1. Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

AIDA64 GPGPU Benchmark

Discussion in 'General Software' started by Aquinus, Jan 7, 2014.

  1. HammerON

    HammerON The Watchful Moderator Staff Member

    Joined:
    Mar 2, 2009
    Messages:
    6,414 (3.28/day)
    Thanks Received:
    3,494
    Here is my contribution:

    [​IMG]
    2 GPUs|Nvidia GTX 780|2|4608|24|901|12130|12033|435708|7514|364.7|1493|1493|314.7|1432|94.64

    X64 CPU|Intel Core i7 4770K|1|4|8|4400|32474|41603|35923|562.9|281.4|140.8|140.8|35.18|174.5|93.41
    Aquinus says thanks.
    Crunching for Team TPU
  2. Aquinus

    Aquinus Resident Wat-man

    Joined:
    Jan 28, 2012
    Messages:
    5,887 (6.59/day)
    Thanks Received:
    1,888
    Location:
    Concord, NH
    Added new scores.

    I pruned this one out because it was longer than the table.
    Code:
    G1620|1|2|SMCU?|2800|trial|12955|12517|trial|22.07|43.95|43.95|64|11.05|trial|11.71
    HammerON says thanks.
  3. james888

    james888

    Joined:
    Jun 27, 2011
    Messages:
    4,246 (3.83/day)
    Thanks Received:
    1,405
    Updated because I now see what SMCU is, and shortened the one in the quote.
    X64|Celeron G1620|1|2|2|2800|trial|12955|12517|trial|22.07|43.95|43.95|64|11.05|trial|11.71
    Gpu|Amd 7970|1|2048|32|1000|3272|trial|235123|4256|1060|4253|trial|202.7|625.2|trial
    Aquinus says thanks.
    Crunching for Team TPU
  4. Aquinus

    Aquinus Resident Wat-man

    Joined:
    Jan 28, 2012
    Messages:
    5,887 (6.59/day)
    Thanks Received:
    1,888
    Location:
    Concord, NH
    The Celeron line is still one element too big. There is an extra | and some data somewhere.
  5. Drone

    Drone

    Joined:
    Sep 1, 2010
    Messages:
    2,473 (1.76/day)
    Thanks Received:
    1,176
    My old laptop

    CPU:
    x64 CPU|Intel Core i7 720QM|1|4|8|2800|15308|12734|14157|55.28|27.65|27.65|27.60|13.80|26.45|12.89
    GPU:
    1 GPU|nVIDIA GeForce GT 240M|1|48|6|1210|3209|2909|20866|115.9|Not Supported|114.9|22.93|5.31|24.20|Not Supported

    [​IMG]
    Aquinus says thanks.
  6. Drone

    Drone

    Joined:
    Sep 1, 2010
    Messages:
    2,473 (1.76/day)
    Thanks Received:
    1,176
    Old desktop

    [​IMG]

    CPU:
    x64 CPU|Intel Pentium E2160|1|2|2|1800|4766|3656|3904|28.89|14.45|14.45|14.45|3.61|10.90|5.63
    GPU:
    1 GPU|nVIDIA GeForce GT 430|1|96|2|1400|408.3|327.4|21096|266.5|22.37|89.47|89.48|22.27|48.70|6.22
    Aquinus says thanks.
  7. agent00skid

    Joined:
    May 6, 2012
    Messages:
    525 (0.66/day)
    Thanks Received:
    322
    [​IMG]

    2 GPUs|AMD R7 250+7850K|2|896|14|1100/720|16981|16625|52420|1579|98.71|1579|316.1|75.21|311.8|32.18

    x64 CPU|AMD A10-7850K|1|4|4|4200|21621|11726|20375|133.8|67.03|33.51|33.51|4.15|35.55|18.27
    Aquinus and james888 say thanks.
    Crunching for Team TPU
  8. broken pixel

