Date: Sat, 4 Jan 2025 18:53:02 +0100 From: "Peter 'PMc' Much" <pmc@citylink.dinoex.sub.org> To: Joe Schaefer <joesuf4@gmail.com> Cc: Chris Torek <chris.torek@gmail.com>, freebsd-hackers@freebsd.org Subject: Re: curious crashes when under memory pressure Message-ID: <Z3l1fnLRmxyNFQI4@disp.intra.daemon.contact> In-Reply-To: <CAOzHqcJ0=So7oeQYnMBjBCeGnbxVeVx_y%2B9k2D7wE7bvD-kSbQ@mail.gmail.com> References: <CAPx1GvfwqFqvMm0RkTvxa4BJStvOdG0=pNx2dCuWEzCXsqhqHw@mail.gmail.com> <slrnvnij2h.26bu.pmc@disp.intra.daemon.contact> <CAPx1GvfOh4g1AeOcYXC26W5jbPAgA_E_t7CW9chg4miC8c15-w@mail.gmail.com> <CAOzHqcJ0=So7oeQYnMBjBCeGnbxVeVx_y%2B9k2D7wE7bvD-kSbQ@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Jan 04, 2025 at 12:27:36PM -0500, Joe Schaefer wrote: ! Intel and AMD cores are now heterogeneous compute engines. The idea that it ! is useful to beat the crap of every core simultaneously for an extended ! period of time is a really dumb idea for capacity testing now. Hm. Not sure what exactly You mean. Also not sure /since when/ this would apply, or to /which series/ of these cores. But I would think they are indeed for beating the crap of every core simulatneously, over days. And in my perception the main issue with this is where to move the thermal. But then also, I am using Xeon EP (and rather old ones, for that) - and if these are not designed to do exactly that, then I really don't know... BTW, I am missing the 2025Q1 - they tend to get later and later. And that is an issue here, because rebuilding yields me some 15 kWh thermal in my hall, and I need some planning to cater for these. cheerio, PMc
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Z3l1fnLRmxyNFQI4>