POV-Ray : Newsgroups : povray.off-topic : should I get Intel this time ? : Re: should I get Intel this time ? Server Time
1 Oct 2024 09:22:29 EDT (-0400)
  Re: should I get Intel this time ?  
From: Severi Salminen
Date: 3 Apr 2008 05:24:25
Message: <47f4b059$1@news.povray.org>
Fa3ien wrote:
> I'm in the process of buying a new computer (wintel) for the firm.
> 
> Until now, I used to select AMD processors, because they were more
> bang for the buck. (and I wanted to encourage Intel to make better
> processors, ykw, it worked !)
> 
> It still looks like the AMD X2 64 5600+ I've initially selected is
> more bang-for-the-buck than an equivalently priced Intel, but the
> difference is becoming small. (checked THG benchmarks)
> 
> Some people are telling me that AMD has some chipset-drivers problems
> these days, and that, all in all, going for Intel is a safer choice...
> (without being able to an objective source)
> 
> Any insights ?

I switched to Intel when Core 2 Duos came and I have been a very happy
user since then. Intel has the clear technology lead now. Intel makes
faster processors, CPUs which overclock a LOT better, and especially:
CPUs which have better performance/watt ratio. I really hate noise so I
want my HW to produce as little heat as possible. I don't mind spending
a few extra euros for that. AMD must compete with price so you can find
good alternatives there if you don't need the best.

If you go for Intel, consider the new 45nm processors which use even
less power and have even better performance. My rendering program
project has made me drool over the new Q9450 quad core processor which
draws basically less power than my E6400 C2D. And I'd get at least a 2x
speedup compared to my current system. 45nm processors are already in
stores in some countries.

Of course, Intel brings completely new architecture (including new
socket, which makes old mobos incompatible) next year: Nehalem. It looks
very promising.


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.