point of view would be to take two systems: a Core i3-3229Y (nowhere lower) and a Core i7-3940XM (no higher) and test them with different memory configurations - at least one and two channels , and at most also with different frequencies. Which is not feasible in practice. Firstly, it’s still difficult to find something with a Y-processor: such models have appeared quite recently, so most tablets in retail chains are equipped with the more familiar U or even M Core. Secondly, there is still no point in searching: the design of the tablet does not imply flexible configuration of the memory system - here you can “run into” memory modules soldered on the board and/or unavoidable single-channel. Thirdly, and at the top end, not everything is smooth - top-end laptops are free of the problems described above, however, processors of the XM and QM families (where maximum frequency graphics is 1.3 GHz) as a rule, are found on sale exclusively in pairs with discrete video cards, which cannot always be disabled. On the other hand, this also leads to the fact that there is simply no need to test extreme options - since the probability of encountering them in practice is zero or (in the case of Y) there are no options for choice anyway.

But the range of 1.05-1.15 GHz, on the contrary, is extremely interesting because it fits the majority possible options. It’s easy to see that three of the five configurations have already been tested by us - today the simply video-related results will be “expanded”. And supplemented by two more implementations - in Core i7-3770S and i7-3770K processors. The clock speed of the video core is 1.15 GHz, typical for many Core i7s, but two different frequencies memory. Plus there is a huge variation in terms of processor performance - let's see how it can affect the graphics results. And for comparison, we added the results of one processor with HDG 2500, but a powerful processor part - it suddenly turns out that ultramobile solutions, despite top-end (formally) graphics, are still significantly slower. If the processor part is equal, this, of course, is not observed, but with such a difference, anything can happen.

And an important point is different levels TDP of the tested processors; fortunately, five out of six support Turbo Boost technology for processor cores and all for GPUs. Why is this important? You may recall that in our power consumption tests, applying a load to the GPU increased it for the Core i7-3770K by 17 W. Naturally, a lot depends on the specific instance of the processor, especially since different series are subject to selection of different degrees of rigidity for this parameter - we also saw 20 W from the HDG 2500 in the budget i5-3450. But the order of magnitude itself is understandable and, in general, not small - dual-core U-series processors are limited to the same 17 W for the entire processor. And the 12 W official difference between the 3770S and 3770K is also bound to affect the operation of Turbo Boost when using the entire processor, and, therefore, performance.

Aliens vs. Predator

As we have written more than once, no integrated graphics can handle this game in this mode, so we get a pure stress test of the video core working at the limit of its capabilities. Moreover, anything can turn out to be a limiter on these capabilities: the equality of results between the Core i3-3217U and i7-3517U is very significant - despite potential differences, both models hit the same TDP. But two qualitative effects are clearly visible - firstly, single-channel memory is like death even for U-family processors (we have already seen that this is true for top models), and secondly, even in this mode the HDG 4000 is still faster, than 2500.

In low-quality mode, you can even try and play, and on any of the subjects. But in different ways: a low-frequency dual-core processor with single-channel DDR3-1333, but with HDG 4000, as it turns out, is suitable for this almost to the same extent as one of the older desktop models with HDG 2500! Despite the fact that the processor also works in this mode, it is not for nothing that two quad-core Core i7s are in first place. The difference between them is already relatively small, despite the fact that one model is generally top-end and works with more fast memory, and the second is energy efficient. 3217U and 3517U are much slower, although in their case there is some performance reserve that can slightly improve picture quality.

Batman: Arkham Asylum GOTY Edition

The relatively old and “light” graphics engine “loads” the GPU to a lesser extent, but has increased requirements for the processor component due to good multi-threaded optimization. As a result, desktop Core i7s already “pull out” the high-quality mode, and ultramobile processors are only close to this level. But they are very close, so with a slight decrease in quality they can reach a “playable” level. Unless, of course, you “press” the memory system - in single-channel mode, HDG 4000 is reduced to almost the level of 2500. But, by the way, not lower - the i5-3570S overtook the i5-3317U only due to the “full” four cores at a higher clock frequency and twice the amount of L3 cache.

