Author: Vincent Diepeveen
Date: 15:07:25 07/08/03
Go up one level in this thread
On July 08, 2003 at 16:43:02, Russell Reagan wrote: >On July 08, 2003 at 16:16:00, Rajen Gupta wrote: > >>they didn't > >Thanks. That's all I needed to know. > > >>and i am talking of the chip's overall performance as most >>end-users dont buy a chip for its chess performance only: (even on this forum of >>die-hard computer chess fanatics its quite probable that there are more intel >>based systems being used) if this so-called pentium killer does not measure up >>(and i have a horrible feeling it won't), it will be the end of AMD as we know >>it; for the last 2 yrs they have been telling us: >> >>"never mind the athlon; it is only stop gap; wait for the hammer; it is going >>to smash intel;" and finally after endless delays it does show up it appears to >>be almost geting a smashing itself! > >You do realize that "most end-users who aren't interested in computer chess that >are going to use the Opteron" aren't interested in 3D games either, right? The >Opteron is a server chip. Hehe. for 3d engines the opteron is of course kicking major butt. There was for some popular game a programmer of a big company who exactly explained why and how. If you look around you'll find his story everywhere. It'll speed up 3d engines of course 2 times to use 64 bits. that's why the SGI company which has all 64 bits servers is selling already for years supercomputers that can render in 64 bits way faster than the 32 bits processors. SGI is having a big name in 3d supercomputers! Oceanic research. Mapping the weather in 3d, you name it! >So I guess all of those web server admins who were planning on playing 3D games >on the companies web server should choose the Pentium 4. Good point. > >I also wonder if they recompiled every benchmark they ran for the Opteron with a >compiler that can take advantage of it's newer features. If so, it would be It is a hard problem that there is no good windows compiler yet in 64 bits for the opteron. basically because there is no 64 bits windows version RELEASED yet (let's not confuse with beta's that some companies have). On the other hand the fact that Cray is going to use opterons and hypertransport for 10000 processor supercomputers should ring a bell everywhere. the alternative for intel addicts that SGI offers is the altix3000 with itanium2 madisons. It's very buggy. Soon i'll be able to know how fast diep is at them when running in parallel. It seems right now, interpretting the benches done by HPC (high performance computing) professor Aad v/d Steen that the latency of the Itanium2 to RAM latencies sucks major ass when compared even with old chips like R14000. I find that a bit hard to believe but i'll soon be able to test it myself. >interesting to know where they got the source code for some of that stuff, and >what compiler they used. They made sure to mention that they tuned everything >for top performance on the P4 benchmarks. If they didn't do the same for the >Opteron (which seems unlikely, if it's possible at all), then the benchmark is >worthless. This wouldn't be the least bit suprising, as you are always quick to >find all of the benchmarks that show Intel chips as being faster and post them >here. It is a hard reality that you need a good compiler and that GCC is not exactly 'good' for good architectures like X86-64. The nalimov guys really are pretty superior compiler guys when compared to those 1 hour a month programming GNU guys. Even putting in vaste amounts of money into the gcc compiler won't help much there. It will improve a lot but it is such in hardware that a 10% lead can mean first spot and GCC is slower by more for such cpu's. In general the more registers the more the GCC sucks. Hard work by gcc to get faster at the K7 did help, but still it is only faster beause of using -fbranch-probabilities and such type of optimizations. Any commercial compiler having those options is outgunning gcc by a large margin. those commercial compilers all work under windows, so we will have to wait until there is a 64 bits version of windows for thos compilers. I understand that some companies who produce compilers also have chosen to not support x86-64 yet. It is incredible for me to hear such stories actually. No i'm not going to mention names here. But it is really true. Who can beat microsoft visual c++ when not many make a x86-64 compiler? This simply shows why windows is leading. Some companies make unforgettable mistakes. > >>if AMD disappears, we will then be subject to intel's monopoly which will be a >>horrible thing for us all: hence my disappointment with AMD: always too little >>and too late! > >I doubt the sincerity of this statement, as you take every opportunity you can >to promote Intel and bash AMD.
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.