The card has 12288 MB of GDDR5 SDRAM located in 24 4 Gbit chips (12 on each side of the PCB).

We used examples from Microsoft and AMD SDKs, as well as the Nvidia demo program for our synthetic DirectX 11 benchmarks. The first is HDRToneMappingCS11.exe and NBodyGravityCS11.exe from the DirectX SDK (February 2010). We also took applications from both video chip manufacturers: Nvidia and AMD. Samples DetailTessellation11 and PNTriangles11 were taken from ATI Radeon SDK (they are also in DirectX SDK). Additionally, Nvidia's demo program, Realistic Water Terrain, also known as Island11, was used.

Synthetic tests were carried out on the following video cards:

• Geforce GTX Titan X GTX Titan X)
• Geforce GTX Titan Z with standard parameters (abbreviated GTX Titan Z)
• Geforce GTX 980 with standard parameters (abbreviated GTX 980)
• Radeon R9 295X2 with standard parameters (abbreviated R9 295X2)
• Radeon R9 290X with standard parameters (abbreviated R9 290X)

To analyze the performance of the new model of the Geforce GTX Titan X video card, these solutions were chosen for the following reasons. The Geforce GTX 980 is based on the GPU of the same Maxwell architecture, but at a lower level - the GM204, and it will be very interesting for us to evaluate what caused the complication of the chip to the GM200. Well, the dual-GPU Geforce GTX Titan Z is just taken as a guideline - as the most productive Nvidia video card based on a pair of GK110 chips from the previous Kepler architecture.

We also chose two video cards from rival AMD for our comparison. They are very different in principle, although they are based on the same Hawaii GPUs - they just have different number of GPUs on the cards and they differ in positioning and price. The Geforce GTX Titan X has no price rivals, so we took the most powerful dual-GPU video card, the Radeon R9 295X2, although such a comparison will not be very interesting from the technical point of view. For the latter, the fastest single-chip video card of a competitor is taken - Radeon R9 290X, although it was released too long ago and is based on a GPU of clearly lower complexity. But there is simply no other choice from AMD solutions.

Direct3D 10: PS 4.0 Pixel Shader Tests (Texturing, Loops)

We have abandoned the outdated DirectX 9 tests, as ultra-powerful solutions like the Geforce GTX Titan X show poor results in them, being always limited by memory bandwidth, fill rate or texturing. Not to mention the fact that dual-GPU video cards do not always work correctly in such applications, and we have two of them.

The second version of RightMark3D includes two already familiar PS 3.0 tests for Direct3D 9, which were rewritten for DirectX 10, as well as two more new tests. The first pair adds the ability to enable self-shadowing and shader supersampling, which additionally increases the load on video chips.

These tests measure the performance of executing pixel shaders with loops with a large number of texture samples (in the heaviest mode, up to several hundred samples per pixel) and a relatively low ALU load. In other words, they measure the texture sampling rate and branching efficiency in a pixel shader.

The first pixel shader test will be Fur. At the lowest settings, it uses 15 to 30 texture samples from the heightmap and two samples from the main texture. Effect detail - “High” mode increases the number of samples up to 40–80, enabling “shader” supersampling - up to 60–120 samples, and the “High” mode, together with SSAA, has a maximum “severity” - from 160 to 320 samples from the height map.

Let's first check the modes without supersampling enabled, they are relatively simple, and the ratio of the results in the “Low” and “High” modes should be approximately the same.

Performance in this test depends on the number and efficiency of TMUs, as well as the efficiency of complex programs execution. And in the variant without supersampling, the effective fill rate and memory bandwidth also have an additional effect on performance. The results at the level of detail "High" are obtained up to one and a half times lower than at "Low".

In the tasks of procedural rendering of fur with a large number of texture fetches, with the release of video chips based on GCN architecture, AMD has long seized the leadership. It is the Radeon boards that are the best in these comparisons to this day, which indicates that they are more efficient in performing these programs. This conclusion is confirmed by today's comparison - the Nvidia video card we are considering lost even to the outdated single-chip Radeon R9 290X, not to mention the closest price competitor from AMD.

In the first Direct3D 10 test, the new video card Geforce GTX Titan X turned out to be slightly faster than its younger sister on a chip of the same architecture in the form of a GTX 980, but the lag of the latter is small - 9-12%. This result can be explained by the noticeably lower texturing speed of the GTX 980, and it lags behind in other parameters, although the matter is clearly not in the performance of ALUs. The dual-GPU Titan Z is faster, but not as fast as the Radeon R9 295X2.

