home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zephyr.ens.tek.com!uw-beaver!news.u.washington.edu!usenet.coe.montana.edu!uakari.primate.wisc.edu!sdd.hp.com!wupost!darwin.sura.net!Sirius.dfn.de!math.fu-berlin.de!news.netmbx.de!Germany.EU.net!mcsun!corton!enst!sphinx!ndoduc
- From: ndoduc@sphinx.enst.fr (Nhuan Doduc)
- Newsgroups: comp.benchmarks
- Subject: Re: Weitek 4167: Speed in MFLOPS?
- Summary: re: Weitek 4167
- Message-ID: <2462@sphinx.enst.fr>
- Date: 16 Aug 92 14:17:36 GMT
- References: <1992Aug14.091124.304%boot.decnet@edwards-tems.af.mil>
- Organization: Telecom-Paris, France
- Lines: 19
-
-
-
- In most of the usual cases, the Weitek 4167 will be *slower* than the 486 (of course
- running at the same frequency -while now you can buy the Overdrive ...): the
- superiority of the Weitek (1167 & 3167 over the 387) is no more true with the
- 486 (but Weitek which has "migrated from a FPU company to a CPU one" may still
- surprise us in the future 8-) ) So my recommendation is that unless the speed
- is 3x faster, then it's not worth the binary incompatibility (but if you still
- want to play with multiple binary, why don't you try the 860 ? )
-
- I don't have in mind exact values of classical benchmarks (Whet, Linpack, SPEC
- -but I can dig my archives if there is any need) but I would say that:
- - at 25MHz, the 4167 is about 5x or 6x the Vax 780 and
- - at 33MHz, it is 9x the 780 (I've never meet one at 50MHz, which seems
- not to exist in France :-) )
-
- --nh
- Nhuan DODUC
- ndoduc@inf.enst.fr or doduc@afuu.fr
-