new P4 vs. AMD and ifc vs. g77 benchmark

July 8, 2002, 16:25

See this link for a recent benchmark of natural convection by FEM on a variety of platforms. ifc is the Intel FORTRAN compiler.

NOTE: I am not the author of this so please do not direct specific questions to me. This is for information purposes. Discuss.

http://www.eas.purdue.edu/~scott/Benchtimes.html

July 8, 2002, 17:24

Too bad that not a single test was run on a Windows platform. It would be nice to have at least ONE comparison. I don't think Windows is that bad when it comes to number crunching.

Regarding the Intel Pentium chips, it is apparent that their FPU performance is simply governed by clock speed, regardless of the chip generation (I, II, III, IV). This confirms my earlier posts about the Pentium processors, and it also explains why they all carry the same name.

Impressive performance though!

July 8, 2002, 17:41

I was just sent this link as well:

http://www.geo.lsa.umich.edu/~keken/...ingle_cpu.html

July 9, 2002, 04:04

Can we assume that this was a 512 kB L2 cache P4? In my experience of comparing the 256 kB L2 P4's to equivalent Athlons there is very little difference, and it seems logical to attribute this P4's improved performance to the big L2 cache. FWIW, when I've compared Windows to Linux results using g77 I've found no difference. The compiler is evidently not really interested in the OS!

July 8, 2002, 16:25	new P4 vs. AMD and ifc vs. g77 benchmark	#1
George Bergantz Guest Posts: n/a	See this link for a recent benchmark of natural convection by FEM on a variety of platforms. ifc is the Intel FORTRAN compiler. NOTE: I am not the author of this so please do not direct specific questions to me. This is for information purposes. Discuss. http://www.eas.purdue.edu/~scott/Benchtimes.html

July 8, 2002, 17:24	Re: new P4 vs. AMD and ifc vs. g77 benchmark	#2
Axel Rohde Guest Posts: n/a	Too bad that not a single test was run on a Windows platform. It would be nice to have at least ONE comparison. I don't think Windows is that bad when it comes to number crunching. Regarding the Intel Pentium chips, it is apparent that their FPU performance is simply governed by clock speed, regardless of the chip generation (I, II, III, IV). This confirms my earlier posts about the Pentium processors, and it also explains why they all carry the same name. Impressive performance though!

July 8, 2002, 17:41	yet another set...	#3
George Bergantz Guest Posts: n/a	I was just sent this link as well: http://www.geo.lsa.umich.edu/~keken/...ingle_cpu.html

July 9, 2002, 04:04	Re: new P4 vs. AMD and ifc vs. g77 benchmark	#4
Charles Crosby Guest Posts: n/a	Can we assume that this was a 512 kB L2 cache P4? In my experience of comparing the 256 kB L2 P4's to equivalent Athlons there is very little difference, and it seems logical to attribute this P4's improved performance to the big L2 cache. FWIW, when I've compared Windows to Linux results using g77 I've found no difference. The compiler is evidently not really interested in the OS!