see http://www.sun.com/desktop/products/ultrafeatures.html $Id: spectable,v 4.9 1995/05/06 22:04:24 jdd Exp $ What this is: A file of reported SPECmarks for various machines. These figures are mostly taken from numbers published on the net, or in manufacturer press releases or reports. This file is organized into four tables, the first reporting SPECbase int92 and fp92, the second reporting SPECbaserate int92 and fp92, the third reporting SPECint92 and SPECFP92, and the fourth reporting integer and FP SPECrate92. SPECmark89 (obsolete) is no longer reported. There are no chip-only entries (as opposed to systems with those chips in them); SPEC is a suite of system benchmarks, not chip benchmarks, and reporting SPEC numbers for a chip alone is misleading. Some specrate numbers have been computed from reported specint/FP92 numbers for various uniprocessor systems. These are indicated by a trailing "c". Manufacturer estimates, or estimates of any sort, are not normally reported. This file is available via anonymous ftp from ftp.cdf.toronto.edu in the file /pub/spectable. A SPEC FAQ describing the SPEC benchmark suite and the SPEC consortium is periodically posted to comp.benchmarks. I strongly recommend reading that document before using these numbers. Note carefully: benchmark results depend not only on processor type, speed, and cache size, but compiler, OS and other machine characteristics that are not reported here. In particular, the compiler can have a significant effect. Quote: " While no one benchmark can fully characterize overall system performance, the results of a variety of realistic benchmarks can give valuable insight into expected real performance. " - SPEC newsl. Disclaimer: These numbers have not been verified. Nobody guarantees their correctness, and there is no guarantee that they accurately reflect the true performance of these systems. Furthermore, this is not a publication of the SPEC consortium and is not endorsed by the SPEC consortium in any way. Please send all corrections, updates, and new entries to jdd@cdf.toronto.edu. John -- John DiMarco Office: EA201B Computing Disciplines Facility Systems Manager Phone: 416-978-1928 University of Toronto Fax: 416-978-1931 http://www.cdf.toronto.edu/personal/jdd/jdd.html ******** Legend ******** Guide to Vendor Acronyms: DEC: Digital Equipment Corporation DG: Data General HP: Hewlett-Packard IBM: International Business Machines RT: Ross Technology SGI: Silicon Graphics Inc. SNI: Siemens-Nixdorf Inc. Guide to processor families: 88000: 88100 68000: 68040 ALPHA: A21064, A21064A, A21066, A21164 HP PA: PA7100, PA7100LC, PA1.1 i86: 80487SX, 80486DX, 80486DX2, 80486DX4, Pentium MIPS: R2000, R3000, R4000, R4400, R4600, R6000, R8000 POWER: POWER, POWER2, MPC601(PowerPC), RSC3308, RSC4608 SPARC: SuperSPARC, SuperSPARC-II, HyperSPARC, MicroSPARC, MicroSPARC-II, FJMB86902(LSI64911), FJMB86903, RT601, RT605, Weitek PowerUp VAX: REX520, SOC, KA46, NVAX, KA650, KA660, KA680 ******** TABLE 1: SPECbase92, SPECbaseFP92 ******** Notes: - SPECint 92 is derived from the results of a set of integer benchmarks, and can be used to estimate a machine's single-tasking performance on integer code. SPECbase92 is a variant of SPECint 92 that reports "baseline" results, using stricter run rules. - SPECfp 92 is derived from the results of a set of floating point benchmarks, and can be used to estimate a machine's single-tasking performance on floating-point code. SPECbase FP 92 is a variant of SPECfp 92 that reports "baseline" results, using stricter run rules. (to sort on specbase int92/FP92, "sort -n +0.48 -0.55"/"sort -n +0.56 -0.63" ) Field start, end column positions: 0,17 18,28 29,36 37,47 48,55 56,63 64,69 70,79 -------------------------------------------------------------------------------- System CPU ClkMHz Cache SPCbase SPCbase Info Source Name (NUMx)Type ext/in Ext+I/D 92 Int. 92 FP. Date Obtained ================= ========== ======= ========== ======= ======= ===== ========== DEC 3000/900 A21064A 39/275 2M+16/16 178.4 244.6 Jul94 Digital DEC 7000/710 A21064A 39/275 4M+16/16 180.0 265.8 Aug94 Digital DEC [24]00/4/166 A21064 33/166 512+8/8 107.5 134.2 Apr95 Digital DEC [24]00/4/233 A21064A 39/233 512+16/16 137.4 174.6 Apr95 Digital DEC 250/4/266 A21064A ??/266 2M+16/16 182.6 246.8 Apr95 www.dec DEC 600/5/266 A21164 38/266 4M+96+8/8 265.9 362.6 Apr95 Digital DEC 1000/4/200 A21064 40/200 2M+8/8 123.3 165.7 Nov94 Digital DEC 2[01]00/4/200 A21064 47/190 1M+8/8 117.5 154.3 Nov94 Digital DEC 2[01]00/4/233 A21064A 38/233 1M+16/16 163.7 192.3 Apr95 Digital DEC 2[01]00/4/275 A21064A 39/275 4M+16/16 187.8 259.5 Apr95 Digital DEC 2100/5/250 A21164 35/250 4M+96+8/8 244.7 356.3 Apr95 Digital DEC 8[24]00/5/300 A21164 42/300 4M+96+8/8 314.4 444.0 Apr95 Digital HP E45 PA7100LC 80 256 74.5 110.6 Mar95 www.hp HP E55 PA7100LC 96 1M 96.1 149.9 Mar95 www.hp SNI PC/E5S Pentium 30/90 256+8/8 82.9 67.9 Jul94 c.bmarks SNI PC/E5S Pentium 33/100 256+8/8 92.4 75.4 Jul94 c.bmarks SNI PC/D5T Pentium 30/60 256+8/8 63.9 48.3 Nov94 c.bmarks SNI PC/D5T Pentium 30/90 256+8/8 83.0 62.4 Nov94 c.bmarks SNI 2-12[05] R4600 50/100 16/16 71.5 ? Nov94 c.bmarks SNI 4-220 R4400 50/100 512+16/16 63.7 ? Nov94 c.bmarks SNI 4-3[34]0 R4400 50/100 1M+16/16 67.7 ? Nov94 c.bmarks SNI 4-420 R4400 75/150 512+16/16 87.1 ? Nov94 c.bmarks SNI 4-4[34]0 R4400 75/150 1M+16/16 94.0 ? Nov94 c.bmarks SNI 4-5[34]0 R4400 75/150 4M+16/16 101.6 ? Nov94 c.bmarks Sun SS20/61 SuprSP 50/60 1M+20/16 ? 