With minimal quality, everything turns into a competition between processors. What is worth noting here is that such settings, as we see, still cannot be called completely irrelevant - for top processors with integrated graphics, the frame rate begins to go “off scale” beyond the threshold of sufficiency, but it is not only them that needs to be tested. On models for nettops and ultrabooks, the FPS is high, but not to say “excessive”.

Crysis: Warhead x64

Another stress test, where it is clearly visible, firstly, the complete incompetence of both systems with single-channel memory, such as the HDG 2500, and secondly, that the processor component, even in such conditions, still matters, affecting the final performance. On the other hand, first of all, still, the GPU, and then everything else.

Including in video modes that are potentially suitable for practical use (if, of course, someone enjoys looking at such a picture). In any case, the Core i7-3517U managed to overtake the Core i5-3570S due to its advantage in the graphics component, despite the fundamentally different processor.

F1 2010

As we have written more than once, the same frame rate in this game does not mean anything if it is equal to 12.5 FPS - a feature of the game engine, which tries to keep it at this level, discarding what is not essential (in its opinion).

In low quality, you can sometimes play on the HDG 4000, however, as we see, for this you need at least a Core i7-3517U (not the worst in its class, to put it mildly, and not cheap), and equipped with dual-channel memory with a frequency of 1600 MHz. Failure to comply with any of these conditions will result in consequences. Excess will change the picture to a lesser extent than the size of the excess :)

Far Cry 2

The performance of the HDG 4000 is still not enough for this old game (which is no longer news), but to a lesser extent than for Crysis or AvP, of course. It’s no wonder that the performance of the older and younger of the tested processors differs by one and a half times. On the other hand, from the point of view of worldly wisdom, we would not be surprised at a greater difference - after all, the CPU parts differ too much. One might even say, fundamentally and in all respects.

And in the minimum quality mode it comes to the fore. And the most curious result is that the Core i3-3217U, even in this case, could not reach the comfort threshold. That is, this game, almost five years old, still in no way lends itself not only to Atom or Brazos, but also to many high-efficiency platforms in general. And it doesn’t matter whether it’s with integrated video or any discrete video: the performance of the processor part itself is not enough. So progress is progress, and a certain minimum of system requirements must be provided. Which, as we see, older CULV processors can cope with without much safety margin, while younger ones cannot cope at all (it will be interesting to see how Kabini and younger Haswell fare with this). In general, a “fresh” tablet or budget ultrabook will not necessarily allow you to play even very old games and even at minimum settings.

Metro 2033

Let's return to the origins in the form of the first diagram - it is clear that not a single one of the subjects is enough for a high-quality mode of this game, and fundamentally not enough. But the influence of performance characteristics on performance is very clear, so we will not describe everything in detail - it’s easy to draw all the conclusions yourself.

Metro 2033 appeared a year and a half later than FC2, so minimum requirements to the equipment of the game is higher. To be fair, the “plinth” quality mode itself has much more high quality:) The minimum for it is Core i3-3225, i.e., in order to somehow play this game, we need a processor with a frequency above 3 GHz and HDG 4000, and both conditions are essential. The HDG 2500 won't run the game even with these settings, regardless of the processor. And weak models with any graphics will not cope with it precisely because they are weak.

We advise many laptop buyers to think about the latter;) Firstly, in light of these trends, the attempts of some manufacturers to equip their products with CULV processors with discrete video cards are beginning to look somewhat strange. In particular, we came across models with a Core i3-3217U paired with a GeForce GT 740M. The latest video card is another example of renaming and optimization, since it is practically the same 640M that has long been familiar to many, but with slightly increased frequencies. Not God knows what, of course, but potentially a couple of times faster than the same HDG 4000. However, as we see, the “processor independence” of games has its limit, especially when it comes to more or less modern projects, i.e. for Metro 2033 there are already few low-voltage dual-core models. Thus, a configuration similar to the one indicated will allow the user, perhaps, to increase the picture quality in old games, but not to play (at least somehow) new ones - you must agree, this is not an achievement for which it makes sense to pay for discrete graphics.