Let's look at the result of the same test, but with “shader” supersampling enabled, which quadruples the work: in such a situation, something should change, and the memory bandwidth with fill rate will have a smaller effect:

In difficult conditions, the new Geforce GTX Titan X video card is already more noticeably ahead of the younger model from the same generation - GTX 980, being faster by a decent 33-39%, which is much closer to the theoretical difference between them. And the gap with competitors in the form of the Radeon R9 295X2 and R9 290X has decreased - the new product from Nvidia has almost caught up with the single-chip Radeon. However, the dual-GPU remained far ahead, because AMD chips prefer per-pixel calculations and are very strong in such calculations.

The next DX10 test measures the performance of complex pixel shaders with loops with a large number of texture samples and is called Steep Parallax Mapping. At low settings, it uses 10 to 50 texture samples from the heightmap and three samples from the main textures. When you turn on the heavy mode with self-shadowing, the number of samples doubles, and supersampling increases this number four times. The most difficult test mode with supersampling and self-shadowing selects from 80 to 400 texture values, that is, eight times more than the simple mode. We first check the simple options without supersampling:

The second pixel-shader test Direct3D 10 is more interesting from a practical point of view, since varieties of parallax mapping are widely used in games, and heavy options like steep parallax mapping have long been used in many projects, for example, in games from the Crysis, Lost Planet and many others series. Besides, in our test, in addition to supersampling, you can enable self-shadowing, which doubles the load on the video chip - this mode is called “High”.

The diagram is generally similar to the previous one, also without supersampling enabled, and this time the new Geforce GTX Titan X turned out to be a little closer to the GTX Titan Z, losing not so much to a dual-GPU board based on a pair of Kepler GPUs. Under different conditions, the new product is 14-19% ahead of the previous top-end model of the current generation from Nvidia, and even if we compare it with AMD video cards, something has changed - in this case, the new GTX Titan X is slightly inferior to the Radeon R9 290X. The two-chip R9 295X2, however, is far ahead of everyone. Let's see what will change the inclusion of supersampling:

When supersampling and self-shadowing are enabled, the task becomes more difficult; enabling two options at once increases the load on the cards by almost eight times, causing a serious drop in performance. The difference between the speed indicators of the tested video cards has changed slightly, although the inclusion of supersampling has less effect than in the previous case.

AMD Radeon graphics solutions perform more efficiently in this D3D10 pixel shader test than competing Geforce cards, but the new GM200 chip changes the situation for the better - Geforce GTX Titan X on a Maxwell architecture chip outperforms Radeon R9 290X in all conditions ( however, based on a noticeably less complex GPU). The dual-GPU solution based on the Hawaii pair remained the leader, but in comparison with other Nvidia solutions, the new product is not bad. It showed speeds almost on par with the dual-GPU Geforce GTX Titan Z, and outperformed the Geforce GTX 980 by 28-33%.

Direct3D 10: PS 4.0 Pixel Shader Benchmarks (Compute)

The next couple of pixel shader tests contain the minimum number of texture fetches to reduce the impact of TMU performance. They use a large number of arithmetic operations, and they measure exactly the mathematical performance of video chips, the speed of execution of arithmetic instructions in a pixel shader.

The first math test is Mineral. This is a complex procedural texturing test that uses only two texture data samples and 65 instructions like sin and cos.

The results of extreme mathematical tests most often correspond to the difference in frequencies and the number of computing units, but only approximately, since the results are influenced by the different efficiency of their use in specific tasks, and driver optimization, and the latest frequency and power control systems, and even the emphasis on memory bandwidth. ... In the case of the Mineral test, the new Geforce GTX Titan X is only 10% faster than the GTX 980 based on the GM204 chip from the same generation, and the dual-GPU GTX Titan Z was not so fast in this test - there is clearly something preventing Nvidia boards from revealing ...

Comparing the Geforce GTX Titan X to competing boards from AMD would not be so frustrating if the GPUs in the R9 290X and Titan X were similar in complexity. But the GM200 is much larger than the Hawaii, and its small win is only natural. Upgrading Nvidia's architecture from Kepler to Maxwell brought the new chips closer to competing solutions from AMD in such tests. But even the cheaper dual-GPU solution Radeon R9 295X2 is noticeably faster.