    broken pixel

    Joined:
    Aug 12, 2012
    Messages:
    287 (0.41/day)
    Thanks Received:
    198
    Location:
    Nebulas
    gpgpu.png
    XFX 290x 1
    Benchmark Result Run Time Build Time
    -------------------------------------------------------------
    Memory Read 11681 MB/s 813 ms
    - Pinned 11681 MB/s 3 ms
    - Pageable 5769 MB/s 5 ms
    Memory Write 10332 MB/s 516 ms
    - Pinned 10332 MB/s 3 ms
    - Pageable 5532 MB/s 5 ms
    Memory Copy 277099 MB/s 2891 ms
    - 15 MB Block 200745 MB/s 0 ms
    - 32 MB Block 243898 MB/s 0 ms
    - 64 MB Block 259735 MB/s 0 ms
    - 128 MB Block 265247 MB/s 1 ms
    - 256 MB Block 273125 MB/s 2 ms
    - 512 MB Block 277099 MB/s 4 ms
    Single-Precision FLOPS 6214 GFLOPS 8516 ms
    - float1 6213 GFLOPS 708 ms 234 ms
    - float2 6087 GFLOPS 722 ms 157 ms
    - float4 6214 GFLOPS 708 ms 125 ms
    - float8 6211 GFLOPS 708 ms 62 ms
    - float16 6209 GFLOPS 708 ms 63 ms
    Double-Precision FLOPS 776.8 GFLOPS 10093 ms
    - double1 775.7 GFLOPS 709 ms 516 ms
    - double2 745.8 GFLOPS 737 ms 500 ms
    - double4 776.8 GFLOPS 708 ms 500 ms
    - double8 776.6 GFLOPS 708 ms 500 ms
    - double16 776.3 GFLOPS 708 ms 531 ms
    24-bit Integer IOPS 6210 GIOPS 7719 ms
    - int1 6204 GIOPS 709 ms 31 ms
    - int2 6210 GIOPS 708 ms 63 ms
    - int4 6210 GIOPS 708 ms 31 ms
    - int8 6208 GIOPS 708 ms 78 ms
    - int16 6209 GIOPS 708 ms 62 ms
    32-bit Integer IOPS 2113 GIOPS 7891 ms
    - int1 2113 GIOPS 520 ms 1703 ms
    - int2 2065 GIOPS 533 ms 47 ms
    - int4 2059 GIOPS 534 ms 62 ms
    - int8 2070 GIOPS 531 ms 156 ms
    - int16 2068 GIOPS 532 ms 125 ms
    64-bit Integer IOPS 475.3 GIOPS 11110 ms
    - long1 474.8 GIOPS 579 ms 829 ms
    - long2 475.3 GIOPS 578 ms 94 ms
    - long4 470.9 GIOPS 584 ms 172 ms
    - long8 377.4 GIOPS 728 ms 360 ms
    - long16 225.9 GIOPS 608 ms 1734 ms
    Single-Precision Julia 987.1 FPS 10750 ms
    - float1 break 456.5 FPS 561 ms 469 ms
    - float1 stay / unroll 3 769.7 FPS 665 ms 63 ms
    - float2 stay / unroll 3 987.1 FPS 519 ms 78 ms
    - float4 stay / unroll 3 774.8 FPS 661 ms 31 ms
    - float1 stay / unroll 9 803.4 FPS 637 ms 62 ms
    - float2 stay / unroll 9 932.9 FPS 549 ms 32 ms
    - float4 stay / unroll 9 742.3 FPS 690 ms 94 ms
    Double-Precision Mandel 122.3 FPS 9484 ms
    - double1 break 89.3 FPS 716 ms 63 ms
    - double1 stay / unroll 3 113.8 FPS 562 ms 31 ms
    - double2 stay / unroll 3 122.3 FPS 523 ms 31 ms
    - double4 stay / unroll 3 113.7 FPS 563 ms 47 ms
    - double1 stay / unroll 9 113.9 FPS 562 ms 32 ms
    - double2 stay / unroll 9 120.6 FPS 531 ms 93 ms
    - double4 stay / unroll 9 84.5 FPS 758 ms 63 ms

