tag:support.primatelabs.com,2011-01-31:/discussions/geekbench/23607-linpack-and-hpcg-benchmarkingPrimate Labs: Discussion 2023-09-09T21:33:31Ztag:support.primatelabs.com,2011-01-31:Comment/445003392018-01-15T18:25:36Z2018-01-15T18:25:36ZLINPACK and HPCG benchmarking<div><p>Hi,</p>
<p>Thanks for your suggestion! I'll pass your suggestions along to the rest of my team for more consideration, and see if we can facilitate these kinds of comparisons in the future.</p>
<p>All the best,<br>
Colin<br>
Primate Labs Inc.</p></div>Colintag:support.primatelabs.com,2011-01-31:Comment/445003392018-01-15T18:33:20Z2018-01-15T18:33:20ZLINPACK and HPCG benchmarking<div><p>Thanks</p>
<p>I’d really like to see how well the new iMac Pro runs at least LINPACK and whether HPCG can be made to run on Radeon GPUs. Not everyone needs a supercomputer but having the capability to act like a small one might be beneficial for colleges and small businesses who would rather run locally than off someone’s (insecure) cloud platform.</p>
<p>Peter</p></div>rob53tag:support.primatelabs.com,2011-01-31:Comment/445003392019-07-17T20:43:48Z2019-07-17T20:43:49ZLINPACK and HPCG benchmarking<div><p>Hi Peter and Colin<br>
Thanks for the question, which I would also like an updated answer to.<br>
Following the TOP500 list change over time I have long felt the need to compare common consumer computers to the capabilities of TOP500 computers.</p>
<p>The TOP500 methodology is available in this document:<br>
<a href="http://www.netlib.org/benchmark/performance.ps">http://www.netlib.org/benchmark/performance.ps</a><br>
This document covers even very old computer systems.</p>
<p>News of intention and possibly progress?<br>
Michael</p></div>michael.osterwaldlenumtag:support.primatelabs.com,2011-01-31:Comment/445003392019-07-29T17:48:15Z2019-07-29T17:48:15ZLINPACK and HPCG benchmarking<div><p>Is there any update from Primate Lab staff on your internal conversation?</p>
<p>Thanks</p></div>rob53tag:support.primatelabs.com,2011-01-31:Comment/445003392019-09-03T13:27:40Z2019-09-03T13:27:40ZLINPACK and HPCG benchmarking<div><p>Just saw Geekbench 5 has been released but nothing about additional benchmarking for LINPACK or HPCG. That's unfortunate. I hope you are still looking into this capability.</p></div>rob53tag:support.primatelabs.com,2011-01-31:Comment/445003392020-01-03T04:33:59Z2020-01-03T04:33:59ZLINPACK and HPCG benchmarking<div><p>We looked at LINPACK and HPCG when designing Geekbench 5 and decided not to include them as both represent HPC workloads whereas Geekbench focuses on client workloads.</p></div>Johntag:support.primatelabs.com,2011-01-31:Comment/445003392020-06-30T19:57:33Z2020-06-30T19:57:33ZLINPACK and HPCG benchmarking<div><p>Unfortunate. Macs are more than clients they are standalone workstations. Sure, the results won't come close to anything in the TOP500 but I know a lot of scientists who use Macs and having the ability to run code normally used on supercomputers would be nice. After Apple finally admitted to the roadmap to Apple Silicon SoCs using ARM and with the newest TOP500 supercomputer being built using---<br>
"The new top system, Fugaku, turned in a High Performance Linpack (HPL) result of 415.5 petaflops, besting the now second-place Summit system by a factor of 2.8x. Fugaku, is powered by Fujitsu’s 48-core A64FX SoC, becoming the first number one system on the list to be powered by ARM processors." ---it just makes sense to begin thinking about a "client" implementation of LINPACK to see how close a single Mac workstation comes. Of course, at the moment only the Mac Pro could come anywhere near but the possibility of using cluster software on Macs might have a resurgence after seeing what the Japanese have done.</p></div>rob53