Consider the second shader computation test called Fire. It is heavier for ALU, and there is only one texture fetch in it, and the number of instructions like sin and cos is doubled, to 130. Let's see what has changed with increasing load:

In the second math test from RigthMark, we see different results of video cards relative to each other. So, the new Geforce GTX Titan X is already stronger (by 20%) ahead of the GTX 980 on a chip of the same graphics architecture, and the dual-GPU Geforce is very close to the new product - Maxwell copes with computational tasks much better than Kepler.

The Radeon R9 290X is left behind, but as we already wrote, the Hawaii GPU is noticeably simpler than the GM200, and this difference is logical. But although the dual-GPU Radeon R9 295X2 continues to be the leader in the tests of mathematical calculations, in general, the new Nvidia video chip performed well in such tasks, although it did not achieve the theoretical difference with the GM204.

Direct3D 10: geometry shader benchmarks

The RightMark3D 2.0 package contains two tests of the speed of geometry shaders, the first version is called "Galaxy", the technique is similar to the "point sprites" from previous versions of Direct3D. It animates a particle system on a GPU, a geometric shader creates four vertices from each point, forming a particle. Similar algorithms should be widely used in future DirectX 10 games.

Changing the balancing in the geometry shader tests does not affect the final rendering result, the final image is always exactly the same, only the scene processing methods change. The "GS load" parameter determines in which of the shaders the calculations are performed - in vertex or geometric. The number of calculations is always the same.

Let's consider the first variant of the Galaxy test, with calculations in a vertex shader, for three levels of geometric complexity:

The ratio of speeds with different geometric complexity of scenes is approximately the same for all solutions, the performance corresponds to the number of points, with each step the FPS drop is close to twofold. This task for powerful modern video cards is very simple, and its performance is limited by the geometry processing speed, and sometimes by the memory bandwidth and / or fill rate.

The difference between the results of video cards based on Nvidia and AMD chips is usually in favor of the solutions of the Californian company, and it is due to differences in the geometric pipelines of the chips of these companies. In this case, the top-end Nvidia video chips have a lot of geometry processing units, so the gain is obvious. In geometry tests, Geforce cards are always more competitive than Radeon cards.

The new Geforce GTX Titan X lags slightly behind the previous generation dual-GPU GTX Titan Z, but the GTX 980 outperforms the GTX 980 by 12-25%. Radeon video cards show noticeably different results, since the R9 295X2 is based on a pair of GPUs, and only it can compete with the novelty in this test, and the Radeon R9 290X has become an outsider. Let's see how the situation will change when transferring part of the calculations to the geometry shader:

When the load changed in this test, the numbers changed slightly for AMD boards and for Nvidia solutions. And that doesn't really change anything. In this test of geometry shaders, video cards react poorly to changes in the GS load parameter, which is responsible for transferring part of the calculations to the geometry shader, so the conclusions remain the same.

Unfortunately, "Hyperlight" is the second geometry shader test that demonstrates the use of several techniques at once: instancing, stream output, buffer load, which uses dynamic geometry creation by drawing in two buffers, as well as a new Direct3D 10 feature - stream output, on all modern graphics cards from AMD just don't work. At some point, another update of Catalyst drivers led to the fact that this test stopped running on boards of this company, and this has not been fixed for several years.

Direct3D 10: speed of fetching textures from vertex shaders

The Vertex Texture Fetch tests measure the speed of a large number of texture fetches from the vertex shader. The tests are similar, in fact, so the ratio between the results of the cards in the tests "Earth" and "Waves" should be approximately the same. Both tests use displacement mapping based on texture fetch data, the only significant difference is that the Waves test uses conditional transitions, while the Earth test does not.

Consider the first test "Earth", first in the "Effect detail Low" mode:

Our previous research showed that the results of this test can be influenced by both the fill rate and the memory bandwidth, which is clearly visible in the results of Nvidia boards, especially in simple modes. The new video card from Nvidia in this test demonstrates the speed clearly lower than it should - all Geforce cards turned out to be approximately at the same level, which clearly does not correspond to theory. In all modes, they obviously run into something like the memory bandwidth. However, the Radeon R9 295X2 is also not twice as fast as the R9 290X.

