• MACHINE, OS, compiler, options, performance
• 21264-500(DS20), OSF1 4.0, DEC C V5.8-009,(-O4 -fast -tune ev6 -Dev6 -DMAIN), 612.269 MFLOPS, Ail_max 48, Blj_max 24, A_row_block 341
• 21164A-600(LX164), Linux, egcs-2.90.29,(-O2 ),359.379MFLOPS, Ail_max=24, Blj_max=12
• 21164A-600(?), OSF1, DEC C,(-O2 ),317.473MFLOPS, Ail_max=24, Blj_max=12
• 21164-300(EB164), Linux, egcs-2.91.60,(-O2 -DKMNLOOP),185.681MFLOPS, Ail_max=24, Blj_max=12
• 21164-300(EB164), Linux, egcs-2.91.60,(-O2),182.182MFLOPS, Ail_max=24, Blj_max=12
• 21064A-300(AlphaPC64), Linux, egcs-2.90.21, (-O2 -Dev4), 80MFLOPS, Ail_max=12,Blj_max=6
• 21064A-100(AS200 4/100), Linux, egcs-2.90.21, (-O2 -Dev4), 32MFLOPS, Ail_max=12,Blj_max=6
• PA8000-180(HP D series), HPUX, c89, (+O4 -DKMNLOOP -DAil_max=72 -DBlj_max=8), 242MFLOPS, Ail_max=72,Blj_max=8
• R10000-200(NEC UP4800), UNIX_SV, cc, (-Kopt=2), 32MFLOPS, Ail_max=32,blj_max=16
• Celeron-300A(450MHz), slakware-3.6, gcc (-O2), 136.986 MFLOPS, Ail_max 24, Blj_max 12, A_row_block 85
• PentiumII-450, RedHat-5.1, egcs-1.1.1 (-O2), 148.038 MFLOPS, Ail_max 24, Blj_max 12, A_row_block 85
• PentiumII-400, WindowsNT, VisualC++6.0 (-O2 -G6 ), 114.593 MFLOPS, Ail_max 24, Blj_max 12, A_row_block 85

C=alpha*(A*B)+beta*C where m=k=n=1000, lda=ldb=ldc=1001, alpha=2.0, beta=0.5