GPU numerical compliance

Article: Predicting AMD and Nvidia GPU Performance
By: Vincent Diepeveen (diep.delete@this.xs4all.nl), April 20, 2011 3:17 am
Room: Moderated Discussions
Sylvain Collange (full.name@gmail.com) on 4/19/11 wrote:
---------------------------
>>Ian Ameline (ian.ameline@(NoSpam).gmail.com) on 4/18/11 wrote:
>>---------------------------
>>>*Fully* compliant? As in <= 1/2 ULP precision guarantees for add, subtract, divide, multiply *and* sqrt?
>
>Yes. Unlike x86-SSE for instance, which is not compliant with the current IEEE-754
>standard. Not until we get the FMA operator...
>
>>>Denormal support?
>
>Implemented in hardware at full speed. Unlike most CPUs, which trap on subnormals
>and handle them in microcode at a much lower speed.
>
>>>+-0 and +- infinity correctly generated and handled?
>
>These ones were already supported long before GPGPU APIs came out. At least since
>GeForce 6 / Radeon X1x00, and likely earlier.
>
>Ping-Che Chen (pcchen@kimicat.com) on 4/19/11 wrote:
>---------------------------
>>Right now, OpenCL 1.1 requires basically the following:
>>
>>1. round to nearest even, other rounding modes are optional. Dynamic rounding mode selection is not supported.
>>2. INF and (non signaling) NaN must be supported.
>>3. Denormalized number is optional.
>>4. No exception handling is supported right now.
>>5. +, -, * must be infinitely precise. Division and sqrt are not required to be infinitely precise.
>
>For double precision, it also requires FMA, subnormals, all 4 mandatory IEEE rounding
>modes, and correctly rounded division and sqrt.
>
>Basically, current-generation GPUs are already far ahead of x86 and most other
>CPU architectures in terms of numerical compliance.
>Power6/7 and Itanium are still arguably one notch ahead of GPUs, but that is about it.

gpu's are so much faster in 32 bits code than anything else; the focus is too much upon floating point in HPC, if you ask me. Every transform you can do more accurately in integers than in floating point; in floating point you always have the problem of erros getting backtracked, no such issues with integer transforms.

Vincent
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
New Article: Predicting GPU Performance for AMD and NvidiaDavid Kanter2011/04/12 12:55 AM
  Graph is not red-green colorblind friendly (NT)RatherNotSay2011/04/12 04:51 AM
    FixedDavid Kanter2011/04/12 09:46 AM
  New Article: Predicting GPU Performance for AMD and NvidiaJames2011/04/12 01:30 PM
    New Article: Predicting GPU Performance for AMD and NvidiaDavid Kanter2011/04/12 03:51 PM
  Try HD6450 or HD6850EduardoS2011/04/12 04:31 PM
    Try HD6450 or HD6850David Kanter2011/04/13 11:25 AM
      Try HD6450 or HD6850EduardoS2011/04/13 04:20 PM
        of causeMoritz2011/04/14 09:03 AM
          of causeEduardoS2011/04/14 02:55 PM
            Barts = 5DMoritz2011/04/14 10:26 PM
              Barts = 5DAntti-Ville Tuunainen2011/04/15 01:38 AM
                Limiting fixed function unitsMoritz2011/04/15 05:28 AM
                  Limiting fixed function unitsVincent Diepeveen2011/04/20 03:38 AM
                    lack of detailMoritz2011/04/20 10:24 AM
                      lack of detailEduardoS2011/04/20 12:45 PM
            gpgpuVincent Diepeveen2011/04/16 03:10 AM
              gpgpuEduardoS2011/04/17 01:31 PM
                gpgpuGroo2011/04/17 01:58 PM
                  gpgpuEduardoS2011/04/17 02:08 PM
                  gpgpuIan Ameline2011/04/18 04:55 PM
                    gpgpuPing-Che Chen2011/04/19 01:59 AM
                      GPU numerical complianceSylvain Collange2011/04/19 12:38 PM
                        GPU numerical complianceVincent Diepeveen2011/04/20 03:17 AM
                gpgpuVincent Diepeveen2011/04/20 03:02 AM
                  gpgpu and core countsHeikki Kultala2011/04/20 05:41 AM
                    gpgpu and core countsVincent Diepeveen2011/04/20 06:52 AM
                      gpgpu and core countsnone2011/04/20 08:05 AM
                        gpgpu and core countsEduardoS2011/04/20 12:36 PM
                      gpgpu and core countsHeikki Kultala2011/04/20 11:16 AM
                        gpgpu and core countsEduardoS2011/04/20 12:34 PM
                          gpgpu and core countsHeikki Kultala2011/04/20 08:24 PM
                            gpgpu and core countsEduardoS2011/04/20 09:55 PM
                              gpgpu and core countsHeikki Kultala2011/04/21 07:48 AM
                                gpgpu and core countsEduardoS2011/04/22 02:41 PM
                              AMD Compute and Texture FetchDavid Kanter2011/04/21 11:42 AM
                                AMD Compute and Texture FetchVincent Diepeveen2011/04/22 02:14 AM
                                  AMD Compute and Texture FetchDavid Kanter2011/04/22 11:53 AM
                                AMD Compute and Texture FetchEduardoS2011/04/22 02:46 PM
                                  AMD Compute and Texture FetchDavid Kanter2011/04/22 03:02 PM
                                    AMD Compute and Texture FetchEduardoS2011/04/22 03:18 PM
                                    AMD Compute and Texture Fetchanon2011/04/22 04:30 PM
                                      AMD Compute and Texture FetchDavid Kanter2011/04/22 10:17 PM
                        gpgpu and core countsVincent Diepeveen2011/04/20 01:12 PM
                          gpgpu and core countsHeikki Kultala2011/04/21 11:23 AM
                            gpgpu and core countsVincent Diepeveen2011/04/22 03:11 AM
                              Keep the crazy politics out of thisDavid Kanter2011/04/22 09:39 AM
                                Keep the crazy politics out of thisVincent Diepeveen2011/04/22 10:12 AM
                                  Keep the crazy politics out of thisDavid Kanter2011/04/22 11:44 AM
                              gpgpu and core countsJouni Osmala2011/04/22 12:06 PM
                  gpgpuEduardoS2011/04/20 12:59 PM
                    gpgpuVincent Diepeveen2011/04/20 01:37 PM
                      gpgpuEduardoS2011/04/20 06:27 PM
                        gpgpuVincent Diepeveen2011/04/21 03:06 AM
                          gpgpuEduardoS2011/04/22 03:00 PM
  New Article: Predicting GPU Performance for AMD and NvidiaPiedPiper2011/04/12 11:05 PM
    New Article: Predicting GPU Performance for AMD and NvidiaDavid Kanter2011/04/12 11:42 PM
      New Article: Predicting GPU Performance for AMD and NvidiaMS2011/04/15 06:04 AM
        New Article: Predicting GPU Performance for AMD and NvidiaKevin G2011/04/16 03:25 AM
          New Article: Predicting GPU Performance for AMD and NvidiaDavid Kanter2011/04/16 09:42 AM
          New Article: Predicting GPU Performance for AMD and NvidiaVincent Diepeveen2011/04/20 03:20 AM
    memoryMoritz2011/04/14 10:03 PM
      memory - moreMoritz2011/04/16 12:11 AM
  New Article: Predicting GPU Performance for AMD and NvidiaKevin G2011/04/14 12:30 PM
Reply to this Topic
Name:
Email:
Topic:
Body: No Text
How do you spell avocado?