By the way, the single-chip board from AMD this time turned out to be stronger than all the boards from Nvidia in the easy mode and approximately at their level in the heavy mode. Well, the dual-GPU Radeon R9 295X2 again became the leader of our comparison. Let's look at the performance in the same test with an increased number of texture fetches:

The situation on the diagram has slightly changed, the single-chip solution from AMD in heavy modes has lost significantly more Geforce cards. The new Geforce GTX Titan X showed speeds up to 14% faster than the Geforce GTX 980, and outperformed the single-GPU Radeon in all but the lightest modes due to the same focus on something. If we compare the new product with a dual-GPU solution from AMD, then Titan X was able to fight in heavy mode, showing similar performance, but lagging behind in light modes.

Let's consider the results of the second test of texture samples from vertex shaders. The Waves test has a smaller number of samples, but it uses conditional jumps. The number of bilinear texture samples in this case is up to 14 ("Effect detail Low") or up to 24 ("Effect detail High") for each vertex. The complexity of the geometry changes in the same way as in the previous test.

The results in the second test of vertex texturing "Waves" are not at all similar to what we saw in the previous diagrams. The speed indicators of all GeForce in this test seriously deteriorated, and the new model Nvidia Geforce GTX Titan X shows speed only slightly faster than the GTX 980, lagging behind the dual-GPU Titan Z. all modes. Let's consider the second variant of the same problem:

With the complication of the task in the second test of texture sampling, the speed of all solutions became lower, but Nvidia video cards suffered more, including the model in question. In the conclusions, almost nothing changes, the new Geforce GTX Titan X is up to 10-30% faster than the GTX 980, lagging behind both the dual-GPU Titan Z and both Radeon boards. The Radeon R9 295X2 turned out to be far ahead in these tests, and from the point of view of theory this is simply inexplicable, except for insufficient optimization from Nvidia.

3DMark Vantage: Feature Benchmarks

The synthetic benchmarks from the 3DMark Vantage suite will show us what we missed earlier. Feature tests from this test suite have DirectX 10 support, are still relevant and interesting because they differ from ours. When analyzing the results of the latest video card model Geforce GTX Titan X in this package, we will draw some new and useful conclusions that eluded us in the tests from the RightMark family of packages.

The first test measures the performance of texture fetch units. Used to fill a rectangle with values ​​read from a small texture using multiple texture coordinates that change every frame.

The efficiency of AMD and Nvidia video cards in the texture test by Futuremark is quite high and the final figures for different models are close to the corresponding theoretical parameters. So, the difference in speed between GTX Titan X and GTX 980 turned out to be 38% in favor of a solution based on GM200, which is close to theory, because the new product has one and a half times more TMUs, but they operate at a lower frequency. Naturally, the lag behind the dual-GPU GTX Titan Z remains, since the two GPUs have a high texturing speed.

When it comes to comparing the texturing speed of the new top-end video card from Nvidia with similarly priced solutions from a competitor, the new product is inferior to a two-GPU rival, which is a conditional neighbor in the price niche, but outperforms the Radeon R9 290X, although not too significantly. Still, AMD graphics cards still do a little better with texturing.

The second task is to test the fill rate. It uses a very simple pixel shader with no performance cap. The interpolated color value is written to the offscreen buffer (render target) using alpha blending. A 16-bit FP16 off-screen buffer is used, which is most often used in games that use HDR rendering, so this test is quite timely.

The numbers of the second 3DMark Vantage subtest show the performance of ROP units without taking into account the amount of video memory bandwidth (the so-called "effective fill rate"), and the test measures ROP performance. The Geforce GTX Titan X board we are reviewing is noticeably ahead of both Nvidia boards, the GTX 980 and even the GTX Titan Z, outperforming the single-chip board based on GM204 by as much as 45% - the number of ROPs and their efficiency in the top GPU of the Maxwell architecture is excellent!

And if we compare the scene fill rate with the new Geforce GTX Titan X video card with AMD video cards, then the Nvidia board we are considering in this test shows the best scene fill rate even in comparison with the most powerful dual-GPU Radeon R9 295X2, not to mention the considerably lagging Radeon R9 290X. A large number of ROPs and optimizations for the compression efficiency of the framebuffer data did their job.

One of the most interesting feature tests, as a similar technique is already used in games. It draws one quadrilateral (more precisely, two triangles) using a special technique called Parallax Occlusion Mapping, which simulates complex geometry. Quite resource-intensive ray tracing and high resolution depth map are used. Also this surface is shaded using the heavy Strauss algorithm. This is a test of a very complex and GPU-heavy pixel shader containing numerous texture selections for ray tracing, dynamic branches and complex lighting calculations by Strauss.