95.8 Jun94 SPEC newsl Sun SS20/612 2xSuprSP 50/60 1M+20/16 ? 111.0 Sep94 SPEC newsl Sun SS20/HS11 HyperSP 50/100 256+8/0 ? 117.8 Dec94 SPEC newsl ================================================================================ ******** TABLE 2: Integer/FP SPECbaserate92 ******** Notes: - Integer SPECrate is derived from the results of a set of integer benchmarks run multiple times simultaneously, and can be used to estimate a machine's overall multi-tasking throughput for integer code. It is typically used on MP machines. - Floating-Point SPECrate is derived from the results of a set of floating-point benchmarks run multiple times simultaneously, and can be used to estimate a machine's overall multi-tasking throughput for FP code. It is typically used on MP machines. - SPECbaserate is a variant of SPECrate that reports "baseline" results, using stricter run rules. - Computed specbaserates are indicated by "c". They're computed from Specbase Int92, FP92 (for uniprocessors) using a scaling factor. This number is usually slightly less than or equal to a measured specbaserate on a uniprocessor. The scaling factor is the number of seconds in a week, divided by the time of the longest-running benchmark on the reference SPEC VAX 11/780, which is 604800/25500, or about 23.7. (to sort on int/FP specbaserate, "sort -n +0.48 -0.55" / "sort -n +0.56 -0.63" ) Field start, end column positions: 0,17 18,28 29,36 37,47 48,55 56,63 64,69 70,79 -------------------------------------------------------------------------------- System CPU ClkMHz Cache Integer FloatPt Info Source Name (NUMx)Type ext/in Ext+I/D SPCbrte SPCbrte Date Obtained ================= ========== ======= ========== ======= ======= ===== ========== DEC 3000/700 A21064A 38/225 2M+16/16 3682 5106 Jul94 Digital DEC 3000/900 A21064A 39/275 2M+16/16 4402 5798 Jul94 Digital DEC 7000/710 A21064A 39/275 4M+16/16 4222 6159 Aug94 Digital DEC 7000/720 2xA21064A 39/275 4M+16/16 8550 12344 Aug94 Digital DEC 7000/740 4xA21064A 39/275 4M+16/16 14922 24711 Aug94 Digital DEC 7000/760 6xA21064A 39/275 4M+16/16 22267 37273 Aug94 Digital DEC [24]00/4/166 A21064 33/166 512+8/8 2521 3167 Apr95 Digital DEC [24]00/4/233 A21064A 39/233 512+16/16 3275 4041 Apr95 Digital DEC 250/4/266 A21064A ??/266 2M+16/16 4331c 5854c Apr95 www.dec DEC 600/5/266 A21164 38/266 4M+96+8/8 6307c 8600c Apr95 Digital DEC 1000/4/200 A21064 40/200 2M+8/8 2944 3906 Nov94 Digital DEC 2[01]00/4/200 A21064 47/190 1M+8/8 2786 3594 Nov94 Digital DEC 2[01]00/4/200 2xA21064 47/190 1M+8/8 5495 6914 Nov94 Digital DEC 2100/4/200 4xA21064 47/190 1M+8/8 10537 12384 Nov94 Digital DEC 2[01]00/4/233 A21064A 38/233 1M+16/16 3842 4575 Apr95 Digital DEC 2[01]00/4/233 2xA21064A 38/233 1M+16/16 7367 8605 Apr95 Digital DEC 2100/4/233 4xA21064A 38/233 1M+16/16 14494 15741 Apr95 Digital DEC 2[01]00/4/275 A21064A 39/275 4M+16/16 4423 6182 Apr95 Digital DEC 2[01]00/4/275 2xA21064A 39/275 4M+16/16 8617 12373 Apr95 Digital DEC 2100/4/275 4xA21064A 39/275 4M+16/16 16963 24273 Apr95 Digital DEC 2100/5/250 A21164 35/250 4M+96+8/8 6175 8448 Apr95 Digital DEC 2100/5/250 2xA21164 35/250 4M+96+8/8 11556 17068 Apr95 Digital DEC 2100/5/250 4xA21164 35/250 4M+96+8/8 22017 33127 Apr95 Digital DEC 8[24]00/5/300 A21164 42/300 4M+96+8/8 7831 10632 Apr95 Digital DEC 8[24]00/5/300 2xA21164 42/300 4M+96+8/8 15691 21225 Apr95 Digital DEC 8[24]00/5/300 4xA21164 42/300 4M+96+8/8 30772 42497 Apr95 Digital DEC 8[24]00/5/300 6xA21164 42/300 4M+96+8/8 46584 63388 Apr95 Digital DEC 8400/5/300 8xA21164 42/300 4M+96+8/8 59901 83108 Apr95 Digital DEC 8400/5/300 10xA21164 42/300 4M+96+8/8 74347 102194 Apr95 Digital DEC 8400/5/300 12xA21164 42/300 4M+96+8/8 82663 121155 Apr95 Digital HP E45 PA7100LC 80 256 1767c 2623c Mar95 www.hp HP E55 PA7100LC 96 1M 2279c 3555c Mar95 www.hp SNI PC/E5S Pentium 30/90 256+8/8 1966c 1610c Jul94 c.bmarks SNI PC/E5S Pentium 33/100 256+8/8 2192c 1788c Jul94 c.bmarks SNI PC/D5T Pentium 30/60 256+8/8 1470c 1111c Nov94 c.bmarks SNI PC/D5T Pentium 30/90 256+8/8 1909c 1435c Nov94 c.bmarks SNI 2-12[05] R4600 50/100 16/16 1645c ? Nov94 c.bmarks SNI 4-220 R4400 50/100 512+16/16 1465c ? Nov94 c.bmarks SNI 4-3[34]0 R4400 50/100 1M+16/16 1557c ? Nov94 c.bmarks SNI 4-420 R4400 75/150 512+16/16 2003c ? Nov94 c.bmarks SNI 4-4[34]0 R4400 75/150 1M+16/16 2162c ? Nov94 c.bmarks SNI 4-5[34]0 R4400 75/150 4M+16/16 2337c ? Nov94 c.bmarks SNI 6-5[34]0 12xR4400 75/150 4M+16/16 22878 ? Nov94 c.bmarks SNI 6-5[34]0 16xR4400 75/150 4M+16/16 29316 ? Nov94 c.bmarks SNI 6-5[34]0 20xR4400 75/150 4M+16/16 35111 ? Nov94 c.bmarks SNI 6-5[34]0 24xR4400 75/150 4M+16/16 39427 ? Nov94 c.bmarks Cray CS6400 48xSuprSP 55/60 2M+20/16 75275 95943 Nov94 Cray Cray CS6400 56xSuprSP 55/60 2M+20/16 82851 109477 Nov94 Cray Cray CS6400 64xSuprSP 55/60 2M+20/16 92844 122061 Nov94 Cray ================================================================================ ******** TABLE 3: SPECint92, SPECFP92 ******** Notes: - SPECint 92 is derived from the results of a set of integer benchmarks, and can be used to estimate a machine's single-tasking performance on integer code. - SPECfp 92 is derived from the results of a set of floating point benchmarks, and can be used to estimate a machine's single-tasking performance on floating-point code. (to sort on specint92/specFP92, "sort -n +0.48 -0.55"/"sort -n +0.56 -0.63" ) Field start, end column positions: 0,17 18,28 29,36 37,47 48,55 56,63 64,69 70,79 -------------------------------------------------------------------------------- System CPU ClkMHz Cache SPECint SPECfp Info Source Name (NUMx)Type ext/in Ext+I/D 92 92 Date Obtained ================= ========== ======= ========== ======= ======= ===== ========== ALR PowerVEISA 80487SX 20 64+8 10.7 4.9 Mar93 SPEC newsl CDC 4330 R3000 33 32/32 24.9 23.9 Sep92 SPEC newsl CDC 4360 R3000 33 64/64 24.9 26.7 Sep92 SPEC newsl CDC 4680 R6000 66 512+64/16 40.6 45.1 Sep92 SPEC newsl Compaq Deskpro 80487SX 16 0+8 9.3 4.3 Mar93 SPEC newsl Compaq Deskpro 80487SX 25 64+8 14.2 6.7 Mar93 SPEC newsl Compaq Deskpro 80486DX 33 128+8 18.2 8.3 Sep92 SPEC newsl Compaq Deskpro 80486DX2 25/50 256+8 25.7 12.2 Mar93 SPEC newsl Compaq Deskpro 80486DX2 33/66 256+8 32.2 16.0 Mar93 SPEC newsl Compaq DeskproXL Pentium 33/66 256+8/8 65.1 63.6 Sep93 SPEC newsl Mobius P5-60 Pentium 30/60 ? 50.0 46.7 Jan94 c.s.sun.hw Nekotech MachI A21066 ??/166 1M+16 70 105 Jun94 m.fsale.wk Nekotech MachI A21066 ??/200 1M+16 105 135 Jun94 m.fsale.wk Nekotech MachII A21066 ??/210 2M+16 130 184 Jun94 m.fsale.wk Nekotech MachII A21066 ??/225 2M+16 135 205 Jun94 m.fsale.wk Nekotech MachII A21066 ??/275 2M+16 170 240 Jun94 m.fsale.wk DEC VAX11/780 ? ? 0 1.0 1.0 Jan89 SPEC Refce DEC VAX3100/38 ? ? ? 3.5 3.8 Mar93 DECinfo DEC VAX3100/76 REX520 ? 128 7.1 6.6 Mar93 DECinfo DEC VAX4000VLC SOC ? 25 5.8 6.3 Mar93 DECinfo DEC VAX4000/60 KA46 22.2 ? 11.1 12.6 Mar93 DECinfo DEC VAX4000/90 NVAX 71 2/8 ? 30.2 Sep92 SPEC newsl DEC VAX6000/410 KA660 36 128 ? 7.1 Feb90 uproc rpt DEC VAX6000/510 KA650 62 512 ? 13.3 Sep92 SPEC newsl DEC VAX6000/610 KA680 83 2M ? 39.2 Sep92 SPEC newsl DEC 5000/900 R3000 40 64/64 27.3 29.9 Sep92 SPEC newsl DEC 5000/20 R3000 20 64/64 13.5 18.4 Jun93 DECinfo DEC 5000/25 R3000 25 64/64 15.7 21.7 Jun93 DECinfo DEC 5000/33 R3000 33 64/128 20.9 23.4 Sep92 SPEC newsl DEC 5000/50,150 R4000 50/100 1M+8/8 46.7 45.9 Sep93 c.arch DEC 5000/120 R3000 20 64/64 13.8 18.4 Jun93 DECinfo DEC 5000/125 R3000 25 64/64 16.1 21.7 Jun93 DECinfo DEC 5000/133 R3000 33 64/128 20.9 29.1 Jun93 DECinfo DEC 5000/200 R3000 25 64/64 19.5 26.7 Jun93 DECinfo DEC 5000/240 R3000 40 64/64 27.9 35.8 Jun93 DECinfo DEC 5000/260 R4400 60/120 1M+16/16 57.1 54.5 Sep93 c.arch DEC 5000/280 R4400 60/120 1M+16/16 56.9 55.6 Jun93 DECinfo DEC 2000/300 A21064 30/150 512+8/8 80.9 110.2 Oct93 c.arch/DEC DEC 3000/300 A21064 30/150 256+8/8 66.2 91.5 Apr93 c.s.sun.mc DEC 3000/300L A21064 20/100 256+8/8 45.9 63.6 Apr93 c.s.sun.mc DEC 3000/300LX A21064 25/125 256+8/8 63.5 75.5 May94 SPEC newsl DEC 3000/300X A21064 35/175 256+8/8 84.4 100.5 May94 SPEC newsl DEC 3000/400 A21064 27/133 512+8/8 74.7 112.2 Apr93 c.arch/DEC DEC 3000/500 A21064 30/150 512+8/8 84.4 127.7 Apr93 c.arch/DEC DEC 3000/500X A21064 40/200 512+8/8 110.9 164.1 Apr93 c.s.sun.mc DEC 3000/600S A21064 35/175 2M+8/8 114.1 162.1 Oct93 c.arch/DEC DEC 3000/700 A21064A 38/225 2M+16/16 162.6 230.6 Jul94 Digital DEC 3000/800S A21064 40/200 2M+8/8 138.4 187.6 May94 c.s.sun.hw DEC 3000/900 A21064A 39/275 2M+16/16 189.3 264.1 Jul94 Digital DEC 4000/610 A21064 40/160 1M+8/8 94.6 137.6 Oct93 Digital DEC 4000/710 A21064 38/190 4M+8/8 122.3 185.4 Oct93 c.arch/DEC DEC 7000/610 A21064 50/200 4M+8/8 132.6 200.1 Oct93 c.arch/DEC DEC 7000/710 A21064A 39/275 4M+16/16 193.8 292.6 Aug94 Digital DEC 10000/610 A21064 50/200 4M+8/8 116.5 193.6 Oct93 Digital DEC 200/4/100 A21064 ??/100 512+8/8 74.6 95.2 Feb95 Digital DEC 250/4/266 A21064A ??/266 2M+16/16 198.6 262.5 Apr95 www.dec DEC [24]00/4/166 A21064 33/166 512+8/8 116.8 139.7 Apr95 Digital DEC [24]00/4/233 A21064A 39/233 512+16/16 157.7 183.9 Apr95 Digital DEC 600/5/266 A21164 38/266 4M+96+8/8 288.6 428.8 Apr95 Digital DEC 1000/4/200 A21064 40/200 2M+8/8 135.8 177.0 Nov94 Digital DEC 2[01]00/4/200 A21064 47/190 1M+8/8 131.8 161.0 Nov94 Digital DEC 2[01]00/4/233 A21064A 38/233 1M+16/16 177.3 215.0 Apr95 Digital DEC 2[01]00/4/275 A21064A 39/275 4M+16/16 202.9 292.6 Apr95 Digital DEC 2100/5/250 A21164 35/250 4M+96+8/8 277.1 410.4 Apr95 Digital DEC 8[24]00/5/300 A21164 42/300 4M+96+8/8 341.4 512.9 Apr95 Digital DG 4100 88100 20 16/16 13.1 ? Sep92 SPEC newsl DG 4300 88100 25 16/16 17.4 ? Sep92 SPEC newsl DG 4600 88100 33 16/16 22.6 ? Sep92 SPEC newsl DG 4605 88100 33 64/32 26.1 ? Sep92 SPEC newsl DG 5225 2x88100 25 128/128 20.3 12.1 May93 c.s.sun.hw DG 5500 88100 40 128/128 32.3 41.4 Oct93 DG 6240 4x88100 25 256/256 20.1 ? Sep92 SPEC newsl HP 425t 68040 25 4/4 12.3 10.3 Jun93 DECinfo HP 425e 68040 25 4/4 12.2 9.3 Jun93 DECinfo HP 705 PA1.1 35 32/64 21.9 33.0 Nov92 Sunflash HP 710 PA1.1 50 32/64 31.6 47.6 Oct92 c.arch HP 712/60 PA7100LC 60 64/64+1 58.1 85.5 Jan94 HP HP 712/80i PA7100LC 80 256/256+1 84.1 79.0 Jan94 HP HP 712/80 PA7100LC 80 256/256+1 84.3 122.3 Jun94 c.bmarks HP 715/33 PA7100 33 64/64 32.5 52.4 Jan94 HP HP 715/50 PA7100 50 64/64 49.2 78.8 Jan94 HP HP 715/64 PA7100LC 64 256 66.6 96.5 Aug94 www.hp HP 715/80 PA7100LC 80 256 83.5 120.9 Aug94 www.hp HP 715/100 PA7100LC 100 256 100.1 137.0 Aug94 www.hp HP 715/75 PA7100 75 256/256 82.6 127.2 Jan94 HP HP 720 PA1.1 50 128/256 38.5 66.1 Jun93 DECinfo HP 725 PA7100 50 64/64 37.1 72.8 Apr93 Sunflash HP 725/75 PA7100 75 256/256 80.3 126.8 May94 HP HP 730 PA1.1 66 128/256 47.8 75.4 May92 c.s.sun.hw HP 7[35]5 PA7100 99 256/256 109.1 167.9 Jan94 HP HP 7[35]5/125 PA7150 125 256/256 136 201 Apr94 HP HP 750 PA1.1 66 256/256 48.1 75.0 Oct92 c.arch HP F10 PA1.1 32 32/64 22.0 36.6 Mar93 SPEC newsl HP [F-I]30 PA1.1 48 256/256 37.8 62.4 Mar93 SPEC newsl HP [FH]20 PA1.1 48 64/64 33.6 56.1 Mar93 SPEC newsl HP [GHI]30 PA1.1 48 256/256 37.8 62.4 Apr94 www.hp HP [GHI]40 PA1.1 64 256/256 65.2 91.3 Apr94 www.hp HP [GHI]50 PA7100 96 256/256 100.0 158.5 Apr94 www.hp HP [GHI]60 PA7100 96 1M/1M 108.8 195.3 Apr94 www.hp HP E25 PA7100LC 48 64 45.0 66.6 Mar95 www.hp HP E35 PA7100LC 64 256 65.6 98.5 Mar95 www.hp HP E45 PA7100LC 80 256 82.1 122.9 Mar95 www.hp HP E55 PA7100LC 96 1M 108.0 163.4 Mar95 www.hp HP 807 PA1.1 32 64/32 20.2 ? Sep92 SPEC newsl HP 827/17 PA1.1 48 64/64 31.4 ? Sep92 SPEC newsl HP 847 PA1.1 ? ? 