Past the 10 Petaflop scale

[PcPerf bot]

<p>Folding@home has historically been an extremely powerful computing resource, hitting many major computing milestones. For example, we were awarded a <strong><a href="http://www.engadget.com/2007/10/31/folding-home-recognized-by-guinness-world-records/" target="_self">Guinness World Record</a></strong> for being the first to reach a petaflop (10^15 floating point operations per second).</p>

<p>We keep track of the power of Folding@home on our <strong><a href="http://fah-web.stanford.edu/cgi-bin/main.py?qtype=osstats2" target="_self">osstats page</a></strong>. Recently, we found an issue with our reporting. In particular, we were investigating a major under-reporting (OSX stats reported that there were no OSX machines at all). A few months ago, Christian Schwantes (FAH Team member) rewrote our main stats scripts and he fixed a very old bug in the stats reporting. That bug had a workaround in the osstats2 page, so now the workaround was failing and that led to finally getting osstats2 working. <br /><br />After checking the db with the raw files from the work servers (it's an important check that we can know what the right answer is from this more laborious manual process of going through the WS logs), this result is now correct, fixing the previous under-reporting. We also now are keeping better track of the Fermi (and later) class of NV GPUs. We were under-reporting their FLOP count and that alone was what make a huge difference in the FLOPs reported.</p>

<p>The upshot of all of this is that we now are no longer under-reporting our stats and the reported FLOPs has gone up significantly, to about 12 PFLOPs. This puts Folding@home past the 10 PetaFLOP scale, which is itself a very significant amount of computer power.</p>

<p>For more information, please see our <a href="http://folding.stanford.edu/English/FAQ-flops">FLOPS FAQ</a>, whichis useful in discussing this topic, which is surprisingly complicated.</p>

 

Voir l'article complet