This test from the 3DMark Vantage package differs from the previous ones in that the results in it depend not only on the speed of mathematical calculations, the efficiency of branching or the speed of texture fetching, but on several parameters simultaneously. To achieve high speed in this task, the correct balance of the GPU is important, as well as the efficiency of the execution of complex shaders.

In this case, both mathematical and texture performance are important, and in this synthetic from 3DMark Vantage the new Geforce GTX Titan X is more than a third faster than the model based on the GPU of the same Maxwell architecture. And even the dual-GPU Kepler in the form of GTX Titan Z won less than 10% over the new product.

The single-chip top-end Nvidia board in this test showed the result clearly better than the one-chip Radeon R9 290X, but both are very seriously inferior to the two-GPU model Radeon R9 295X2. AMD graphics processors work somewhat more efficiently in this task than Nvidia chips, while the R9 295X2 has two of them.

The fourth test is interesting in that it calculates physical interactions (tissue imitation) using a video chip. Vertex simulation is used, using combined work of vertex and geometry shaders, with several passes. Use stream out to transfer vertices from one simulation pass to another. Thus, the performance of vertex and geometry shaders execution and stream out speed are tested.

Rendering speed in this test also depends on several parameters at once, and the main factors of influence should be geometry processing performance and geometry shader execution efficiency. That is, the strengths of Nvidia chips should be manifested, but alas - we saw a very strange result (we rechecked), the new Nvidia video card showed not very high speed, to put it mildly. Geforce GTX Titan X in this subtest showed the worst result of all solutions, lagging almost 20% behind even the GTX 980!

Well, the comparison with Radeon boards in this test is just as unsightly for a new product. Despite the theoretically smaller number of geometric execution units and the geometric performance lag of AMD chips compared to competing solutions, both Radeon cards perform very efficiently in this test and outperform all three Geforce cards presented in comparison. Again, it looks like a lack of optimization in Nvidia drivers for a specific task.

Physical simulation test of effects based on particle systems calculated using a video chip. Vertex simulation is also used, each vertex represents a single particle. Stream out is used for the same purpose as in the previous test. Several hundred thousand particles are calculated, all are animated separately, and their collisions with the height map are also calculated.

Similar to one of the tests in our RightMark3D 2.0, particles are rendered using a geometry shader, which creates four vertices from each point that form a particle. But the test most of all loads shader units with vertex calculations, stream out is also tested.

In the second "geometric" test from 3DMark Vantage, the situation has changed significantly, this time all Geforce already show more or less normal results, although the dual-GPU Radeon still remains in the lead. The new GTX Titan X model is 24% faster than its sister GTX 980 and is about the same lagging behind the two-GPU Titan Z on the previous generation GPU.

Comparison of the new Nvidia with competing video cards from AMD this time is more positive - it showed the result between two boards from a rival company, and turned out to be closer to the Radeon R9 295X2, which has two GPUs. The novelty is significantly ahead of the Radeon R9 290X and this clearly shows us how different two seemingly similar tests can be: simulating fabrics and simulating a particle system.

The latest feature test of the Vantage package is a mathematically intensive test of the video chip, it calculates several octaves of the Perlin noise algorithm in a pixel shader. Each color channel uses its own noise function for more load on the video chip. Perlin noise is a standard algorithm often used in procedural texturing and uses a lot of math.

In this case, the performance of solutions does not quite correspond to theory, although it is close to what we saw in similar tests. In the mathematical test from the Futuremark suite, which shows the peak performance of video chips in extreme tasks, we see a different distribution of results compared to similar tests from our test suite.

We have known for a long time that GPUs from AMD with GCN architecture still cope with such tasks better than competing solutions, especially in cases where intensive "math" is performed. But the new top model from Nvidia is based on a large GM200 chip, and therefore the Geforce GTX Titan X in this test showed a result noticeably higher than the Radeon R9 290X.

If we compare the new product with the best model of the Geforce GTX 900 family, then in this test the difference between them was almost 40% - in favor of the video card we are considering today, of course. This is also close to the theoretical difference. Not a bad result for the Titan X, only the dual-GPU Radeon R9 295X2 is ahead, and far ahead.

Direct3D 11: Compute Shaders

We used examples from the SDKs and demos from Microsoft, Nvidia, and AMD to test Nvidia's recently released top-of-the-line solution for tasks using DirectX 11 features such as tessellation and compute shaders.