34.8 ? Apr93 DECinfo HP 867 PA1.1 64 256/256 45.6 ? Sep92 SPEC newsl HP 877 PA1.1 64 256/256 45.8 ? Sep92 SPEC newsl HP 897S PA7100 96 ? 78.3 141.6 Sep92 SPEC newsl IBM N40 MPC601 50 32 41.7 51.0 Mar95 www.ibm IBM [2M]20 RSC3308 33.3 8 20.4 29.1 Sep93 c.arch IBM 230 RSC4608 45.5 128+8 28.5 39.9 Sep93 c.arch IBM 250 MPC601 66 32 62.6 72.2 Jul94 www.ibm IBM 250 MPC601 80 32 77.6 89.4 Jul94 www.ibm IBM 25T MPC601 66 32 62.6 78.8 Mar95 www.ibm IBM 25T MPC601 80 32 72.2 90.4 Mar95 www.ibm IBM C10 MPC601 80 32 78.8 90.4 Jul94 www.ibm IBM C10 MPC601 80 1M+32 90.5 100.8 Jul94 www.ibm IBM 320H POWER 25 8/64 20.9 39.4 Nov92 Sunflash IBM 340 POWER 33 8/32 27.7 51.9 Oct92 c.arch IBM 350 POWER 41.6 8/32 35.4 74.2 Nov92 DEC AXP an IBM 355 POWER 41.6 32/32 48.1 83.3 Sep93 c.arch IBM 365,570 POWER 50 32/32 57.5 99.2 Sep93 c.arch IBM 37[05] POWER 62.5 32/32 70.3 121.1 Sep93 c.arch IBM 380 POWER2 59 32/64 99.3 187.2 Mar95 www.ibm IBM 390 POWER2 67 1M+32/64 114.3 205.3 Mar95 www.ibm IBM 39H POWER2 67 2M+32/64 130.2 266.6 Mar95 www.ibm IBM 3AT POWER2 59 32/64 98.2 185.2 May94 IBM IBM 3BT POWER2 67 1M+32/64 113.2 204.3 May94 IBM IBM 3CT POWER2 67 2M+32/64 130.2 266.6 Mar95 www.ibm IBM 40P MPC601 66 32 63.7 67.8 Mar95 www.ibm IBM 40P MPC601 66 256+32 75.1 77.0 Mar95 www.ibm IBM 41T MPC601 80 512+32 88.1 98.7 Jul94 www.ibm IBM 41T MPC601 80 32 78.8 90.4 Jul94 www.ibm IBM 520H POWER 25 8/32 20.9 39.6 May92 c.s.sun.hw IBM 530H POWER 41.6 8/64 28.5 64.6 Mar93 c.s.sun.hw IBM 550 POWER 41.6 8/64 35.4 71.7 May92 c.s.sun.hw IBM 560 POWER 50 8/64 42.0 85.6 Oct92 c.arch IBM [59]80 POWER 62.5 32/64 73.3 134.6 Sep93 c.arch IBM 580H POWER2 55 32/256 97.6 203.9 Sep93 c.arch IBM 590 POWER2 66.6 32/256 121.6 259.7 Jul94 c.bmarks IBM 59H POWER2 66.6 1M+32/128 122.4 250.7 Mar95 www.ibm IBM 970B POWER 50 32/64 58.8 108.9 Sep93 c.arch IBM 990 POWER2 71.5 32/256 126.0 260.4 Sep93 c.arch IBM R24 POWER2 71.5 2M+32/128 134.1 273.8 Jul94 c.bmarks Mips Magnum R4000 50/100 16 36.8 40.0 Oct92 c.arch SGI 4D/25 R3000 20 64/32 14.0 11.1 Jun93 DECinfo SGI 4D/35 R3000 36 64/64 28.0 33.4 Jun93 DECinfo SGI Challenge R4400 50/100 1M+16/16 62.4 66.5 Apr93 c.arch SGI PowerChl,Onyx R8000 75 4M+16/16 108.7 310.6 Jun94 c.arch SGI Crimson R4000 50/100 1M+8/8 61.7 63.4 Oct92 c.arch SGI Crimson R4400 75/150 1M+16/16 86.0 93.2 Nov94 SGI PdcTbl SGI Indigo R3000 33 32/32 22.4 24.2 Nov92 Sunflash SGI Indigo2 R4600 66/133 512+16/16 94.8 72.0 Nov94 SGI PdcTbl SGI Indigo2 R4400 75/150 1M+16/16 90 87 Apr93 c.bmarks SGI Indigo2 R4400 100/200 1M+16/16 119 131 Nov94 SGI PdcTbl SGI PowerIndigo2 R8000 75 2M+16/16 107 265 Nov94 SGI PdcTbl SGI IndigoR4000 R4000 50/100 1M+8/8 57.6 60.3 Mar93 c.s.sun.hw SGI IndyPC R4000 50/100 8/8 34 35 Jul93 SGI anno SGI IndyPC R4600 50/100 16/16 62.8 49.9 May94 SGI anno SGI IndyPC R4600 44/133 16/16 84.9 61.0 Feb95 SGI anno SGI IndySC R4600 44/133 512+16/16 93.7 72.9 Aug94 SGI anno SGI IndySC R4600 50/150 512+16/16 91.7 97.5 Nov94 SGI PdcTbl SGI IndySC R4000 50/100 1M+8/8 59 61 Jul93 SGI anno SGI IndySC R4400 44/175 1M+16/16 122.6 115.5 Feb95 SGI anno Sun SS/ELC FJMB86903 33 64 18.2 17.9 Nov92 Sunflash Sun SS/IPC FJMB86902 25 64 13.8 11.1 Nov92 Sunflash Sun SS/IPX FJMB86903 40 64 21.8 21.5 Nov92 Sunflash Sun SS2 RT601 40 64 21.8 22.8 Oct92 c.arch Sun SS2/PowerUp WeitekPwUP 40/80 16/8 32.2 31.1 Jun93 c.s.sun.an Sun SS10/20 SuprSP 33 20/16 39.8 46.6 Nov92 Sunflash Sun SS10/30 SuprSP 36 20/16 45.2 54.0 Apr93 Cockcroft Sun SS10/40 SuprSP 40 20/16 50.2 60.2 Apr93 Sunflash Sun SS10/41 SuprSP 40/40.3 1M+20/16 53.2 67.8 Apr93 Cockcroft Sun SS10/51 SuprSP 40/50 1M+20/16 65.2 83.0 Apr93 Sunflash Sun Classic,LX MicroSP 50 4/2 26.4 21.0 Nov92 Sunflash Sun Voyager MicroSP2 60 16/8 43.2 36.2 Mar94 Sun Sun SS4/70 MicroSP2 70 16/8 59.6 46.8 Jan95 Sunflash Sun SS4/85 MicroSP2 85 16/8 65.3 53.1 May95 SunIntro Sun SS5/70 MicroSP2 70 16/8 57.0 47.3 Mar94 Sunflash Sun SS5/85 MicroSP2 85 16/8 65.3 53.1 May95 SunIntro Sun SS5/110 MicroSP2 110 16/8 78.6 65.3 May95 SunIntro Sun SS20/50 SuprSP 50 20/16 76.9 80.1 May95 SunIntro Sun SS20/51 SuprSP 40/50 1M+20/16 81.8 89.0 May95 SunIntro Sun SS20/61 SuprSP 50/60 1M+20/16 98.2 107.2 May95 SunIntro Sun SS20/71 SuprSP2 50/75 1M+20/16 125.8 121.2 Jan95 SunIntro Sun SS20/612 2xSuprSP 50/60 1M+20/16 ? 127.1 Sep94 SPEC newsl Sun SS20/HS11 HyperSP 50/100 256+8/0 104.5 127.6 Nov94 SunIntro Sun SS20/HS21 HyperSP 50/125 256+8/0 131.2 153.0 May95 SunIntro Sun SS1000 SuprSP 40/50 1M+20/16 ? 79.9 Jan95 Cockcroft Sun SS1000 2xSuprSP 40/50 1M+20/16 ? 92.3 Jan95 Cockcroft Sun SS1000 4xSuprSP 40/50 1M+20/16 ? 112.8 Jan95 Cockcroft Sun SS1000 8xSuprSP 40/50 1M+20/16 ? 