Hopefully they can get Timespy onto Vulkan as well.

Or Firestrike. 😉 For now its just a another api overhead test.

DirectX 11 single-thread 1 278 132 Draw calls per second DirectX 11 multi-thread 1 868 381 Draw calls per second DirectX 12 27 583 473 Draw calls per second Vulkan 29 721 036 Draw calls per second

DirectX 11 single-thread 1 278 132 Draw calls per second DirectX 11 multi-thread 1 868 381 Draw calls per second DirectX 12 27 583 473 Draw calls per second Vulkan 29 721 036 Draw calls per second

That's interesting, we both have GTX 1070's and our DX12 & Vulkan scores are almost an exact match, yet the DX11 multi-thread is miles different, must be the different CPUs we have, maybe this shows that DX12 & Vulkan can utilise your many cores, whereas DX11 multi cannot - thinking about it, that's probably not news!: DX11 mult:i 4,507,018 DX12: 27,231,630 Vulkan: 29,736,051

I started asking a few question about some of these descrepancies and got directed to the updated 3dMark Technical Guide. There is a lot more details on each test and all in this guide. I suggest taking a look at it to really see some eye opening things that we have been comparing but yet futuremark states we cannot compare using their tests. Link: http://s3.amazonaws.com/download-aws.futuremark.com/3DMark_Technical_Guide.pdf Specifically pages 112 and 113 to answer my questions on the API testing. Correct use of the API Overhead feature test The API Overhead feature test is not a general-purpose GPU benchmark, and it should not be used to compare graphics cards from different vendors. The test is designed to make API overhead the performance bottleneck. It does this by maximizing the number of draw calls in a scene, (by drawing a huge number of individual ‘buildings’), while minimizing the GPU load, (by using simple shaders and no lighting effects). This an artificial scenario that is unlikely to be found in games, which typically aim to achieve high levels of detail and exceptional visual quality. The benefit of reducing API overhead is greatest when the CPU is the limiting factor. With modern APIs and fast CPUs, the test can become GPU bound, but not always in a way that is meaningful from a general GPU performance perspective. The point at which the test moves from being CPU-bound to GPUbound changes from system to system. It is not easy to tell from the test results whether the run was CPU or GPU limited. And what's more, it is difficult to isolate the relative impact of GPU performance and driver performance. As a result, you should be careful making conclusions about GPU performance when comparing API Overhead test results from different systems. For instance, we would advise against comparing the Vulkan score from an AMD GPU with the DirectX 12 score from an NVIDIA GPU. Likewise, it could be misleading to credit the GPU for any difference in DirectX 12 performance between an AMD GPU and an NVIDIA GPU. Another scenario, for example, would be to test DirectX 12 performance with a range of CPUs in a system with a fixed GPU. Or, you could test a vendor's range of GPUs, from budget to high-end, and keep the CPU fixed. But in both cases, the nature of the test means it will not show you the extent to which the performance differences are due to the hardware and how much is down to the driver. The proper use of the test is to compare the relative performance of each API on a single system, rather than the absolute performance of different systems. The focus on single-system testing is one reason why the API Overhead test is called a feature test rather than a benchmark.

Hope they will update the benchmark for upcoming changes in the Windows SDK too. I would like to see if current hardware/driver could benefit for things like priority on command buffer queues.. My guess is current hardware are far from benefit a lot on different scheduling priorities..