First, we'll look at benchmarks using Compute shaders. Their appearance is one of the most important innovations in the latest versions of the DX API, they are already used in modern games to perform various tasks: post-processing, simulations, etc. The first test shows an example of HDR rendering with tone mapping from the DirectX SDK, with post-processing using pixel and compute shaders.

The calculation speed in the computational and pixel shaders is approximately the same for all AMD and Nvidia boards, the differences were observed only in video cards based on GPUs of previous architectures. Judging by our previous tests, the results in a task often depend not so much on the mathematical power and computational efficiency as on other factors, such as memory bandwidth.

In this case, the new top-end video card is faster than the single-chip versions Geforce GTX 980 and Radeon R9 290X, but lags behind the two-chip R9 295X2, which is understandable, because it has the power of a pair of R9 290X. If we compare the new product with the Geforce GTX 980, then the board of the Californian company being considered today is 34-36% faster - exactly according to theory.

The second computational shader test, also taken from the Microsoft DirectX SDK, shows the computational N-body gravity problem - a simulation of a dynamic particle system that is acted upon by physical forces such as gravity.

In this test, the emphasis is most often observed on the speed of execution of complex mathematical calculations, geometry processing and the efficiency of code execution with branches. And in this DX11 test, the balance of forces between the solutions of two different companies turned out to be completely different - obviously in favor of Geforce video cards.

However, the results of a couple of Nvidia solutions on different chips are also strange - Geforce GTX Titan X and GTX 980 are almost equal, they are separated by only 5% of the performance difference. Dual-GPU rendering does not work in this task, so the rivals (single-GPU and dual-GPU Radeon models) are roughly equal in speed. Well, the GTX Titan X is three times ahead of them. It seems that this task is much more efficiently calculated on GPUs of the Maxwell architecture, which we noted earlier.

Direct3D 11: Tessellation Performance

Compute shaders are very important, but another major innovation in Direct3D 11 is hardware tessellation. We examined it in great detail in our theoretical article about the Nvidia GF100. Tessellation has long been used in DX11 games such as STALKER: Call of Pripyat, DiRT 2, Aliens vs Predator, Metro Last Light, Civilization V, Crysis 3, Battlefield 3 and others. Some of them use tessellation for character models, while others use tessellation to simulate a realistic water surface or landscape.

There are several different schemes for partitioning graphic primitives (tessellations). For example phong tessellation, PN triangles, Catmull-Clark subdivision. So, the PN Triangles partitioning scheme is used in STALKER: Call of Pripyat, and in Metro 2033 - Phong tessellation. These methods are relatively quickly and easily introduced into the game development process and existing engines, and therefore have become popular.

The first tessellation test will be the Detail Tessellation example from the ATI Radeon SDK. It implements not only tessellation, but also two different per-pixel processing techniques: simple overlay normal maps and parallax occlusion mapping. Well, let's compare DX11 solutions from AMD and Nvidia in different conditions:

In the simple bumpmapping test, the speed of the boards is not very important, since this task has long become too easy, and the performance in it is limited by memory bandwidth or fill rate. Today's hero of the review is 23% ahead of the previous top-end model Geforce GTX 980 based on the GM204 chip and is slightly inferior to the competitor in the form of the Radeon R9 290X. The dual-chip version is even slightly faster.

In the second subtest with more complex pixel-by-pixel calculations, the new product has already become 34% faster than the Geforce GTX 980, which is closer to the theoretical difference between them. But Titan X this time is already a little faster than a single-chip conditional competitor based on a single Hawaii. Since the two chips in the Radeon R9 295X2 work perfectly, this task is performed even faster on it. Although the efficiency of performing mathematical calculations in pixel shaders is higher for chips of the GCN architecture, the release of solutions of the Maxwell architecture has improved the position of Nvidia solutions.

In the subtest using a light degree of tessellation, the recently announced Nvidia board is again only a quarter faster than the Geforce GTX 980 - perhaps the speed is limited by the memory bandwidth, since texturing has almost no effect in this test. If we compare the new product with AMD boards in this subtest, then the Nvidia board is again inferior to both Radeon cards, since in this tessellation test the partitioning of triangles is quite moderate and the geometric performance does not limit the overall rendering speed.

The second test of tessellation performance will be another example for 3D developers from ATI Radeon SDK - PN Triangles. Actually, both examples are also included in the DX SDK, so we are sure that game developers create their code on their basis. We tested this example with a different tessellation factor to see how much of a change would have an effect on overall performance.