123.1 Jan95 Cockcroft RT 100S-55 HyperSP 40/55 256+8/0 57 74 Aug94 Ross RT 100S-66 HyperSP 40/66 256+8/0 67 87 Aug94 Ross RT 100S-72 HyperSP 40/72 256+8/0 75 96 Aug94 Ross RT 200S-66 HyperSP 50/66 256+8/0 72 94 Aug94 Ross RT 200S-72 HyperSP 50/72 256+8/0 80 105 Aug94 Ross RT 200S-90 HyperSP 50/90 256+8/0 101 120 Apr95 Ross RT 200S-110 HyperSP 50/110 256+8/0 122 142 Apr95 Ross Solbourne 6/901 SuprSP 33 16+1M+20/16 44.0 52.5 Dec92 SPEC newsl SNI PC/E5S Pentium 30/60 256+8/8 60.6 55.1 Sep93 SPEC newsl SNI PC/E5S Pentium 33/66 256+8/8 67.4 61.5 Sep93 SPEC newsl SNI PC/E5S Pentium 30/90 256+8/8 86.3 72.7 Jul94 c.bmarks SNI PC/E5S Pentium 33/100 256+8/8 96.2 81.2 Jul94 c.bmarks SNI PC/D5T Pentium 30/60 256+8/8 65.9 52.4 Nov94 c.bmarks SNI PC/D5T Pentium 30/90 256+8/8 86.0 68.3 Nov94 c.bmarks SNI 2-12[05] R4600 50/100 16/16 76.3 ? Nov94 c.bmarks SNI 4-120 R4400 50/100 16/16 45.6 ? Oct93 c.bmarks SNI 4-120 R4400 50/100 128+16/16 49.7 ? Jan94 c.bmarks SNI 4-220 R4400 50/100 512+16/16 68.2 ? Nov94 c.bmarks SNI 4-3[34]0 R4400 50/100 1M+16/16 71.4 ? Nov94 c.bmarks SNI 4-420 R4400 75/150 512+16/16 92.0 ? Nov94 c.bmarks SNI 4-4[34]0 R4400 75/150 1M+16/16 100.4 ? Nov94 c.bmarks SNI 4-5[34]0 R4400 75/150 4M+16/16 108.7 ? Nov94 c.bmarks SNI 6-120 R4400 50/100 1M+16/16 55.8 ? Nov93 Siemens SNI 6-220 R4400 75/150 4M+16/16 94.2 ? Nov93 Siemens Intel 486DX 80486 50 256+8 30.1 14.0 Oct92 c.arch Intel 486DX2 80486DX2 33/66 0+8 32.4 16.1 Sep92 uproc rpt Micronics M4P 80486DX4 33/100 256+16 51.4 26.6 Mar94 c.arch Intel Xpress Pentium 60 256+8/8 70.4 55.1 Mar95 www.intel Intel Xpress Pentium 66 256+8/8 78.0 63.6 Mar95 www.intel Intel Xpress Pentium 50/75 512+8/8 89.1 68.5 Mar95 www.intel Intel Xpress Pentium 60/90 512+8/8 106.5 81.4 Mar95 www.intel Intel Xpress Pentium 60/90 1M+8/8 110.1 84.4 Mar95 www.intel Intel Xpress Pentium 66/100 512+8/8 118.1 89.9 Mar95 www.intel Intel Xpress Pentium 66/100 1M+8/8 121.9 93.2 Mar95 www.intel Intel Xpress Pentium 60/120 512+8/8 133.7 99.5 Mar95 www.intel Intel Xpress Pentium 60/120 1M+8/8 140.0 103.9 Mar95 www.intel ================================================================================ ******** TABLE 4: Integer/FP SPECrate92 ******** Notes: - Integer SPECrate is derived from the results of a set of integer benchmarks run multiple times simultaneously, and can be used to estimate a machine's overall multi-tasking throughput for integer code. It is typically used on MP machines. - Floating-Point SPECrate is derived from the results of a set of floating-point benchmarks run multiple times simultaneously, and can be used to estimate a machine's overall multi-tasking throughput for FP code. It is typically used on MP machines. - The SS10/51 SPECrate numbers are computed from the SS10/41 numbers. Sun's announcement claims both SPECrate figures for the SS10/51 are identical to those of the SS600/140, which is almost certainly wrong. - Computed specrates are indicated by "c". They're computed from SpecInt92, SpecFP92 (for uniprocessors) using a scaling factor. This number is usually slightly less than or equal to a measured specrate on a uniprocessor. The scaling factor is the number of seconds in a week, divided by the time of the longest-running benchmark on the reference SPEC VAX 11/780, which is 604800/25500, or about 23.7. (to sort on int/FP specrate, "sort -n +0.48 -0.55" / "sort -n +0.56 -0.63" ) Field start, end column positions: 0,17 18,28 29,36 37,47 48,55 56,63 64,69 70,79 -------------------------------------------------------------------------------- System CPU ClkMHz Cache Integer FloatPt Info Source Name (NUMx)Type ext/in Ext+I/D SPCrate SPCrate Date Obtained ================= ========== ======= ========== ======= ======= ===== ========== ALR PowerVEISA 80487SX 20 64+8 254c 116c Mar93 SPEC newsl CDC 4330 R3000 33 32/32 591c 567 Sep92 SPEC newsl CDC 4360 R3000 33 64/64 591c 633 Sep92 SPEC newsl CDC 4680 R6000 66 512+64/16 963c 1070c Sep92 SPEC newsl CDC 4680 2xR6000 66 512+64/16 ? 2232 Sep92 SPEC newsl Compaq Deskpro 80487SX 16 8 221c 102c Mar93 SPEC newsl Compaq Deskpro 80487SX 25 64+8 337c 159c Mar93 SPEC newsl Compaq Deskpro 80486DX 33 128+8 432c 197c Sep92 SPEC newsl Compaq Deskpro 80486DX2 25/50 256+8 610c 289c Mar93 SPEC newsl Compaq Deskpro 80486DX2 33/66 256+8 764c 379c Mar93 SPEC newsl Compaq DeskproXL Pentium 33/66 256+8/8 1544c 1508c Sep93 SPEC newsl Mobius P5-60 Pentium 30/60 ??+8/8 1186c 1108c Jan94 c.s.sun.hw Convex SPP1000 PA7100 100 1M/1M ? 3478 Sep94 Convex Convex SPP1000 8xPA7100 100 1M/1M ? 27701 Sep94 Convex Convex SPP1000 32xPA7100 100 1M/1M ? 95108 Sep94 Convex Nekotech MachI A21066 ??/166 1M+16 1660c 2490c Jun94 c.fsale.wk Nekotech MachI A21066 ??/200 1M+16 2490c 3202c Jun94 c.fsale.wk Nekotech MachII A21066 ??/210 2M+16 3083c 4364c Jun94 c.fsale.wk Nekotech MachII A21066 ??/225 2M+16 3202c 4862c Jun94 c.fsale.wk Nekotech MachII A21066 ??/275 2M+16 4032c 5692c Jun94 c.fsale.wk DEC VAX11/780 ? ? 0 24c 24c Jan89 SPEC Refce DEC VAX3100/38 ? ? ? 83c 90c Mar93 DECinfo DEC VAX3100/76 REX520 ? 128 168c 157c Mar93 DECinfo DEC VAX4000VLC SOC 25 ? 138c 149c Mar93 DECinfo DEC VAX4000/60 KA46 22.2 ? 263c 299c Mar93 DECinfo DEC VAX4000/90 NVAX 71 2/8 ? 716c Sep92 SPEC newsl DEC VAX6000/410 KA660 36 128 ? 168 Feb90 uproc rept DEC VAX6000/510 KA650 62 512 ? 315 Sep92 SPEC newsl DEC VAX6000/610 KA680 83 2M ? 