    XFX 290x 2
    Benchmark Result Run Time Build Time
    -------------------------------------------------------------
    Memory Read 11686 MB/s 844 ms
    - Pinned 11686 MB/s 3 ms
    - Pageable 5969 MB/s 5 ms
    Memory Write 10201 MB/s 594 ms
    - Pinned 10201 MB/s 3 ms
    - Pageable 5514 MB/s 5 ms
    Memory Copy 276955 MB/s 3062 ms
    - 15 MB Block 202480 MB/s 0 ms
    - 32 MB Block 242419 MB/s 0 ms
    - 64 MB Block 261433 MB/s 0 ms
    - 128 MB Block 265864 MB/s 1 ms
    - 256 MB Block 273686 MB/s 2 ms
    - 512 MB Block 276955 MB/s 4 ms
    Single-Precision FLOPS 6347 GFLOPS 8500 ms
    - float1 6347 GFLOPS 693 ms 109 ms
    - float2 6347 GFLOPS 693 ms 78 ms
    - float4 6214 GFLOPS 708 ms 63 ms
    - float8 6212 GFLOPS 708 ms 125 ms
    - float16 6209 GFLOPS 708 ms 125 ms
    Double-Precision FLOPS 793.4 GFLOPS 10093 ms
    - double1 793.4 GFLOPS 693 ms 266 ms
    - double2 776.9 GFLOPS 708 ms 250 ms
    - double4 776.8 GFLOPS 708 ms 250 ms
    - double8 776.6 GFLOPS 708 ms 250 ms
    - double16 776.3 GFLOPS 708 ms 266 ms
    24-bit Integer IOPS 6343 GIOPS 7734 ms
    - int1 6343 GIOPS 693 ms 63 ms
    - int2 6343 GIOPS 693 ms 32 ms
    - int4 6210 GIOPS 708 ms 63 ms
    - int8 6209 GIOPS 708 ms 47 ms
    - int16 6210 GIOPS 708 ms 125 ms
    32-bit Integer IOPS 2113 GIOPS 7891 ms
    - int1 2113 GIOPS 520 ms 812 ms
    - int2 2109 GIOPS 521 ms 94 ms
    - int4 2059 GIOPS 534 ms 109 ms
    - int8 2070 GIOPS 531 ms 78 ms
    - int16 2068 GIOPS 532 ms 250 ms
    64-bit Integer IOPS 486.9 GIOPS 11110 ms
    - long1 486.9 GIOPS 565 ms 1657 ms
    - long2 475.3 GIOPS 578 ms 188 ms
    - long4 470.7 GIOPS 584 ms 343 ms
    - long8 377.4 GIOPS 728 ms 719 ms
    - long16 224.5 GIOPS 612 ms 890 ms
    Single-Precision Julia 991.2 FPS 10719 ms
    - float1 break 457.5 FPS 560 ms 438 ms
    - float1 stay / unroll 3 776.0 FPS 660 ms 31 ms
    - float2 stay / unroll 3 991.2 FPS 517 ms 110 ms
    - float4 stay / unroll 3 706.1 FPS 725 ms 78 ms
    - float1 stay / unroll 9 813.5 FPS 629 ms 31 ms
    - float2 stay / unroll 9 952.4 FPS 538 ms 78 ms
    - float4 stay / unroll 9 789.8 FPS 648 ms 47 ms
    Double-Precision Mandel 123.6 FPS 9469 ms
    - double1 break 89.5 FPS 715 ms 31 ms
    - double1 stay / unroll 3 113.6 FPS 563 ms 62 ms
    - double2 stay / unroll 3 121.8 FPS 525 ms 78 ms
    - double4 stay / unroll 3 112.8 FPS 567 ms 78 ms
    - double1 stay / unroll 9 114.5 FPS 559 ms 78 ms
    - double2 stay / unroll 9 123.6 FPS 518 ms 47 ms
    - double4 stay / unroll 9 84.5 FPS 758 ms 125 ms
    Last edited: Feb 5, 2014
  9. james888

    james888

    Joined:
    Jun 27, 2011
    Messages:
    4,246 (3.83/day)
    Thanks Received:
    1,405
    I have been waiting to see a keveri gpgpu benched by a reliable source. If you wouldn't mind I would love to see how just the integrated does with compute.
    Crunching for Team TPU
  10. Aquinus

    Aquinus Resident Wat-man

    Joined:
    Jan 28, 2012
    Messages:
    5,887 (6.59/day)
    Thanks Received:
    1,888
    Location:
    Concord, NH
    Updated.