In this test, more complex geometry is used, therefore, a comparison of the geometric power of various solutions yields different conclusions. The modern solutions presented in the material cope quite well with light and medium geometric loads, showing high speed. But although in light conditions, the Hawaii GPUs in the Radeon R9 290X and R9 295X2 in the number of one and two pieces work fine, in heavy conditions the Nvidia boards go far ahead. So, in the most difficult modes, the Geforce GTX Titan X presented today shows the speed noticeably better than the dual-GPU Radeon.

As for the comparison of Nvidia boards based on GM200 and GM204 chips with each other, the Geforce GTX Titan X model under consideration today increases its advantage with an increase in geometric load, since in light mode everything depends on memory bandwidth. As a result, the new product outperforms the Geforce GTX 980 board, depending on the mode complexity, by up to 31%.

Let's take a look at the results of another test, the Nvidia Realistic Water Terrain demo program, also known as Island. This demo uses tessellation and displacement mapping to render realistic looking ocean and terrain surfaces.

The Island test is not a purely synthetic test for measuring exclusively the geometric performance of the GPU, since it contains complex pixel and computational shaders, including, and such a load is closer to real games in which all GPU units are used, and not just geometric ones, as in previous geometry tests. Although the main one is still the load on the geometry processing units, the same memory bandwidth, for example, can also affect.

We test all video cards at four different tessellation ratios - in this case, the setting is called Dynamic Tessellation LOD. At the first partitioning factor of triangles, the speed is not limited by the performance of geometric units, and Radeon video cards show a fairly high result, especially the dual-GPU R9 295X2, even surpassing the result of the announced Geforce GTX Titan X card, but already at the next stages of geometric load, the performance of Radeon cards decreases, and solutions Nvidia comes out ahead.

The advantage of the new Nvidia board based on the GM200 video chip over its competitors in such tests is already quite decent, and even multiple. If we compare Geforce GTX Titan X with GTX 980, then the difference between their performance reaches 37-42%, which is perfectly explained by theory and exactly corresponds to it. Maxwell GPUs are noticeably more efficient in mixed load mode, quickly switching from graphics to computing tasks and vice versa, and the Titan X is much faster than even the dual-GPU Radeon R9 295X2 in this test.

After analyzing the results of synthetic tests of the new Nvidia Geforce GTX Titan X video card based on the new top-end GM200 GPU, as well as considering the results of other video card models from both discrete GPU manufacturers, we can conclude that the video card under consideration should become the fastest on the market. competing with the strongest dual-GPU video card from AMD. In general, this is a good follower of the Geforce GTX Titan Black model - the most powerful single-chip.

The new video card from Nvidia shows quite strong results in synthetic - in many tests, though not in all. Radeon and Geforce traditionally have different strengths. In a large number of tests, the two graphics processors in the Radeon R9 295X2 model turned out to be faster, also due to the higher final memory bandwidth and texturing speed with very efficient execution of computational tasks. But in other cases, the top GPU of the Maxwell architecture plays out, especially in geometry tests and examples with tessellation.

However, in real gaming applications everything will be a little different, compared to the "synthetics" and the Geforce GTX Titan X should show there speed noticeably higher than the one-chip Geforce GTX 980 and even more so the Radeon R9 290X. And it is difficult to compare the new product with a two-chip Radeon R9 295X2 - systems based on two or more GPUs have their own unpleasant features, although they provide an increase in the average frame rate with proper optimization.

But the architectural features and functionality are clearly in favor of the premium Nvidia solution. Geforce GTX Titan X consumes much less power than the same Radeon R9 295X2, and in terms of energy efficiency, the new model from Nvidia is very strong - this is a distinctive feature of the Maxwell architecture. Do not forget about the greater functionality of the new Nvidia: there is support for Feature Level 12.1 in DirectX 12, hardware acceleration VXGI, a new anti-aliasing method MFAA and other technologies. We already talked about the market point of view in the first part - in the elite segment, not so much depends on the price. The main thing is that the solution is as functional and productive as possible in gaming applications. Quite simply - it was the best in everything.

Just in order to assess the speed of the new item in games, in the next part of our material we will determine the performance of the Geforce GTX Titan X in our set of gaming projects and compare it with the indicators of competitors, including assessing the reasonableness of the retail price of the new item from the point of view of enthusiasts, and we will also find out how much faster it is than the Geforce GTX 980 already in games.