I started asking a few question about some of these descrepancies and got directed to the updated 3dMark Technical Guide. There is a lot more details on each test and all in this guide. I suggest taking a look at it to really see some eye opening things that we have been comparing but yet futuremark states we cannot compare using their tests. Link: http://s3.amazonaws.com/download-aws.futuremark.com/3DMark_Technical_Guide.pdf Specifically pages 112 and 113 to answer my questions on the API testing. Correct use of the API Overhead feature test The API Overhead feature test is not a general-purpose GPU benchmark, and it should not be used to compare graphics cards from different vendors. The test is designed to make API overhead the performance bottleneck. It does this by maximizing the number of draw calls in a scene, (by drawing a huge number of individual ‘buildings’), while minimizing the GPU load, (by using simple shaders and no lighting effects). This an artificial scenario that is unlikely to be found in games, which typically aim to achieve high levels of detail and exceptional visual quality. The benefit of reducing API overhead is greatest when the CPU is the limiting factor. With modern APIs and fast CPUs, the test can become GPU bound, but not always in a way that is meaningful from a general GPU performance perspective. The point at which the test moves from being CPU-bound to GPUbound changes from system to system. It is not easy to tell from the test results whether the run was CPU or GPU limited. And what's more, it is difficult to isolate the relative impact of GPU performance and driver performance. As a result, you should be careful making conclusions about GPU performance when comparing API Overhead test results from different systems. For instance, we would advise against comparing the Vulkan score from an AMD GPU with the DirectX 12 score from an NVIDIA GPU. Likewise, it could be misleading to credit the GPU for any difference in DirectX 12 performance between an AMD GPU and an NVIDIA GPU. Another scenario, for example, would be to test DirectX 12 performance with a range of CPUs in a system with a fixed GPU. Or, you could test a vendor's range of GPUs, from budget to high-end, and keep the CPU fixed. But in both cases, the nature of the test means it will not show you the extent to which the performance differences are due to the hardware and how much is down to the driver. The proper use of the test is to compare the relative performance of each API on a single system, rather than the absolute performance of different systems. The focus on single-system testing is one reason why the API Overhead test is called a feature test rather than a benchmark.

Good points, I noticed that in small sections of the benchmarks that my GPU hit 100% load - towards the end of the tests when there are shed loads of draw calls on the screen. It was at those points that my GPU was the bottleneck, so I can see why more powerful GPUs can increase the scores slightly, and it's probably why I have very similar scores to the guy I replied to earlier who has a GTX 1070 - we probably became GPU limited at the same point in the test because both of our CPUs were able to deliver enough draw calls - mine through high CPU clocks and 8 threads & the other guy by having a bazzillion cores on his server CPU (GeniusPr0 and his BWE Xeon 18C/36T 2.745Ghz).

Good points, I noticed that in small sections of the benchmarks that my GPU hit 100% load - towards the end of the tests when there are shed loads of draw calls on the screen. It was at those points that my GPU was the bottleneck, so I can see why more powerful GPUs can increase the scores slightly, and it's probably why I have very similar scores to the guy I replied to earlier who has a GTX 1070 - we probably became GPU limited at the same point in the test because both of our CPUs were able to deliver enough draw calls - mine through high CPU clocks and 8 threads & the other guy by having a bazzillion cores on his server CPU (GeniusPr0 and his BWE Xeon 18C/36T 2.745Ghz).

While it answered my questions on the DX11 from my HD 7770 and RX 480 it did not explain very well reviewers much lower AMD scores across the board then real people are getting out in the real world. My question to futuremark was very specific on on the DX 11 results alone...as to reviewers results was to ask the reviewer,

for me, DX12 was consistently hitting just over 17,000,000 draw-calls per second. (most 17,220,483) Vulcan was consistently hitting just under 16,000,000 draw-calls per second. (most 15,985,491)

Is that gtx780 and how far is it overclocked? This is the best i could do with 280X. I dont think this translates to real world performance. Im positive my card outperformed yours in Doom using Vulkan according to the web benchmarks and my experience. DX12 - 14 073 975 Vulkan - 13 710 626 [spoiler]https://s28.postimg.org/hipq6aki5/fdgfdgd.png[/spoiler]

I got ~17mill with my old gtx780 @ 1150MHz boost, http://www.3dmark.com/compare/aot/198728/aot/192915/aot/62970/aot/46129 last result was best off dx11 ST, now latest 378.92 driver went a bit further 🤓