Jump to content
HWBOT Community Forums

_mat_

Members
  • Posts

    1003
  • Joined

  • Last visited

  • Days Won

    41

Posts posted by _mat_

  1. With CUDA 7.0 NVIDIA removed the support for Geforce 200 cards. GPUPI 1.4 should still work, because it comes with CUDA 6.5 which allows to compile kernels for CUDA 2.0 devices, for example the GTX 285. I would not have introduced a new CUDA version if it wasn't the best for performance.

     

    If it's popular I can compile GPUPI 2.1 with the old CUDA toolkit especially for Geforce 200 cards. :)

     

    Btw, here is a full list of currently supported devices: https://developer.nvidia.com/cuda-gpus

  2. RV670 (HD 3850) doesn't support opencl on my system either.

     

    _mat_:

    If I understand it correctly the oldest supported GPU for the GPUPI is RV770 and GT200, right?

    Too bad the whole GeForce 8 and 9 family is out because of the lack of dual-precision.

    GeForce 8 and 9 both support the first versions of CUDA, but I guess I would have to write a whole new code path to support them. It would definitely be _very_ slow.

     

    OpenCL 1.0 specifications were release in 2008 so any cards before that have no support. That sadly includes HD 38xx cards that have been launched in November 2007. Although there is one chance to support them with the first version of AMD's Stream SDK and the Close To Metal interface. I tried that years ago without any luck. :P

     

    so this meens a lot of amd gpu's whil not work? like a 4870x2
    HD 4xxx are the first cards to introduce OpenCL 1.0. They should work if the recent Catalyst drivers still support them. With old drivers you could run into the "Invalid result" error due to their many bugs with double precision calculation.
  3. dam ther gose my plan for running 4x 3870x2.have alredy 2 cards here and 2 more on the way.

    edit; acording to http://en.wikipedia.org/wiki/List_of_AMD_graphics_processing_units#Radeon_R700_.28HD_4xxx.29_Series

    the 3870x2 support cl 1.0 and have a Double Precision of 211.2

    It could also be a driver issue. The bench only detects what the driver knows. If a device is ignored (no double precision support), you will be notified in the output window when starting GPUPI.

     

    Edit: Have a look at your screen again ... GPU-Z also claims no OpenCL support for this card.

  4. I've just finished working on GPUPI 2.1. It's now in the hands of the HWBOT staff to test it thoroughly! :)

     

    Changelog:

     

    • Fixed timer bugs in Windows 8/10 - it uses RTC for XP and 7 and QPC (+ validation of frequency) on 8/10. Additionally the used timer and frequency/precision is written to the top of the output window.
    • The data file can now be saved to disk.
    • Detects ~50 AMD graphics cards now (about 30 more than 2.0).
    • Supports UNICODE file paths for logging and configuration now
    • Fixed hardware detection of Xeon CPUs

  5. HWBOT does not know the processor you are using. Please contact the moderators to add it.

     

    I have also enabled the manual submission again, I hope that's okay with the HWBOT staff. It has to stay enabled for good because there will always be cases where the detection fails. It's just not possible with the amount of hardware out there, lots of exceptions and inconsistencies ... sometimes even the vendors have inconsisent naming schemes. I highly doubt that it's possible to avoid those detection problems.

  6. It's currently not possible, but will be very soon. Hopefully today or maybe tomorrow. It's is currently tested. :)

     

    I was not aware until today that the manual submission for the bench was disabled. I will talk to Massman about this, because it avoids submission of older AMD graphics cards like HD 58xx, which are currently only detected with their GPU codename and will therefor not be recognized by the HWBOT API.

  7. Interesting bench, tried 1,2 & 3 GTX 280s Invalid Result – switched to a 3870 X2 no go, last tried a 4890 valid result, kept locking up when setting up screen shot, same for 7 32 & 64. Went to 2D kicked me out of the submission saying I needed a cpu to submit for 2D, a bit frustrating.
    Seems like the driver is too old. Update it to the newest one and it will work.

     

    I had no problem running single HD 4850 and HD 4830 + 4850 combination (win7 x64, cat 14.9, X58 platform). At the moment it is impossible to submit score due to the CPU detection bug.
    What's your CPU?

     

    What was that?

    OpenCL, and VS2013 was installed.

    Strange error, seems like your system has some general problems. The benchmark needs to create a few (worker) threads by using _beginthreadex(). It's a very basic task for a Win32 application so this should work on a healthy installation. But have a look at GPUPI.log in the directory of the bench, it might give you more insight.
  8. Seems it can't detect my Graphic card right, it shows AMD Radeon R9 280 instead of AMD Radeon HD 7950 (because 280 and 7950 are the same)

    Can i edit my GPU name after submitting the score as "R9 280"?