930 Sep92 SPEC newsl DEC 5000/900 R3000 40 64/64 646 709 Sep92 SPEC newsl DEC 5000/20 R3000 20 64/64 320c 351 Jun93 DECinfo DEC 5000/25 R3000 25 64/64 372c 415 Jun93 DECinfo DEC 5000/33 R3000 33 64/128 496c 556c Sep92 SPEC newsl DEC 5000/50,150 R4000 50/100 1M+8/8 1107c 1088c Sep93 c.arch DEC 5000/120 R3000 20 64/64 327c 436c Jun93 DECinfo DEC 5000/125 R3000 25 64/64 382c 514c Jun93 DECinfo DEC 5000/133 R3000 33 64/128 495c 690c Jun93 DECinfo DEC 5000/200 R3000 25 64/64 462c 633c Jun93 DECinfo DEC 5000/240 R3000 40 64/64 661c 848c Jun93 DECinfo DEC 5000/260 R4400 60/120 1M+16/16 1353c 1292c Sep93 c.arch DEC 5000/280 R4400 60/120 1M+16/16 1349c 1318c Jun93 DECinfo DEC 2000/300 A21064 30/150 512+8/8 1930 2634 Oct93 c.arch/DEC DEC 3000/300 A21064 30/150 256+8/8 1535 2137 Apr93 c.s.sun.mc DEC 3000/300L A21064 20/100 256+8/8 1081 1480 Apr93 c.s.sun.mc DEC 3000/300LX A21064 25/125 256+8/8 1506c 1791c May94 SPEC newsl DEC 3000/300X A21064 35/175 256+8/8 2002c 2384c May94 SPEC newsl DEC 3000/400 A21064 27/133 512+8/8 1763 2662 Apr93 c.arch/DEC DEC 3000/500 A21064 30/150 512+8/8 1997 3023 Apr93 c.arch/DEC DEC 3000/500X A21064 40/200 512+8/8 2611 3910 Apr93 c.s.sun.mc DEC 3000/600S A21064 35/175 2M+8/8 2722 3857 Oct93 c.arch/DEC DEC 3000/700 A21064A 38/225 2M+16/16 3944 5482 Jul94 Digital DEC 3000/800S A21064 40/200 2M+8/8 3137 4377 Oct93 c.arch/DEC DEC 3000/900 A21064A 39/275 2M+16/16 4702 6293 Jul94 Digital DEC 4000/610 A21064 40/160 1M+8/8 2198 3247 Oct93 Digital DEC 4000/620 2xA21064 40/160 1M+8/8 3861 6215 May93 sunflash DEC 4000/710 A21064 38/190 4M+8/8 2900 4340 Oct93 c.arch/DEC DEC 4000/720 2xA21064 38/190 4M+8/8 5144 8272 Apr94 DECinfo DEC 7000/610 A21064 50/200 4M+8/8 3250 4701 Apr94 DECinfo DEC 7000/620 2xA21064 50/200 4M+8/8 6347 9329 Apr94 DECinfo DEC 7000/640 4xA21064 50/200 4M+8/8 12463 18719 Apr94 DECinfo DEC 7000/660 6xA21064 50/200 4M+8/8 18956 28157 Apr94 DECinfo DEC 7000/710 A21064A 39/275 4M+16/16 4522 6680 Aug94 Digital DEC 7000/720 2xA21064A 39/275 4M+16/16 8621 13395 Aug94 Digital DEC 7000/740 4xA21064A 39/275 4M+16/16 17450 27008 Aug94 Digital DEC 7000/760 6xA21064A 39/275 4M+16/16 24735 40103 Aug94 Digital DEC 10000/610 A21064 50/200 4M+8/8 2761c 4588c Oct93 Digital DEC 10000/660 6x21064 50/200 4M+8/8 12865 24748 Nov92 c.arch DEC 200/4/100 A21064 ??/100 512+8/8 1769c 2258c Feb95 Digital DEC 250/4/266 A21064A ??/266 2M+16/16 4710c 6226c Apr95 www.dec DEC [24]00/4/166 A21064 33/166 512+8/8 2782 3346 Apr95 Digital DEC [24]00/4/233 A21064A 39/233 512+16/16 3772 4415 Nov94 Digital DEC 600/5/266 A21164 38/266 4M+96+8/8 6845c 10170c Apr95 Digital DEC 1000/4/200 A21064 40/200 2M+8/8 3136 4230 Nov94 Digital DEC 2[01]00/4/200 A21064 47/190 1M+8/8 3123 3835 Nov94 Digital DEC 2[01]00/4/200 2xA21064 47/190 1M+8/8 6178 7296 Nov94 Digital DEC 2[01]00/4/233 A21064A 38/233 1M+16/16 4135 5112 Apr95 Digital DEC 2[01]00/4/233 2xA21064A 38/233 1M+16/16 8284 9676 Apr95 Digital DEC 2100/4/233 4xA21064A 38/233 1M+16/16 15538 17361 Apr95 Digital DEC 2[01]00/4/275 A21064A 39/275 4M+16/16 4711 6827 Apr95 Digital DEC 2[01]00/4/275 2xA21064A 39/275 4M+16/16 9423 13242 Apr95 Digital DEC 2100/4/275 4xA21064A 39/275 4M+16/16 18036 25997 Apr95 Digital DEC 2100/5/250 A21164 35/250 4M+96+8/8 6551 9795 Apr95 Digital DEC 2100/5/250 2xA21164 35/250 4M+96+8/8 13112 18802 Apr95 Digital DEC 2100/5/250 4xA21164 35/250 4M+96+8/8 24996 37928 Apr95 Digital DEC 8[24]00/5/300 A21164 42/300 4M+96+8/8 8551 11981 Apr95 Digital DEC 8[24]00/5/300 2xA21164 42/300 4M+96+8/8 16769 24329 Apr95 Digital DEC 8[24]00/5/300 4xA21164 42/300 4M+96+8/8 33201 48526 Apr95 Digital DEC 8[24]00/5/300 6xA21164 42/300 4M+96+8/8 50778 71286 Apr95 Digital DEC 8400/5/300 8xA21164 42/300 4M+96+8/8 63418 94686 Apr95 Digital DEC 8400/5/300 10xA21164 42/300 4M+96+8/8 80707 117493 Apr95 Digital DEC 8400/5/300 12xA21164 42/300 4M+96+8/8 91580 140571 Apr95 Digital DG 4100 88100 20 16/16 310c ? Sep92 SPEC newsl DG 4300 88100 25 16/16 412c ? Sep92 SPEC newsl DG 4600 88100 33 16/16 536c ? Sep92 SPEC newsl DG 4605 88100 33 64/32 619c ? Sep92 SPEC newsl DG 5225 2x88100 25 128/128 868 532 May93 c.s.sun.hw DG 5500 88100 40 128/128 766c 981c Oct93 DG 5240 4x88100 25 64/64 1591 971 May93 c.s.sun.hw DG 6240 4x88100 25 256/256 1591 ? Sep92 SPEC newsl DG 6280 8x88100 25 512+64/64 3245 ? Sep92 SPEC newsl HP 425t 68040 25 4/4 292c 244c Jun93 DECinfo HP 425e 68040 25 4/4 289c 220c Jun93 DECinfo HP 705 PA1.1 35 32/64 519c 782c Nov92 Sunflash HP 710 PA1.1 50 32/64 749c 1128c Oct92 c.arch HP 712/60 PA7100LC 60 64/64+1 1378c 2028c May94 HP HP 712/80i PA7100LC 80 256/256+1 1995c 1874c Jan94 HP HP 712/80 PA7100LC 80 256/256+1 1999c 2901c Jun94 c.bmarks HP 715/33 PA7100 33 64/64 574c 1067c Mar93 c.s.sun.hw HP 715/50 PA7100 50 64/64 866 1710 Apr93 Sunflash HP 715/64 PA7100LC 64 256 1498 2281 Aug94 www.hp HP 715/75 PA7100 75 256/256 1446c 2678c Sep93 cl.nb.unix HP 715/80 PA7100LC 80 256 1866 2865 Aug94 www.hp HP 715/100 PA7100LC 100 256 2237 3226 Aug94 www.hp HP 720 PA1.1 50 128/256 912 1567c Jun93 DECinfo HP 725 PA7100 50 64/64 866 1710 Apr93 Sunflash HP 725/75 PA7100 75 256/256 1905c 3007c May94 HP HP 730 PA1.1 66 128/256 1133c 1787c May92 c.s.sun.hw HP 7[35]5 PA7100 99 256/256 1832 2950 Nov92 c.arch HP 7[35]5/125 PA7150 125 256/256 3226c 4767c Apr94 HP HP 750 PA1.1 66 256/256 1141 1778c Oct92 c.arch HP F10 PA1.1 32 32/64 521c 867c Mar93 SPEC newsl HP [F-I]30 PA1.1 48 256/256 896c 1479c Mar93 SPEC newsl HP [FH]20 PA1.1 48 64/64 796c 1330c Mar93 SPEC newsl HP [GHI]30 PA1.1 48 256/256 869c 1435c Apr94 www.hp HP [GHI]40 PA1.1 64 256/256 1500c 2100c Apr94 www.hp HP [GHI]50 PA7100 96 256/256 2300c 3646c Apr94 www.hp HP [GHI]60 PA7100 96 1M/1M 2502c 4492c Apr94 www.hp HP E25 PA7100LC 48 64 1067c 1580c Mar95 www.hp HP E35 PA7100LC 64 256 1556c 2336c Mar95 www.hp HP E45 PA7100LC 80 256 1947c 2915c Mar95 www.hp HP E55 PA7100LC 96 1M 2562c 3875c Mar95 www.hp HP 827/17 PA1.1 48 64/64 744c ? Sep92 SPEC newsl HP 847 PA1.1 ? ? 