Author: Aaron Gordon
Date: 21:12:10 04/10/03
Go up one level in this thread
On April 10, 2003 at 20:37:39, Tom Kerrigan wrote: >On April 10, 2003 at 12:57:10, Aaron Gordon wrote: > >>It's possible because Intel engineered their chips to have high MHz, low >>instructions per cycle. Why? Great for marketing. Most people don't know MHz > >I don't think so. > >It can't be a coincidence that this design principle that's "great for >marketing" also yielded a processor that's faster than the competition on most >benchmarks. > >-Tom I've been over this before, a LOT of pages are corrupt. Tomshardware for one fakes reviews, there's already proof of that. Anandtech still runs the biased (and everyone who knows benchmarks knows this) Sysmark 2002, plus he lied about processor wattages. Mistake perhaps? No, because I informed him multiple times and posted the information out of the Intel P4 tech documents.. he still refused to change the page. This isn't even getting started, there's so much stuff going on you wouldn't believe. Just try getting a real AMD system with full bios tweaks infront of you along with a P4 with full bios tweaks. Run everything you can get your hands on, you'll be surprised. The AthlonXP will beat the P4 by a pretty decent margin. These are kids running these review pages, and yes, they're in it for free stuff and money. Their wanting $50 per day, $750 month, etc just to ADVERTISE on their pages. Also, when Intel sends them $5000 worth of hardware do you think they're going to let the AthlonXP smoke it in the benchmarks? Of course not, they're going to want to give "Good ol' Intel" an excellent review, that way in a month they get ANOTHER $5000 of free hardware. I've tested just about everything you can think of... chess, audio encoding/decoding, 3d games, video encoding/decoding, encryption/decryption, molecular dynamics simulations, simulations dealing with airflow over a wing, numerous raytracing programs, DNA string comparing, etc. All (yes, *ALL*) were faster on the Athlon. When I get the money I'll soup up my box further and get a P4 board/chip and run those same tests with newer hardware and make a page (similar to my Crafty benchmark page) showing the results. Almost forgot to mention.. Another reason I don't like the P4 is it has the potential to be *EXTREMELY* slow in certain applications. Take Distributed.net's RC5 decryption for example. It would take nearly a P4 running 6GHz to equal my AthlonXP 2.5GHz. Now, go take netlib's simple fibonacci benchmark. Only a handful of lines of code.. guess what? It runs (MHz for MHz) half as fast on a P4 vs an Athlon. At least with my AthlonXP I don't have to wonder, "Gee, if this was run on an Athlon cpu would I get a score 2x better?". It's always fast, all the time.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.