guru3dDoom Vulkan Benchmarks

[Hitman928]

But the minimum is a fair bit higher so to me its still the faster card with all things considered.

Those aren't minimums, those are OpenGL results.

 

[Unreal123]

Look at the OpenGL and Vulkan numbers. At 1920X1080,the GTX1060 is still slower than a RX480,but only shows a 3% increase over OpenGL using Vulkan. At 2560X1440,it moves a bit ahead but the Vulkan and OpenGL numbers are exactly the same. The Fury X is still faster than a GTX1070.

No Guru3d has done Doom Benchmark with the old driver (389.05) and still GTX 1070 is on par with Fury X. The new driver gains 10-15% performance for Nvidia on Vulkan ,which you add in this benchmark.

http://www.guru3d.com/articles_pages/msi_geforce_gtx_1070_gaming_z_review,17.html

http://forums.guru3d.com/showthread.php?t=409228&page=6

 

[Hitman928]

Here are my results. Latest nvidia driver definitely fixed Vulkan synch issue so now Vulkan is the preferred API for both cards. I'll have to do 1440p at some point as well.

 

[Unreal123]

Here are my results. Latest nvidia driver definitely fixed Vulkan synch issue so now Vulkan is the preferred API for both cards. I'll have to do 1440p at some point as well.

Can you old driver performance vs new drivers on Nvidia cards?

 

[Hitman928]

I'll see if I can find some time, I have a ton of stuff to do that's higher priority but I'll try to squeeze it in.

 

[Bacon1]

No Guru3d has done Doom Benchmark with the old driver (389.05) and still GTX 1070 is on par with Fury X. The new driver gains 10-15% performance for Nvidia on Vulkan ,which you add in this benchmark.

http://www.guru3d.com/articles_pages/msi_geforce_gtx_1070_gaming_z_review,17.html

http://forums.guru3d.com/showthread.php?t=409228&page=6

What is the point of the forums link you posted? Which post are you referring to? The guy with a CPU bound 1080?

You can't just say that the new drivers give 10-15% without showing that happen in an actual benchmark. That review from guru 3d might be with the newer driver, unless they purposefully used a month old driver during their review instead of the one from 2 weeks prior to the review. Considering how close the performance was in that review, I don't see how anyone would be complaining.

 

[Bacon1]

sweclockers is using 372.54

Those might be the ones that increased it then, I don't have a nvidia card and sadly no one saying it was impossible for drivers to fix performance is willing to test it out and admit they were wrong. Maybe some other sites will test it out to see what drivers it was.