    I'm not going to add it unless you format it yourself. Call me lazy, if you will.
  11. Durvelle27

    Durvelle27

    Joined:
    Jul 10, 2012
    Messages:
    2,394 (3.28/day)
    Thanks Received:
    594
    Location:
    Memphis, TN
    x64 CPU | Intel i7 4770 | 1 | 4 | 8 | 4134mhz | trial version | 36489MB/s | 33370MB/s | trial version | 262,9 GFLOPS | 131.9 GIOPS | 131.9 GIOPS | 32.88 GIOPS | Trial Version | 87.75FPS
    1 GPU | Nvidia GTX780 | 1 | 2304 | 15 | 1267mhz | 12752 MB/s | Trial Version | 245354 MB/s | 4810 GFLOPS | 233.5 GFLOPS | 955.5 GIOPS | Trial Version | 201.4 GFLOPS | 916.5FPS | Trial Version


    [​IMG]
    Aquinus says thanks.
  12. Drone

    Drone

    Joined:
    Sep 1, 2010
    Messages:
    2,473 (1.76/day)
    Thanks Received:
    1,176
    wait for it .... it's Pentium4 !!!!

    [​IMG]



    x86 CPU|Intel Pentium 4|1|1|2|2800|3123|3095|3089|11.23|5.61|3.06|3.06|0.37|312.6|4.63|2.34
    Aquinus says thanks.
  13. Aquinus

    Aquinus Resident Wat-man

    Joined:
    Jan 28, 2012
    Messages:
    5,887 (6.59/day)
    Thanks Received:
    1,888
    Location:
    Concord, NH
    Table has been updated. I dig the P4 results. Anyone have a skt478 P4 just for giggles?

    Also, the table is predominantly Intel for CPUs. For those of you with systems with AMD cpus, I would love to see how AMD's modules fair on GPGPU. My hunch is that integer performance will be amazing.

    Edit: When I say that, I mean 3 module (6 core) and 4 module (8 core) AMD cpus is what I would like to see. Kaveri looks promising considering it's only two modules.
  14. james888

    james888

    Joined:
    Jun 27, 2011
    Messages:
    4,246 (3.83/day)
    Thanks Received:
    1,405
    For a rough estimate on and gpgpu results in comparison to Intel look up how well they hash in cryptocurrencies. https://litecoin.info/Mining_hardware_comparison

    Just looking over it myself, an AMD 8350 gets ~60 k/hash and an intel 3770 gets about ~60 k/hash. Its a decent comparison. Integer compute performance.
    Last edited: Mar 6, 2014
    Crunching for Team TPU
  15. agent00skid

    Joined:
    May 6, 2012
    Messages:
    525 (0.66/day)
    Thanks Received:
    322
    Unavngivet.jpg

    x86 CPU|Intel Pentium 4|1|1|1|2000|1952|2057|1903|6.97|3.68|3.84|3.84|0.18|3.32|1.77
    Last edited: Mar 8, 2014
    Aquinus says thanks.
    Crunching for Team TPU
  16. Aquinus

    Aquinus Resident Wat-man

    Joined:
    Jan 28, 2012
    Messages:
    5,887 (6.59/day)
    Thanks Received:
    1,888
    Location:
    Concord, NH
    Thanks for the follow up, but...

  17. agent00skid

    Joined:
    May 6, 2012
    Messages:
    525 (0.66/day)
    Thanks Received:
    322
    Updated with formatting. :p

    Might get my hand on a 2,4 Celeron soon too.

