
<html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;" class=""><div class=""><div><br class=""><blockquote type="cite" class=""><div class="">On 1 Sep 2021, at 14:37, 陈克 <<a href="mailto:joykechen@163.com" class="">joykechen@163.com</a>> wrote:</div><br class="Apple-interchange-newline"><div class=""><div style="line-height: 1.7; font-size: 14px; font-family: Calibri;" class=""><div style="margin: 0;" class="">Hi, </div><div style="margin: 0;" class=""></div><div style="margin: 0;" class="">I recommend a tool to test GPU performance, which depends on CUDA installation.</div><div style="margin: 0;" class=""><br class=""></div><div style="margin: 0;" class="">link: <a href="https://github.com/GPUburn/gpuburn/tree/master/GPUBurn" class="">https://github.com/GPUburn/gpuburn/tree/master/GPUBurn</a></div><div style="margin: 0px;" class=""><br class=""></div></div></div></blockquote><div><br class=""></div>gpuburn is also the tool that CERN uses also when new hardware is delivered and we want to check the hardware is working under load.<div class=""><br class=""></div><div class="">It’s also a good way to stress your PDUs and cooling if you have several cards (see <a href="https://indico.cern.ch/event/950196/contributions/3993259/attachments/2113752/3556174/GPUs_in_CERN_IT.pdf" class="">https://indico.cern.ch/event/950196/contributions/3993259/attachments/2113752/3556174/GPUs_in_CERN_IT.pdf</a>)</div><div class=""><br class=""></div><div class="">We also have some specific High Energy Physics benchmarks - <a href="https://gitlab.cern.ch/hep-benchmarks/hep-workloads-gpu" class="">https://gitlab.cern.ch/hep-benchmarks/hep-workloads-gpu</a></div><div class=""><br class=""></div><div class="">Cheers</div><div class="">Tim</div><div class=""><br style="caret-color: rgb(0, 0, 0); color: rgb(0, 0, 0);" class=""></div><blockquote type="cite" class=""><div class=""><div style="line-height: 1.7; font-size: 14px; font-family: Calibri;" class=""><div style="margin: 0px;" class=""><br class=""></div><div style="margin: 0;" class="">Thanks,</div><div style="margin: 0;" class="">Ke Chen</div><div style="margin: 0px;" class=""><br class=""></div><div style="position:relative;zoom:1" class=""></div><div id="divNeteaseMailCard" class=""></div><div style="margin: 0px;" class=""><br class=""></div><p class="">在 2021-08-31 23:10:03，"Sylvain Bauza" <<a href="mailto:sbauza@redhat.com" class="">sbauza@redhat.com</a>> 写道：</p><blockquote id="isReplyContent" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid" class=""><div dir="ltr" class=""><div dir="ltr" class=""><br class=""></div><br class=""><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Tue, Aug 31, 2021 at 4:34 PM Ildiko Vancsa <<a href="mailto:ildiko.vancsa@gmail.com" class="">ildiko.vancsa@gmail.com</a>> wrote:<br class=""></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Hi,<br class="">

<br class="">

As we are approaching the end of the holiday season I wanted to surface back my question about GPU performance testing. Does anyone have any hints to find the best tools to do some benchmarking with?<br class="">

<br class=""></blockquote><div class=""><br class=""></div><div class=""><br class=""></div><div class="">You made the point, Ildiko, I was on a long-running time-off so I didn't had time to look at your question yet.</div><div class=""><br class=""></div><div class="">Good concern tho, I have no knowledge about this, but I can ping a few other folks to get you an answer.</div><div class="">-Sylvain</div><div class=""><br class=""></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">

Thanks,<br class="">

Ildikó<br class="">

<br class="">

<br class="">

> On Aug 9, 2021, at 08:46, Ildiko Vancsa <<a href="mailto:ildiko.vancsa@gmail.com" target="_blank" class="">ildiko.vancsa@gmail.com</a>> wrote:<br class="">

> <br class="">

> Hi,<br class="">

> <br class="">

> I got a question about tools and practices to check GPU performance in an OpenStack environment that I need some help to answer.<br class="">

> <br class="">

> The question is about recommended GPU performance testing/benchmarking tools if there are a few that people in the community are using and would recommend? The scope of the testing work is to check GPU performance in OpenStack VMs (both virtualized and passthrough).<br class="">

> <br class="">

> All the help and pointers are very much appreciated!<br class="">

> <br class="">

> Thanks,<br class="">

> Ildikó<br class="">

> <br class="">

> <br class="">

<br class="">

<br class="">

</blockquote></div></div>

</blockquote></div><br class=""><br class=""><span title="neteasefooter" class=""><div class=""> <br class="webkit-block-placeholder"></div></span></div></blockquote></div><br class=""></div></body></html>