<html><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8"></head><body><div>Meltdown - AMD not affected - specific to Intel products</div><div><br></div><div>Spectre - all cpus</div><div><br></div><div>https://blogs.manageengine.com/desktop-mobile/2018/01/05/meltdown-and-spectre-battling-the-bugs-in-intel-amd-and-arm-processors.html</div><div><br></div><div style="font-size:100%;color:#000000"><!-- originalMessage --><div>-------- Message d'origine --------</div><div>De : Jonathan Aquilina <jaquilina@eagleeyet.net> </div><div>Date : 05/01/2018 17:47 (GMT+01:00) </div><div>À : beowulf@beowulf.org </div><div>Objet : Re: [Beowulf] [upgrade strategy] Intel CPU design bug & security flaw - kernel fix imposes performance penalty </div><div><br></div></div>Chris on a number of articles I read they are saying AMD's are not<br>affected by this.<br><br>On 05/01/2018 17:27, Christopher Samuel wrote:<br>> On 05/01/18 10:48, Jörg Saßmannshausen wrote:<br>> <br>>> What I would like to know is: how about compensation? For me that is<br>>> the same as the VW scandal last year. We, the users, have been<br>>> deceived.<br>> <br>> I think you would be hard pressed to prove that, especially as it seems<br>> that pretty much every mainstream CPU is affected (Intel, AMD, ARM, Power).<br>> <br>>> Specially if the 30% performance loss which have been mooted are not<br>>> special corner cases but are seen often in HPC. Some of the chemistry<br>>> code I am supporting relies on disc I/O, others on InfiniBand and<br>>> again other is running entirely in memory.<br>> <br>> For RDMA based networks like IB I would suspect that the impact will be<br>> far less as the system calls to set things up will be impacted but that<br>> after that it should be less of an issue (as the whole idea of RDMA was<br>> to get the kernel out of the way as much as possible).<br>> <br>> But of course we need real benchmarks to gauge that impact.<br>> <br>> Separating out the impact of various updates will also be important,<br>> I've heard that the SLES upgrade to their microcode package includes<br>> disabling branch prediction on AMD k17 family CPUs for instance.<br>> <br>> All the best,<br>> Chris<br><br><br>_______________________________________________<br>Beowulf mailing list, Beowulf@beowulf.org sponsored by Penguin Computing<br>To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf<br></body></html>