    PS: Also, the other P4 has SHA1 hash in it's data too, so it's broken the table.
    Crunching for Team TPU
  18. TRINITAS

    TRINITAS

    Joined:
    Nov 5, 2012
    Messages:
    9 (0.01/day)
    Thanks Received:
    7
    My computer:

    [​IMG] [​IMG]
    [​IMG]
    [​IMG]
  19. Aquinus

    Aquinus Resident Wat-man

    Joined:
    Jan 28, 2012
    Messages:
    5,887 (6.59/day)
    Thanks Received:
    1,888
    Location:
    Concord, NH
    I'll add it when you format it. ;)
  20. james888

    james888

    Joined:
    Jun 27, 2011
    Messages:
    4,246 (3.83/day)
    Thanks Received:
    1,405
    Since this thread started, they added a SHA-1 Hash in the benchmark. Anyways, an a10-5800k might be interesting.

    x64|A10-5800k|1|4|4|4200|trial|10519|17397|trial|1.64|33.18|33.41|4.16|Trial|18.45

    GPU|A10-5800k|1|384|6|800|7072|trial|19713|607.5|38.14|602.3|trial|16.09|76.61|trial
    Crunching for Team TPU
  21. Aquinus

    Aquinus Resident Wat-man

    Joined:
    Jan 28, 2012
    Messages:
    5,887 (6.59/day)
    Thanks Received:
    1,888
    Location:
    Concord, NH
    AMD CPUs have always excelled at cryptography and I'm not sure why. I suspect AMD has optimized their CPUs for these kinds of workloads.
    Yeah... I don't really want to change the table though, lets leave it the way that it is now for the time being.
  22. BRiX

    BRiX

    Joined:
    Mar 20, 2014
    Messages:
    3 (0.03/day)
    Thanks Received:
    2
    cpuid.png gpgpu.png

    X64 CPU|Intel Core i7-3930k|1|6|12|4704.4|66361|70235|63530|445.5|222.8|222.3|222.3|55.73|186.0|98.63

    2 GPU|Nvidia GTX 780|2|4608|12|1254|23518|22229|505482|9513|472.4|1911|1911|477.1|1801|122.3
    Aquinus says thanks.
  23. itsakjt

    itsakjt

    Joined:
    Oct 9, 2010
    Messages:
    1,164 (0.85/day)
    Thanks Received:
    365
    Location:
    Kolkata, India
    [​IMG]

    GPU
    1xGPU|AMD Radeon HD 6770|1|800|16|960|5451|5328|62343|1231|Not Supported|304.4|304.5|61.30|170.7|Not Supported

    CPU
    x64 CPU|AMD Phenom II X4 955BE|1|4|4|3740|20880|11178|17634|119.6|59.80|38.28|38.28|14.44|47.57|24.23

    Don't know how I almost missed this thread. Good work. :)
    Aquinus says thanks.
  24. Aquinus

    Aquinus Resident Wat-man

    Joined:
    Jan 28, 2012
    Messages:
    5,887 (6.59/day)
    Thanks Received:
    1,888
    Location:
    Concord, NH
    @itsakjt You know, threads like these like to hide and then they just loose interest. A shame. Maybe I'll start doing some graphical stuff with it soon to entice more people to contribute.

    BTW, the table has been updated. Thanks! :cool:
    itsakjt says thanks.
  25. agent00skid

    Joined:
    May 6, 2012
    Messages:
    525 (0.66/day)
    Thanks Received:
    322
    Guess I can throw my laptop in as well. With a picture bonus of the GPU at 300/300 MHz because I forgot to lift the performance limit. :p

    1xGPU|AMD Radeon Mobility HD 5650|1|400|5|550|3079|2653|21178|349.1|Not Supported|86.65|86.65|17.47|60.95|Not Supported
    x64 CPU|AMD Phenom II X3 N830|1|3|3|2100|13173|6709|10809|50.24|25.12|16.08|16.08|6.06|19.99|10.18

    Edit: What's up with some of the columns sorting correctly and some of them not?
    Also, the Q9550 entry is missing a x64 CPU in front.

    Edit2: Derp me. It's probably the text in the columns that make them go from number order to character order.

    Edit3: Now that I'm going so well. :p One of the GTX 780 entry's have the total core count in the wrong place, switched with what should be CU count, but is incorrect.

    Attached Files:

    Last edited: Mar 24, 2014
    Aquinus says thanks.
    Crunching for Team TPU

Currently Active Users Viewing This Thread: 1 (0 members and 1 guest)

Share This Page