Compute Performance

Shifting gears, we'll look at the compute aspects of the Radeon VII. Though it is fundamentally similar to first generation Vega, there has been an emphasis on improved compute for Vega 20, and we may see it here.

Beginning with CompuBench 2.0, the latest iteration of Kishonti's GPU compute benchmark suite offers a wide array of different practical compute workloads, and we’ve decided to focus on level set segmentation, optical flow modeling, and N-Body physics simulations.

Compute: CompuBench 2.0 - Level Set Segmentation 256

Compute: CompuBench 2.0 - N-Body Simulation 1024K

Compute: CompuBench 2.0 - Optical Flow

Moving on, we'll also look at single precision floating point performance with FAHBench, the official Folding @ Home benchmark. Folding @ Home is the popular Stanford-backed research and distributed computing initiative that has work distributed to millions of volunteer computers over the internet, each of which is responsible for a tiny slice of a protein folding simulation. FAHBench can test both single precision and double precision floating point performance, with single precision being the most useful metric for most consumer cards due to their low double precision performance.

Compute: Folding @ Home (Single and Double Precision)

Next is Geekbench 4's GPU compute suite. A multi-faceted test suite, Geekbench 4 runs seven different GPU sub-tests, ranging from face detection to FFTs, and then averages out their scores via their geometric mean. As a result Geekbench 4 isn't testing any one workload, but rather is an average of many different basic workloads.

Compute: Geekbench 4 - GPU Compute - Total Score

Lastly, we have SiSoftware Sandra, with general compute benchmarks at different precisions.

Compute: SiSoftware Sandra 2018 - GP Processing (OpenCL)

Compute: SiSoftware Sandra 2018 - GP Processing (DX11)

Compute: SiSoftware Sandra 2018 - Pixel Shader Compute (DX11)

 

Total War: Warhammer II Synthetics
Comments Locked

289 Comments

View All Comments

  • eva02langley - Thursday, February 7, 2019 - link

    Love my 2400g, agree with an iGPU like that, however you will not make me believe that an HD 520 is "enough".

    My only GPUs that died were Nvidia ones... 3 in total. 0 AMD.
  • TheinsanegamerN - Thursday, February 7, 2019 - link

    Read what he said again. "For majority 90+ % of people Integrated graphics are good enough for spreadsheets, internet and word processing"

    Guess what nearly every office laptop and desktop uses? If it wasnt "good enough" there would be a push for more powerful iGPUs in widespread circulation.

    The basic intel iGPU if far mroe then enough to do office work, stream video, or normal workstation content. more powerful GPUs are only needed in specific circumstances.
  • eva02langley - Friday, February 8, 2019 - link

    I know what he said and I know what he means... and I know he is painted Intel all over his body.

    And no, basic HD 520 is not enough, period. You can barely do office work and play videos.

    If it was so true, the mobile market would not be the most lucrative for games/entertainment. As of now, a smart phone is having more GPU power than an HD520.

    So basically, I am not agreeing at all.
  • AdhesiveTeflon - Friday, February 8, 2019 - link

    We do plenty of CAD work and GIS functions with Intel's iGPUs, so what were you saying about them barely able to do office work and videos?

    We also have a lot more issues with AMD's mobile and professional cards than nVidia's.
  • Icehawk - Saturday, February 9, 2019 - link

    Out of 500 PCs at my job a whopping two have video cards, they are random low ends ones purchased to add additional video ports for two stations that run quad monitors. Otherwise there is zero need for a dGPU.

    This is typical of the vast majority of businesses.
  • ksec - Thursday, February 7, 2019 - link

    1. I believe there will be better drivers for VII, it was quite clear that there are many optimisation not done in time, although I don't know how long it will take. The new AMD seems to be quick to react though.

    2. What if AMD decided to release the MI60 VII at $899.
  • TheinsanegamerN - Thursday, February 7, 2019 - link

    VII is the VEGA arch, with more ROPs. If AMD managed to leave that much performance on the table, they must be the most incompetent code writers in all of existence.

    The VEGA arch has long been optimized for, adding some ROPs isnt going to require much work to optimize for, and AMD has likely already done that.
  • ksec - Friday, February 8, 2019 - link

    Optimisation are now nearly done on a per AAA game level. And more importantly not only the drivers but the game itself. Whether the developer are willing to optimise the game ( at the help of AMD ) will be another story.l
  • crotach - Thursday, February 7, 2019 - link

    Nice to see AMD being competitive again. It's a pity they've priced the card so high in Europe that you can get a RTX 2080 for 100 euros less. At that price point they won't be selling many.
  • Manch - Thursday, February 7, 2019 - link

    Need to see the VAT free price.

Log in

Don't have an account? Sign up now