Correct, one difference between the pgi 3.0 and 3.1 releases is that version 3.1 supports sse on PIII hardware. But as you see on the results for the stream_d bechmark compiled with the PGI 3.0 Fortran compiler this makes not much difference in the results: ------------------------------------------------------------- Function Rate (MB/s) RMS time Min time Max time Copy: 312.2868 0.0636 0.0512 0.0924 Scale: 313.4120 0.0621 0.0511 0.1009 Add: 372.2549 0.0752 0.0645 0.1145 Triad: 369.7024 0.0715 0.0649 0.0850 So, I think the direct comparison isn't so unfair. /Herbert - Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/dmentre/smp-howto/ To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
- Re: SMP Celeron or UP P2? Herbert Wengatz 42850
- Re: Athlon and floating point (was: SMP Celeron... Neil Conway
- Re: Athlon and floating point (was: SMP Cel... Alan Cox
- Re: Athlon and floating point (was: SMP... Herbert Wengatz 42850
- Re: Athlon and floating point (was:... Don Holmgren
- Re: Athlon and floating point ... Herbert Huber
- Re: Athlon and floating po... Alan Cox
- Re: Athlon and floating po... Brian D. Haymore
- Re: Athlon and floating po... Brian D. Haymore
- Re: Athlon and floating po... Alan Cox
- Re: Athlon and floating po... Herbert Huber
- Re: Athlon and floating po... Brian D. Haymore
- Re: Athlon and floating point ... Neil Conway
- Re: Athlon and floating po... Don Holmgren
- Re: Athlon and floating po... Chris Pirih
- Re: Athlon and floating po... Don Holmgren
- Re: Athlon/Intel floating ... Neil Conway
- Re: Athlon and floating point (was:... Herbert Huber
- Re: SMP Celeron or UP P2? xrr
- Re: SMP Celeron or UP P2? Marc Schneiders
- Re: SMP Celeron or UP P2? Greg Youngblood
