Gentoo Archives: gentoo-performance

From: "M. Edward (Ed) Borasky" <znmeb@×××××××.net>
To: gentoo-performance@l.g.o
Subject: Re: [gentoo-performance] performance testing
Date: Sun, 29 Apr 2007 04:24:16
Message-Id: 46341D4B.5030502@cesmail.net
In Reply to: [gentoo-performance] performance testing by "Peter A. H. Peterson"
1 -----BEGIN PGP SIGNED MESSAGE-----
2 Hash: SHA1
3
4 Peter A. H. Peterson wrote:
5 > Hi Everyone,
6 >
7 > My name is Peter Peterson and I represent a group of a grad students
8 > at UCLA. We're in a computer systems performance analysis course and we
9 > were hoping to do a general performance comparison of gentoo vs. a
10 > popular binary i386-compatible distribution (probably ubuntu) in some
11 > "real-world" server tests to try and meaningfully calculate the
12 > performance gains that local compilation provides. (For example,
13 > apache2 requests processed per second on the same hardware.)
14 >
15 > I've subscribed to this list because we want the gentoo community to
16 > be involved in helping us design the tests so that we can hopefully
17 > all feel good about what and how we are testing the systems.
18 >
19 > We have no particular outcome in mind; our group represents a wide
20 > range of computer users, from Mac, Linux, and Windows enthusiasts, and
21 > we have all used a wide variety of Linux distributions. We have simply
22 > noticed that much of the discussion of gentoo's performance advantage
23 > is anecdotal and we're genuinely hoping to provide some meaningful
24 > experimental data for discussion. Also, if anyone knows of any
25 > available benchmark data or papers on this subject, we'd love to hear
26 > about them. There was apparently a paper on slashdot a couple of years
27 > ago, but the host it was on appears to now be squatted. For that
28 > matter, if this is a well understood or closed issue (for example, if
29 > the statistics that people quote are actually from good experimental
30 > data) please let us know.
31 >
32 > Is anyone here interested in discussing this project? We are
33 > specifically interested in discussing methodology, testing suits,
34 > CFLAGS and other options. Our desire is not to "trick out" gentoo or
35 > ubuntu, but rather quantify the performance benefit that gentoo has
36 > over binary distributions with "normal" compile flags (whatever normal
37 > is).
38 >
39 > Thanks for your time,
40 >
41 > Peter Peterson (et al)
42 >
43
44 The one thing I know is that the Ruby interpreter gains quite a bit
45 (about 20 - 30 percent) from being compiled with the appropriate
46 "-march=" flag in GCC, and that most of the other compilation options
47 have much less payoff. I don't have the link handy, but it certainly
48 makes sense.
49
50 Today's chips are optimized to run lots of streams of 386 code, so a
51 "server" test might not show substantial differences. If you've got
52 enough threads to keep the processor(s) busy, you aren't going to see
53 much of an advantage from compiling to "native" code. Where you will see
54 a pronounced advantage is on single-threaded applications, like
55 scientific workstations run.
56
57 Quite frankly, I'd be surprised if a Gentoo server was significantly
58 faster than a CentOS 5 (RHEL 5 clone) server on a high-intensity server
59 workload. But I have tried a lot of distros for scientific workstations,
60 and Gentoo does seem to have an advantage there.
61
62 What's the title of the course? Is this a compiler course?
63 -----BEGIN PGP SIGNATURE-----
64 Version: GnuPG v2.0.3 (GNU/Linux)
65 Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
66
67 iD8DBQFGNB1LaZUb+jwczfoRAprsAKC08NiyZSNnzOV2vAYIEgShQnFRigCgzddR
68 6htoqVXzJE2lSb6R5+0ftZA=
69 =GFdy
70 -----END PGP SIGNATURE-----
71 --
72 gentoo-performance@g.o mailing list

Replies

Subject Author
Re: [gentoo-performance] performance testing Daniel Armyr <daniel.armyr@××××.se>