    That one is on AMD. They only pass a code name to the OpenCL drivers which is obviously the same for these reused chips. I tried to divide those cards by using the reference gpu frequency of the old models and let everything else be detected as the newer ones. So if you don't have a reference card or flashed it with another BIOS you'll get an R7/R9 card for now.

     

    I might be able to distinguish these better by using WMI or a graphics API to read out the family string (HD 7900 series ...) and use it as a helper. Actually I am already using WMI for the HWBOT device detection, if it's needed. But I am not sure if the whole benchmark should be depending on having WMI installed and the service available on the system. :(

  9. Yes, Andi seems like it's only CPU power. Same goes for VGAs. Mems didn't help on my testing at least on 290X.
    It's actually not that easy. :)

    The CPU implementation depends heavily on the OpenCL drivers. You are using the Intel drivers, which aren't the best choice. Try again with AMD's OpenCL 2.0 drivers, it should be faster (I used them in this score).

     

    Yes, the bench loves raw MHz and I am glad about that (that's actually what I like most about SuperPI. :D). But that's not the whole game. Memory can be important too and can make a difference of 10 seconds or more, depending on the platform. You can optimize it by having a look at the reduction time in the statistics (last line).

  10. Even for Single GPUs if we look at the Generational increase we might reach to zero fast maybe even with the upcoming 390X

     

    Just look at the difference between 7970 & 290X at similar clockspeeds

     

    http://hwbot.org/submission/2719953_vadimua_gpupi___1b_radeon_hd_7970_35sec_410ms

    http://hwbot.org/submission/2760482_d0minat0r_gpupi___1b_radeon_r9_290x_17sec_929ms

     

    A 390X on Ln2 might just get to zero.

     

    And for multi GPU, you already saw this

    http://hwbot.org/submission/2820501_pulse88_gpupi___1b_3x_radeon_r9_290x_5sec_907ms

     

    4 x 290Xs on LN2 & its over. In the begining of the thread, people were quite ready for a longer GPU bench, don't know what changed.

    I wouldn't be too sure about that. Hawaii got a big boost in integer performance, which might have been related to the Bitcoin mining hype back then. Yes, rumours indicate more compute cores for 390X, but those seem to be designed for handling virtual reality technologies. They might cut back on integer performance and prefer floating point instead, which means more frames per second for 3d in-game graphics. It's not that straight forward, you know? So let's calm down and see what the vendors got up their sleeves for this year and adjust as needed. The HWBOT staff will take care of that, I have no doubts. :)
  11. We have discussed this recently and came to the conclusion that 32B makes no sense today. It's just not fun to run this benchmark on single cards for more than 15 minutes to get a result. Yeah, multiple GPUs have a lot of performance headroom and will push the times faster towards zero than what's possible with other benchmarks. If that happens 32B might become more popular for multiple graphics cards and we will react on it properly, may add or even shift points to 32B. But that will take a good while in computer years. :)

  12. GPUPI 2.0 is official! The benchmark now supports multiple graphics cards and CPUs and integrates automatic HWBOT submission including hardware detection and screenshot upload. HWBOT just announced it as their first GPGPU benchmark and gives world record points, globals and hardware points for GPUPI 1B and GPUPI for CPUs - 1B.

     

    To see what's possible with version 2.0 have a look at this result with 3x GTX 980 and 1x GTX 780 Ti:

     

    gpupi-2-0-quad-sli-release_202258.jpg

     

    If you have any feedback or questions, just let me know. :)

     

    Full changelog:

     

    • Multi-GPU and Multi-CPU support
    • New platform and device selection with tree view
    • HWBOT integration with automatic result submission including hardware detection (Graphics Card, CPU, OS) and screenshot upload
    • Improved detection of AMD graphics cards (HD 7xxx und Rx xxx only)
    • Encrypted kernel code using AES/CBC
    • Support for CUDA 7.0
    • Improvements for error logging (GPUPI.log in the working directory)
    • Compiled with Visual Studio 2013 Community (Update 4) and AMD OpenCL 2.0 headers
    • New about dialog including benchmark and support information

     

    Links: Download it here | Details on hwbot.org

  13. Run into that a few days ago. Be sure to insert the score correctly into the form. Especially minutes, seconds and milliseconds have to be put into their own input fields, no comma or dot allowed!

×
×
  • Create New...