debra@alice.UUCP (Paul De Bra) (07/20/89)
In article <1216@gvgpsa.GVG.TEK.COM> davew@gvgpsa.gvg.tek.com (David C. White) writes: >Does anyone have any hard data comparing the VAXstation 3100 (running >Ultrix) against the SUN 3/80? SUN is claiming the performance is the >same as the VAXstation 3100 but I don't have a 3/80 to compare against >our VAXstation. If anyone has done any benchmarks between these two >systems I would appreciate any data you may have gathered. Here's some hard data, not what you want maybe, bue anyway: I compared a large number of systems, including: Microvax 3200 with the ninth edition unix. (this is more or less the same then the vaxstation 3100) Sun 3/80 with Sun OS 4.0.3. Let me only give you some cpu results as system calls and disk-related stuff is too OS specific: test Sun 3/80 Microvax 3200 sieve 5.1 4.9 sine 10.1 6.0 loop 10.9 11.0 So I would say that they are a pretty close match, except for floating point where the Microvax excells. (yes, the 3/80 was using the fp chip). Paul. -- ------------------------------------------------------ |debra@research.att.com | uunet!research!debra | ------------------------------------------------------
chris@skat.usc.edu (Christopher Ho) (07/21/89)
In article <9642@alice.UUCP> debra@alice.UUCP () writes: >I compared a large number of systems, including: >Microvax 3200 with the ninth edition unix. (this is more or less the same > then the vaxstation 3100) Actually, if I recall correctly, the VAX 3100, 3300 and 3400 incorporate a slightly slower version of DEC's CMOS VAX chipset that is rated at 2.4 VUPs (versus 2.7 for the 3200/3500/3600/62x0). Thus, your 3200 numbers timings should be divided by 0.89. That makes integer stuff slightly slower than the 3/80, but FP calculations still look significantly faster.
rwood@granite.dec.com (Richard Wood) (07/25/89)
In article <18651@usc.edu> chris@skat.usc.edu (Christopher Ho) writes: > In article <9642@alice.UUCP> debra@alice.UUCP () writes: >> I compared a large number of systems, including: >> Microvax 3200 with the ninth edition unix. (this is more or less >> the same then the vaxstation 3100) > > Actually, if I recall correctly, the VAX 3100, 3300 and 3400 incorporate > a slightly slower version of DEC's CMOS VAX chipset that is rated at 2.4 > VUPs (versus 2.7 for the 3200/3500/3600/62x0). Thus, your 3200 numbers > timings should be divided by 0.89. That makes integer stuff slightly > slower than the 3/80, but FP calculations still look significantly > faster. The VAX 3100 machines (VAXstation 3100 and the new microVAX 3100 and VAXserver 3100) have the same clock speed as the VAXstation 3200, and thus should have generally identical performance. Occasional deviations will be the result of significant differences in the internal bus structure; i.e., the fact that the "busless" 3100 doesn't channel disk and graphics traffic through the Qbus as the 3200 does. The microVAX 3300 and 3400 both use the lower clock speed, as noted above. I believe the faster clock speed is 90ns, and the slower is 100ns. -- -- ----------------------------------------------------------------------------- Richard Wood Corporate Worksystems Team Digital Equipment Corp. ================================================================================
awpsys@ultb.UUCP (Andrew W. Potter) (07/26/89)
In article <511@granite.dec.com> rwood@granite.DEC.COM (Richard Wood) writes: > >The VAX 3100 machines (VAXstation 3100 and the new microVAX 3100 and >VAXserver 3100) have the same clock speed as the VAXstation 3200, and >thus should have generally identical performance. From what I understand this is true. However the 3200 (and MicroVAX 3600) systems have the 64 Kbyte secondary memory cache. The 3100 has to survive with the on-chip cache. By comparision the VAX 6200 (don't know about the 6300 and 6400) have a 256 Kbyte secondary cache. The larger cache sizes in the 3200 can make for better performance in memory intensive applications. I personally have noticed that the 3200 that I have with the Wren IV disk (601 MB) and the DILOG SCSI controller performs quite a bit snappier than a VAXstation 3100. - Andy -- Andrew W. Potter Bitnet: awpsys@ritvax.BITNET Systems Programmer Internet: awpsys%ritvax.bitnet@cunyvm.cuny.edu Information Systems and Computing Rochester Institute of Technology, Rochester NY, 14623 (716) 475-6994
hall@gilroy.pa.dec.com (Jon "Maddog" Hall) (07/29/89)
While I do not know the speed of the CVAX chipset for the 3200/3500/3600/62x0 series of processors, the VAXstation 3100 used an 80 nsec. chip, which figured out to be 3.2 VUPs. If you were right about the 3200's CPU speed, then the 3100's integer benchmarks would have been faster than the 3200s. maddog Product Manager ULTRIX Worksystems (Who usually gets so much mail he does not have time to read news)
davew@gvgpsa.GVG.TEK.COM (David C. White) (08/16/89)
Does anyone have any hard data comparing the VAXstation 3100 (running Ultrix) against the SUN 3/80? SUN is claiming the performance is the same as the VAXstation 3100 but I don't have a 3/80 to compare against our VAXstation. If anyone has done any benchmarks between these two systems I would appreciate any data you may have gathered. -- David White Grass Valley Group, Inc. VOICE: +1 916.478.3052 P.O. Box 1114 Grass Valley, CA 95945 FAX: +1 916.478.3778 Internet: davew@gvgpsa.gvg.tek.com UUCP: ...!tektronix!gvgpsa!davew