Re: Surprizing performances for Postgres on Centrino

From: Dawid Kuroczko <qnex42(at)gmail(dot)com>
To: Jean-Max Reymond <jmreymond(at)gmail(dot)com>
Cc: Postgresql Performance <pgsql-performance(at)postgresql(dot)org>
Subject: Re: Surprizing performances for Postgres on Centrino
Date: 2005-07-07 13:48:06
Message-ID: 758d5e7f05070706481ef034d4@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

On 7/7/05, Jean-Max Reymond <jmreymond(at)gmail(dot)com> wrote:
> On my laptop whith Centrino 1.6 GHz, 512 Mb RAM,
> - it is solved in 1h50' for Linux 2.6
> - it is solved in 1h37' for WXP Professionnal (<troll on> WXP better
> tan Linux ;-) <troll off>)
[...]
> I test CPU, memory performance on my laptop and it seems that the
> performances are not perfect except for one single test: String sort.

Well, Pentium 4 is not the most efficient processor around (despite
all the advertisiing and all the advanced hyper features). Sure it
reaches high GHz rates, but that's not what matters the most.
This is why AMD stopped giving GHz ratings and instead uses numbers
which indicate how their processor relate to Pentium 4s. For instance
AMD Athlon XP 1700+ is running at 1.45 GHz, but competes with
Pentium 4 1.7 GHz.

Same is with Intels Pentium-III line (which evolved into Pentium-M
Centrino actually). Like AMD Athlon, Pentium-M is more efficient
about its clockspeed than Pentium 4. In other words, you shouldn't
compare Pentium 4 and Pentium-M clock-by-clock. Pentium 4
just needs more GHz to do same job as Pentium-M or Athlon.

If you want to get some better (more technical) information, just
google around for reviews and articles. There are plenty of them
recently since Apple intends to use Pentium-M as their future
platform, at least for notebooks. As for technical stuff, for instance
look at:
http://www.tomshardware.com/howto/20050621/37watt-pc-02.html

What really is interesting is the performance difference between
WXP and L26... Are you sure they use exactly the same config
parameters (shared buffers) and have similar statistics (both
VACUUM ANALYZEd recently)?

Regards,
Dawid

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Tom Lane 2005-07-07 13:59:33 Re: CURSOR slowes down a WHERE clause 100 times?
Previous Message Madison Kelly 2005-07-07 13:13:30 Need suggestion high-level suggestion on how to solve a performance problem