825c ? Apr93 DECinfo HP 867 PA1.1 64 256/256 1201 ? Sep92 SPEC newsl HP 870 2xPA1.1 50 512/512 1515 ? Sep92 SPEC newsl HP 870 3xPA1.1 50 512/512 2051 ? Sep92 SPEC newsl HP 870 4xPA1.1 50 512/512 2479 ? Sep92 SPEC newsl HP 877 PA1.1 64 256/256 1085c ? Sep92 SPEC newsl HP 897S PA7100 96 ? 1857 1937 Sep92 SPEC newsl HP 890 PA1.1 60 2M/2M 1215 1180 Sep92 SPEC newsl HP 890 2xPA1.1 60 2M/2M 2253 2360 Sep92 SPEC newsl HP 890 3xPA1.1 60 2M/2M 3306 3529 Sep92 SPEC newsl HP 890 4xPA1.1 60 2M/2M 4301 4685 Sep92 SPEC newsl HP T500 4xPA7100 90 1M/1M 9017 15341 Jan94 HP HP T500 8xPA7100 90 1M/1M 17114 28341 Jan94 HP HP T500 12xPA7100 90 1M/1M 23717 38780 Jan94 HP IBM N40 MPC601 50 32 989c 1210c Mar95 www.ibm IBM [2M]20 RSC3308 33.3 8 377 543 Sep93 c.arch IBM 230 RSC4608 45.5 128+8 675c 946c Sep93 c.arch IBM 250 MPC601 66 32 1485c 1712c Jul94 www.ibm IBM 250 MPC601 80 32 1840c 2120c Jul94 www.ibm IBM 25T MPC601 66 32 1485c 1869c Mar95 www.ibm IBM 25T MPC601 80 32 1712c 2144c Mar95 www.ibm IBM C10 MPC601 80 32 1869c 2144c Jul94 www.ibm IBM C10 MPC601 80 1M+32 2146c 2391c Jul94 www.ibm IBM 320H POWER 25 8/64 496 935 Nov92 Sunflash IBM 340 POWER 33 8/32 657 1231 Oct92 c.arch IBM 350 POWER 41.6 8/32 821 1542 Nov92 DEC anno IBM 355 POWER 41.6 32/32 961 1936 Sep93 c.arch IBM 365,570 POWER 50 32/32 1148 2301 Sep93 c.arch IBM 37[05] POWER 62.5 32/32 1332 2612 Sep93 c.arch IBM 380 POWER2 59 32/64 2355c 4440c Mar95 www.ibm IBM 390 POWER2 67 1M+32/64 2711c 4869c Mar95 www.ibm IBM 39H POWER2 67 2M+32/64 3088c 6323c Mar95 www.ibm IBM 3AT POWER2 59 32/64 2329c 4393c May94 IBM IBM 3BT POWER2 67 1M+32/64 2685c 4846c May94 IBM IBM 3CT POWER2 67 2M+32/64 3088c 6323c Mar95 www.ibm IBM 40P MPC601 66 32 1511c 1608c Mar95 www.ibm IBM 40P MPC601 66 256+32 1781c 1826c Mar95 www.ibm IBM 41T MPC601 80 512+32 2090c 2341c Jul94 www.ibm IBM 41T MPC601 80 32 1898c 2144c Jul94 www.ibm IBM 520H POWER 25 8/32 495c 939 May92 c.s.sun.hw IBM 530H POWER 41.6 8/64 669 1364 Mar93 c.s.sun.hw IBM 550 POWER 41.6 8/64 840 1701 May92 c.s.sun.hw IBM 560 POWER 50 8/64 999 2028 Oct92 c.arch IBM [59]80 POWER 62.5 32/64 1404 2960 Sep93 c.arch IBM 580H POWER2 55 32/256 2313c 4832c Sep93 c.arch IBM 590 POWER2 66.6 32/256 2884c 6159c Jul94 c.bmarks IBM 59H POWER2 66.6 1M+32/128 2903c 5946c Mar95 www.ibm IBM 970B POWER 50 32/64 1117 2220 Sep93 c.arch IBM 990 POWER2 71.5 32/256 2986c 6171c Sep93 c.arch IBM R24 POWER2 71.5 2M+32/128 3181c 6494c Jul94 c.bmarks Mips Magnum R4000 50/100 8/8 872c 948c Oct92 c.arch SGI 4D/25 R3000 20 64/32 332c 263c Jun93 DECinfo SGI 4D/35 R3000 36 64/64 664c 792c Jun93 DECinfo SGI Challenge R4400 50/100 1M+16/16 1479c 1576c Apr93 c.arch SGI PowerChl,Onyx R8000 75 4M+16/16 2578c 7367c Jun94 c.arch SGI Crimson R4000 50/100 1M+8/8 1383 1459 Oct92 c.arch SGI Crimson R4400 75/150 1M+16/16 2040c 2210c Nov94 SGI PdcTbl SGI Indigo R3000 33 32/32 531 574 Nov92 Sunflash SGI Indigo2 R4600 66/133 512+16/16 2248c 1708c Nov94 SGI PdcTbl SGI Indigo2 R4400 75/150 1M+16/16 2133c 2062c Apr93 c.bmarks SGI Indigo2 R4400 100/200 1M+16/16 2822c 3107c Nov94 SGI PdcTbl SGI PowerIndigo2 R8000 75 2M+16/16 2538c 6285c Nov94 SGI PdcTbl SGI IndigoR4000 R4000 50/100 1M+8/8 1366 1430 Mar93 c.s.sun.hw SGI IndyPC R4000 50/100 8/8 806c 830c Jul93 SGI anno SGI IndyPC R4600 50/100 16/16 1489c 1184c May94 SGI anno SGI IndyPC R4600 44/133 16/16 2014c 1447c Feb95 SGI anno SGI IndySC R4600 44/133 512+16/16 2222c 1729c Aug94 SGI anno SGI IndySC R4600 50/150 512+16/16 2175c 2312c Nov94 SGI PdcTbl SGI IndySC R4000 50/100 1M+8/8 1398c 1446c Jul93 SGI anno SGI IndySC R4400 44/175 1M+16/16 2908c 2739c Feb95 SGI anno SGI Challenge/L 12xR4400 50/100 1M+16/16 13406 17370 May93 c.s.sun.hw SGI Challenge/XL R4400 75/150 1M+16/16 2221 2306 Oct93 SGI/Mashey SGI Challenge/XL 4xR4400 75/150 1M+16/16 8679 9079 Oct93 SGI/Mashey SGI Challenge/XL 8xR4400 75/150 1M+16/16 16849 17854 Oct93 SGI/Mashey SGI Challenge/XL 12xR4400 75/150 1M+16/16 23696 25171 Oct93 SGI/Mashey SGI Challenge/XL 16xR4400 75/150 1M+16/16 27242 33956 Oct93 SGI/Mashey SGI Challenge/XL 20xR4400 75/150 1M+16/16 31073 40013 Oct93 SGI/Mashey SGI Challenge/XL 24xR4400 75/150 1M+16/16 ? 45776 Oct93 SGI/Mashey SGI Challenge/XL 28xR4400 75/150 1M+16/16 ? 53796 Oct93 SGI/Mashey SGI Challenge/XL 32xR4400 75/150 1M+16/16 ? 56840 Oct93 SGI/Mashey Sun SS/ELC FJMB86903 33 64 432 425 Nov92 Sunflash Sun SS/IPC FJMB86902 25 64 327 263 Nov92 Sunflash Sun SS/IPX FJMB86903 40 64 517 510 Nov92 Sunflash Sun SS2 RT601 40 64 517 541 Oct92 c.arch Sun SS2/PowerUp WeitekPwUP 40/80 16/8 763c 737c Jun93 c.s.sun.an Sun SS10/20 SuprSP 33 20/16 943c 1104c Nov92 Sunflash Sun SS10/30 SuprSP 36 20/16 1072 1282 Apr93 Cockcroft Sun SS10/40 SuprSP 40 20/16 1191 1427 Apr93 Sunflash Sun SS10/402 2xSuprSP 40 20/16 2112 2378 Apr93 Sunflash Sun SS10/41 SuprSP 40/40.3 1M+20/16 1264 1607 Apr93 Cockcroft Sun SS10/412 2xSuprSP 40/40.3 1M+20/16 2411 2854 Apr93 Cockcroft Sun SS10/51 SuprSP 40/50 1M+20/16 1580 2008 Apr93 Sunflash Sun SS10/512 2xSuprSP 40/50 1M+20/16 2950 3744 Apr93 Sunflash Sun SS10/514 4xSuprSP 40/50 1M+20/16 5155 5809 Dec93 Sun Sun Classic,LX MicroSP 50 4/2 626 498 Nov92 Sunflash Sun Voyager MicroSP2 60 16/8 1025c 859c Mar94 Sun Sun SS4/70 MicroSP2 70 16/8 1414c 1110c Jan95 Sunflash Sun SS4/85 MicroSP2 85 16/8 1549c 1259c May95 SunIntro Sun SS5/70 MicroSP2 70 16/8 1352c 1122c Mar94 Sunflash