Gentoo Archives: gentoo-user

From: Joshua Murphy <poisonbl@×××××.com>
To: gentoo-user@l.g.o
Subject: Re: [gentoo-user] Gentoo speed comparison to other distros
Date: Sat, 16 Mar 2013 06:44:46
Message-Id: CAOTuDKo3ufHgbePQ7=DB2pyuywPRDcV5CnsZvo+=vVHWLwVmTg@mail.gmail.com
In Reply to: [gentoo-user] Gentoo speed comparison to other distros by Dale
1 On Thu, Mar 14, 2013 at 4:15 AM, Dale <rdalek1967@×××××.com> wrote:
2
3 > Howdy,
4 >
5 > I was wondering. Has anyone ever seen where a test as been done to
6 > compare the speed of Gentoo with other distros? Maybe Gentoo compared
7 > to Redhat, Mandrake, Ubuntu and such?
8 >
9 > Also, I read that Nasdaq runs a modified version of Gentoo. Do any
10 > other large corps run it that we know of?
11 >
12 > I googled a bit but couldn't find anything. Maybe my search terms
13 > wasn't good enough.
14 >
15 > Links would be nice.
16 >
17 > Dale
18 >
19 > --
20 > I am only responsible for what I said ... Not for what you understood or
21 > how you interpreted my words!
22 >
23 >
24 >
25 While I'll start by backing up everything said by others regarding the
26 differences being nearly negligible in a truly equal test, same feature set
27 from source vs the feature set provided by a binary distro, and even a loss
28 in terms of total time when you include the compile times involved, I do
29 have a bit of anecdotal evidence in Gentoo's favor.
30
31 On the majority of x86 or x86_64 hardware there's very little room for
32 across the board gains in performance over otherwise standard cflags. On
33 slightly less 'normal' hardware, like, say, an Atom N270 based netbook with
34 1GB of ram, however, a few cflags go a *long* way towards having a usable
35 system. My Mini9 shipped with a variant of Ubuntu that's actually built
36 with general optimizations to make it usable on that hardware, and having
37 run the same version of Ubuntu without those optimizations for a day or two
38 on it, the amount of stutter and stalling was almost unbearable. Then, with
39 the help of a desktop (or three) to handle the bulk of the compilation, I
40 moved to Gentoo on it. I hadn't sorted out what cflags would be best, and
41 simply built what I needed to get back to work on it with fairly minimal
42 use flags, and I was rather frustrated to find that it still ran worse than
43 the factory install, once programs had started (though that process
44 was noticeably faster, as it generally is with so much less running in the
45 background). Once I adjusted to the appropriate cflags, the stutter cleared
46 up, things didn't stall frequently, and the system was simply more
47 responsive. I could even watch flash videos full screen without it
48 stuttering, which I'd given up on as a possibility on the system. A vast
49 majority of the gains I saw were simply from clearing away the 80% of
50 Ubuntu's features I have no use for, but when you have a processor that
51 approaches things just a little differently, like an Atom, you really can
52 gain a bit from letting the compiler put things in an order the processor
53 will agree with better.
54
55 --
56 Poison [BLX]
57 Joshua M. Murphy