Date: Mon, 24 Nov 2008 23:51:16 +0100 From: Roland Smith <rsmith@xs4all.nl> To: Chris Maness <chris@chrismaness.com> Cc: Wojciech Puchar <wojtek@wojtek.tensor.gdynia.pl>, freebsd-questions@freebsd.org Subject: Re: Xeon Quad Core (Was: Server Freezing Solid) Message-ID: <20081124225115.GA8033@slackbox.xs4all.nl> In-Reply-To: <492B1361.4050604@chrismaness.com> References: <b2e0a4330811110642k688c2c9aq9fbb9832f5382c53@mail.gmail.com> <BMEDLGAENEKCJFGODFOCCEPNCFAA.tedm@toybox.placo.com> <20081112100640.GA21560@icarus.home.lan> <492AF564.5050605@chrismaness.com> <20081124200044.A1528@wojtek.tensor.gdynia.pl> <492B0275.6000300@chrismaness.com> <20081124200755.GA3659@slackbox.xs4all.nl> <492B1361.4050604@chrismaness.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--MGYHOYXEY6WxJCY8 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Mon, Nov 24, 2008 at 12:49:37PM -0800, Chris Maness wrote: > >> Would there be a major performance gain with amd64 over that of the i3= 86=20 > >> build on a Xeon Quad Core? > >> =20 > > > > It will depend on your workload. If your machines were strapped fo > > address space on i386, switching to amd64 (with enough RAM) will help. > > > > In "long" (64-bit) mode, amd64 compatible CPUs have more registers > > available, so that will speed up things. On the other hand, pointers and > > longs are 64-bit numbers instead of 32-bit, which will make the code > > somewhat larger. Run some benchmarks that are relevant for you on i386 > > and re-run them after you've switched to amd64 to know for sure. <snip> > Another thought. Would a Quad Core chip help with compiling=20 > applications -- or would it be the same as a dual core or single core=20 > chip running at the same clock speed because the compiler is running=20 > single thread?=20 Again, it depends. If you have to compile a lot of C files via a Makefile without much interdependencies you could start make with the "-j 8" flag so it can start 8 jobs concurrently. (The number of cores x 2 seems to be the best option). > Would php processing be benefited by quad a quad core=20 > over a dual core. If not, then I guess I should just purchase a dual=20 > core chip and save the cabbage up front and wattage to boot. It could very well benefit. It depends where the bottleneck is in your current setup. It e.g. depends on how many apache and php instances you have running, and how you have compiled apache. Apache 22 is standard compiled with the prefork MPM, which starts 2 processes by default, and can start up to 16 IIRC (both numbers are configurable). A quad processor could make this run faster as long as the rest of the system can keep up. Roland --=20 R.F.Smith http://www.xs4all.nl/~rsmith/ [plain text _non-HTML_ PGP/GnuPG encrypted/signed email much appreciated] pgp: 1A2B 477F 9970 BA3C 2914 B7CE 1277 EFB0 C321 A725 (KeyID: C321A725) --MGYHOYXEY6WxJCY8 Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.9 (FreeBSD) iEYEARECAAYFAkkrL+MACgkQEnfvsMMhpyWHzgCePBBp8xZQJt0bqS2O0olKeW/Z O6kAnjdg5SaTUkQhpK+WTolorGKjW8Nr =2e1r -----END PGP SIGNATURE----- --MGYHOYXEY6WxJCY8--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20081124225115.GA8033>