Sun SS5/85 MicroSP2 85 16/8 1549c 1259c May95 SunIntro Sun SS5/110 MicroSP2 110 16/8 1864c 1549c May95 SunIntro Sun SS20/50 SuprSP 50 20/16 1628 1842 Mar94 Sunflash Sun SS20/51 SuprSP 40/50 1M+20/16 1731 1995 Mar94 Sunflash Sun SS20/61 SuprSP 50/60 1M+20/16 2092 2418 Mar94 Sunflash Sun SS20/502 2xSuprSP 50 20/16 3218 3193 May95 SunIntro Sun SS20/612 2xSuprSP 50/60 1M+20/16 4492 4888 May95 SunIntro Sun SS20/712 2xSuprSP2 50/75 1M+20/16 5726 5439 Jan95 SunIntro Sun SS20/514 4xSuprSP 40/50 1M+20/16 7072 7341 May95 SunIntro Sun SS20/HS11 HyperSP 50/100 256+8/0 2478c 3026c Nov94 SunIntro Sun SS20/HS21 HyperSP 50/125 256+8/0 3112c 3629c May95 SunIntro Sun SS20/HS22 2xHyperSP 50/125 256+8/0 5600 6399 May95 SunIntro Sun SS20/HS24 4xHyperSP 50/125 256+8/0 8124 8906 May95 SunIntro Sun 600-120 2xRT605 40 64 1043 1066 Sep92 SPEC newsl Sun 600-140 4xRT605 40 64 1847 1930 Sep92 SPEC newsl Sun SS1000 2xSuprSP 40/50 1M+20/16 2730 3681 May93 c.s.sun.hw Sun SS1000 4xSuprSP 40/50 1M+20/16 5318 7076 May93 c.s.sun.hw Sun SS1000 8xSuprSP 40/50 1M+20/16 10113 12710 May93 c.s.sun.hw Sun SS1000E 2xSuprSP 50/60 1M+20/16 3999 4584 Oct94 Sunflash Sun SS1000E 8xSuprSP 50/60 1M+20/16 15414 17113 Oct94 Sunflash Sun SC2000 8xSuprSP 40/40.3 1M+20/16 8047 10600 May93 c.s.sun.hw Sun SC2000 16xSuprSP 40/50 2M+20/16 21196 28064 Oct93 sunflash Sun SC2000E 2xSuprSP 50/60 2M+20/16 4282 4952 Oct94 sunflash Sun SC2000E 20xSuprSP 50/60 2M+20/16 38213 44722 Oct94 sunflash Cray CS6400 24xSuprSP 55/60 2M+20/16 41967 55734 Mar94 c.s.sun.hw Cray CS6400 32xSuprSP 55/60 2M+20/16 54186 72177 Mar94 c.s.sun.hw Cray CS6400 48xSuprSP 55/60 2M+20/16 82522 102235 Nov94 Cray Cray CS6400 56xSuprSP 55/60 2M+20/16 95262 115802 Nov94 Cray Cray CS6400 64xSuprSP 55/60 2M+20/16 101969 129843 Nov94 Cray Solbourne 6/901 SuprSP 33 16M+1M+20/1 1043c 1244c Dec92 SPEC newsl RT 100S-55 HyperSP 40/55 256+8/0 1352c 1755c Aug94 Ross RT 100D-55 2xHyperSP 40/55 256+8/0 2368 2838 Aug94 Ross RT 100Q-55 4xHyperSP 40/55 256+8/0 4554 5457 Aug94 Ross RT 100S-66 HyperSP 40/66 256+8/0 1589c 2063c Aug94 Ross RT 100D-66 2xHyperSP 40/66 256+8/0 2817 3377 Aug94 Ross RT 100Q-66 4xHyperSP 40/66 256+8/0 5419 6470 Aug94 Ross RT 100S-72 HyperSP 40/72 256+8/0 1779c 2277c Aug94 Ross RT 100D-72 2xHyperSP 40/72 256+8/0 3073 3684 Aug94 Ross RT 100Q-72 4xHyperSP 40/72 256+8/0 5912 7058 Aug94 Ross RT 200S-66 HyperSP 50/66 256+8/0 1708c 2229c Aug94 Ross RT 200D-66 2xHyperSP 50/66 256+8/0 3042 3647 Aug94 Ross RT 200Q-66 4xHyperSP 50/66 256+8/0 5853 6988 Aug94 Ross RT 200S-72 HyperSP 50/72 256+8/0 1897c 2490c Aug94 Ross RT 200D-72 2xHyperSP 50/72 256+8/0 3318 3979 Aug94 Ross RT 200Q-72 4xHyperSP 50/72 256+8/0 6385 7623 Aug94 Ross RT 200S-90 HyperSP 50/90 256+8/0 2395c 2846c Apr95 Ross RT 200Q-90 4xHyperSP 50/90 256+8/0 8457 10024 Apr95 Ross RT 200Q-100 4xHyperSP 50/100 256+8/0 9132 11389 Apr95 SunExpert RT 200S-110 HyperSP 50/110 256+8/0 2894c 3368c Apr95 Ross RT 200Q-110 4xHyperSP 50/110 256+8/0 9988 12026 Apr95 Ross Marix DTH802 2xHyperSP ??/80 256+8/0 3684 4613 Jan95 Marix Marix DSH904 4xHyperSP ??/90 256+8/0 7972 8842 Jan95 Marix SNI PC/E5S Pentium 30/60 256+8/8 1436c 1306c Sep93 SPEC newsl SNI PC/E5S Pentium 33/66 256+8/8 1597c 1458c Sep93 SPEC newsl SNI PC/E5S Pentium 30/90 256+8/8 2047c 1724c Jul94 c.bmarks SNI PC/E5S Pentium 33/100 256+8/8 2282c 1926c Jul94 c.bmarks SNI PC/D5T Pentium 30/60 256+8/8 1516c 1205c Nov94 c.bmarks SNI PC/D5T Pentium 30/90 256+8/8 1978c 1571c Nov94 c.bmarks SNI 2-12[05] R4600 50/100 16/16 1755c ? Nov94 c.bmarks SNI 4-120 R4400 50/100 16/16 1081 ? Oct93 c.bmarks SNI 4-120 R4400 50/100 128+16/16 1177 ? Jan94 c.bmarks SNI 4-220 R4400 50/100 512+16/16 1569c ? Nov94 c.bmarks SNI 4-3[34]0 R4400 50/100 1M+16/16 1642c ? Nov94 c.bmarks SNI 4-4[34]0 R4400 75/150 1M+16/16 2309c ? Nov94 c.bmarks SNI 4-5[34]0 R4400 75/150 4M+16/16 2500c ? Nov94 c.bmarks SNI 6-120 R4400 50/100 1M+16/16 1293 ? Nov93 Siemens SNI 6-120 2xR4400 50/100 1M+16/16 2486 ? Nov93 Siemens SNI 6-120 3xR4400 50/100 1M+16/16 3549 ? Nov93 Siemens SNI 6-120 4xR4400 50/100 1M+16/16 4798 ? Nov93 Siemens SNI 6-140 8xR4400 50/100 1M+16/16 9352 ? Jan94 c.bmarks SNI 6-220 R4400 75/150 4M+16/16 2193 ? Nov93 Siemens SNI 6-220 2xR4400 75/150 4M+16/16 4196 ? Nov93 Siemens SNI 6-220 3xR4400 75/150 4M+16/16 6218 ? Nov93 Siemens SNI 6-220 4xR4400 75/150 4M+16/16 8073 ? Nov93 Siemens SNI 6-240 8xR4400 75/150 4M+16/16 15197 ? Jan94 c.bmarks SNI 6-5[34]0 12xR4400 75/150 4M+16/16 24759 ? Nov94 c.bmarks SNI 6-5[34]0 16xR4400 75/150 4M+16/16 31803 ? Nov94 c.bmarks SNI 6-5[34]0 20xR4400 75/150 4M+16/16 36968 ? Nov94 c.bmarks SNI 6-5[34]0 24xR4400 75/150 4M+16/16 42536 ? Nov94 c.bmarks Intel 486DX 80486 50 256+8 713c 332c Oct92 c.arch Intel 486DX2 80486DX2 33/66 0+8 768c 382c Sep92 uproc rpt Micronics M4P 80486DX4 33/100 256+16 1219c 631c Mar94 c.ar/Intel Intel Xpress Pentium 60 256+8/8 1670c 1307c Mar95 www.intel Intel Xpress Pentium 66 256+8/8 1850c 1508c Mar95 www.intel Intel Xpress Pentium 50/75 512+8/8 2113c 1625c Mar95 www.intel Intel Xpress Pentium 60/90 512+8/8 2526c 1931c Mar95 www.intel Intel Xpress Pentium 60/90 1M+8/8 2611c 2002c Mar95 www.intel Intel Xpress Pentium 66/100 512+8/8 2801c 2132c Mar95 www.intel Intel Xpress Pentium 66/100 1M+8/8 2891c 2210c Mar95 www.intel Intel Xpress Pentium 60/120 512+8/8 3171c 2350c Mar95 www.intel Intel Xpress Pentium 60/120 1M+8/8 3320c 2464c Mar95 www.intel ================================================================================