The second problem is from the same area: AMD never tires of repeating that, although its APU has lower processor performance, its graphics are more powerful than Intel's. As you can see, there are limits to everything - including the weak dependence of the results on the processor. And then the partners add fuel to the fire by adding to some A8-4555M (which at least feeds the built-in GPU) a discrete video card on something like Radeon HD 7550M/8550M. There is no doubt - Dual Graphics is sometimes the only way to increase the performance of the graphics subsystem, but this is only relevant when it is precisely insufficient. As you can see, not only this is possible in the low-consumption segment.

Summary results

Let's try to assess the situation in general, and also look not only at games, for which we will use diagrams with average results for a group of tests/applications (you can find out more about the full testing methodology in a separate article). The results in the diagrams are given in points, per 100 points in this article The performance of the Core i3-3217U is accepted as the slowest of the four processors tested. For those who are interested in more detailed information, again, it is traditionally proposed to download a table in Microsoft Excel format, in which all the results are presented both converted into points and in “natural” form.

So let's start with games. It is immediately clear that the single-channel memory mode instantly drops HDG 4000 to the level of 2500 and others similar decisions, so it’s not very relevant for practical use. Under normal conditions, the difference in results is 33%. On the one hand, there is a lot, on the other, everything is different. Even TDP is 4.5 times different. But if such freedom is not given, and the same memory type DDR3-1333 is used, then 15% will not be gained. Which is easily explained - after all, the video core itself is the same (adjusted for the influence of the thermal package on its actual clock frequency), and taking into account its power, heavy gaming applications are the stress test for it in the first place.

But in practice, as we have already seen, in such conditions the frame rate is almost universally too low to be used, so modes with reduced graphics quality are more relevant. For many solutions - reduced to a minimum: this mode is too easy for top solutions, but CULV processors, as we see, do not always cope even in it. And here the dependence of the results on the processor part is visible to the naked eye, so that 33% turns into 128% - no comments necessary. Moreover, we note that a “normal desktop” processor with HDG 2500 outperforms even the CULV Core i7 (3517U, of course, is a junior model, but the older 3687U only differs by a 10% increase in maximum clock frequency, which may not be enough), but by one and a half times behind a “normal desktop” processor with HDG 4000.

If this load were multi-threaded, most likely we would get a spread of results as in the previous case, but “only” 1.87 times. But the situation inside is different: there is practically no difference between HDG 2500 and 4000. It is not surprising that the memory operating mode has an effect, but only weakly - the higher clock frequency of the processor more than covers this difference.

At the time of GMA and the first versions of HDG, these results also depended on the video core, but now, as we see, they have stopped. Well, we will take this into account when developing the next versions of test methods :)

Total

So, as you would expect, we confirmed the dependence of the performance of integrated graphic solutions from the processors in which they are integrated. However, we note that it is not always so strong. As one would expect, when the load falls on the GPU, a large scatter of results can only be detected when comparing processors with fundamentally different thermal packages, since it also affects the frequencies of the graphics core. But such modes are guaranteed to be too “heavy” not only for IGP, but also for younger models of discrete video cards, so in order to play on them in practice (and not just watch a slide show), you have to reduce the picture quality, i.e. i.e. reduce the load on the GPU and increase it on the CPU. While the latter belong to the same class, the determining factor continues to be the power of the graphics core itself (which we have already seen in the example of desktop solutions, where a pair of high-frequency cores and a TDP margin allowed the same HDG 4000 to deploy to the full extent of its weak strengths and paired with different processors ), but you should no longer expect the same level of performance from ultrabook and desktop processors. In principle, it would be difficult to assume the opposite, but it is never superfluous to make sure that this is exactly the state of affairs. The love of naming solutions that are similar in architecture but different in performance began, of course, not with Intel, but in most cases, manufacturers still at least somehow hint at the existence of a difference. Yes, the company itself adheres to the same practice in the system of naming processors - giving them non-overlapping numbers and not forgetting to add the letter “M” or “U” at the end, sometimes dramatically affecting the family number (a hackneyed example: the vast majority of desktop Core i5s are to quad-core processors, but all Core i5-M are only dual-core). But with graphics there is not even such clarity: one can judge only by indirect signs - such as the name of the processor in which it is built.

Is there any hope of stopping the resulting mess in the future? Maybe in a distant, but definitely not in the next generation of processors. That is, we, of course, have no doubt that the Iris 5100 is a more powerful GPU than the HDG 4600. However, will this allow playing on the Core i7-4558U (dual-core SoC with a TDP of 15 W) with greater comfort than on the Core i7-4700HQ? not to mention the older desktop Core i7-4770K (quad-core processors, which are also faster than the 4558U in clock frequency and less “squeezed” by the thermal package) - the question is open. And the complete equality of processors with the so-called equally integrated GPU is even more doubtful. However, it is impossible to accurately understand these issues without direct testing, and this is a topic for completely different testing.

Intel HD 4000 – integrated graphics installed in Intel Core i3, Core i5 and Core i7 processors of the Ivy Bridge generation, which appeared in 2011. The video core is already quite old and cannot boast of decent characteristics and performance.

Graphics card specifications

The characteristics of the HD 4000 are quite modest even at the time the graphics chip was released; at the moment they look truly ridiculous.

The device can provide 16 unified processors. The highest clock frequency of the graphics chip can reach 1350MHz. The actual frequency depends on many factors, including the model of the processor into which the chip will be integrated, as well as the type of device. Laptops and other unproductive devices are almost always reduced in frequency CPU operation and video cores.

The amount of memory available for the needs of the video adapter will depend on two factors: BIOS settings and the amount of RAM installed on the computer. If you are seriously going to use this particular graphics chip, you should fork out for good RAM sticks with higher frequencies.

The memory bus width reaches 128 bits (in dual-channel RAM mode, single-channel mode will allow you to get only 64 bits).

Intel HD 4000 has support for DirectX 11.1, OpenGL 4.1 and Quick Sync. You can’t even dream of any DirectX 12, OpenGL 4.5, OpenCL and Vulcan with this video card, it doesn’t support them.

What tasks is the Intel HD 4000 suitable for?

First of all - office work in undemanding applications or ensuring the browser works. Almost any more or less current video card can cope with such tasks, and the Intel HD 4000 is no exception.

It’s suitable for watching movies, but it’s better not to use it for high resolutions. It will perfectly play movies and other videos in HD or FullHD resolution, but it won’t cope with the increasingly popular UltraHD (4K), it simply won’t have enough performance. If you don't have a monitor or TV that supports UltraHD, then the Intel HD 4000 is quite enough for watching movies. Owners of modern 4K panels should take a closer look at a video card that has better performance than HD 4000.

With games, the HD 4000 is even worse. Even at the time of release (in 2011), the video card could not run absolutely all current games with sufficient performance.

The Intel HD 4000 will handle games from 2010 or earlier very well, although not perfectly. Some projects fundamentally refuse to run normally on older integrated video cards, which can result in some rather strange problems.

The Intel HD 4000 is practically unsuitable for working in specific software (video editors, 3D modeling, rendering). The video core only supports Intel Quick Sync technology, which cannot be called particularly widespread. The more common OpenCL on this graphics chip not supported. Even if the desired application allows you to use Quick Sync, the Intel HD 4000 does not have enough performance to run such software.

Drivers

Installing a driver on Windows is quite simple; all you need to do is download it and run the installation package; nothing more is required from you. The update can be done in two ways. The first is to use Intel settings or automatic update programs. Second - manually download new version driver and reinstall it.

In operating systems of the Linux family, everything is quite sad. The proprietary driver (developed by Intel) is available only on newer models of Intel HD video cards; this video adapter is not supported. Therefore, under Linux, you can only use a free driver, which is inferior in almost all aspects to the driver on Windows. The proprietary driver is updated automatically along with the operating system, but if you want to install a version that is not available on your distribution, you will need to update the Mesa 3D kernel and libraries.

Comparison with discrete video cards

If we compare with, then the Intel HD 4000 can compete only with the weakest video adapters, such as the GT 620. More powerful graphics adapters are already more powerful than the HD 4000.

In general, the Intel HD 4000 can provide only the most basic functionality of a video card and serve instead of the weakest plug.