huff@kuhub.cc.ukans.edu (Steve Huff, University of Kansas, Lawrence) (12/06/90)
Can somebody clue me in on the following two questions about math coprocessors for a 386 clone (Gateway 386/25): 1. I have heard of other companies besides Intel that produce and sell them. Any recommendations? Costs? Compatibility? 2. Is there any way to fool a program (e.g. CAD) into thinking a math coprocessor exists when it really doesn't? Besides the extreme delay in running everything through the 386, would it cause any other problems? Thanks. E-mail replies appreciated. Summaries available upon request. ------------------------------------------------------------------------------- Steve Huff, student, University of Kansas HomeNet: 913 749 4720 Internet: HUFF@kuhub.cc.ukans.edu Bitnet: HUFF@Ukanvax.Bitnet Don't_hold_your_breath_net: P.O. Box 1225, Lawrence, KS 66044-8225 -------------------------------------------------------------------------------
ndoduc@framentec.fr (Nhuan Doduc) (12/07/90)
In <27265.275d7c84@kuhub.cc.ukans.edu> huff@kuhub.cc.ukans.edu (Steve Huff, University of Kansas, Lawrence) writes: > that produce and sell them. Any recommendations? > Costs? Compatibility? yes, IIT sells 3C87 and Cyrix sells 83D87: complete compatibility, good price ... until now, that is until new things occur ... > 2. Is there any way to fool a program (e.g. CAD) > into thinking a math coprocessor exists when it > really doesn't? Besides the extreme delay in > running everything through the 386, would it > cause any other problems? yes, I've heard about such emulator, and if you REALLY need it, I can try and search thru my vault ... or can try to fetch from some place (in Paris!). --nh Nhuan DODUC, Framentec-Cognitech, Paris, France, ndoduc@framentec.fr or ndoduc@cognitech.fr, Association Francaise des Utilisateurs d'Unix, France, doduc@afuu.fr
tom@bears.ucsb.edu (Tom Weinstein) (12/08/90)
In article <1556@ftc.framentec.fr>, ndoduc@framentec.fr (Nhuan Doduc) writes: > yes, IIT sells 3C87 and Cyrix sells 83D87: complete compatibility, good > price ... until now, that is until new things occur ... Cyrix also has the new 83EMC87, their EMC chip. Very nice from the performance they quote. -- He is Bob...eager for fun. | Tom Weinstein tom@bears.ucsb.edu He wears a smile... Everybody run! | tweinst@polyslo.calpoly.edu
janeri@Lise.Unit.NO (Jan Eri) (12/08/90)
So much for the complete compatibility of the Cyrix math coprocessors: We have already (first attempt) found a 386 clone that the chip doesn't work with. The fact was acknowledged by the dealer. So please, everybody, try before you buy.... Jan Eri jan.eri@sintef.no janeri@lise.unit.no
goes@vx.acs.umn.edu (JAMES GOES) (12/09/90)
In article <27265.275d7c84@kuhub.cc.ukans.edu>, huff@kuhub.cc.ukans.edu (Steve Huff, University of Kansas, Lawrence) writes... >Can somebody clue me in on the following two questions about >math coprocessors for a 386 clone (Gateway 386/25): > > 1. I have heard of other companies besides Intel > that produce and sell them. Any recommendations? > Costs? Compatibility? Just a note for those looking to upgrade '286 machines. AMD is currently offering a pretty good deal on their 80C287 co-processors - $99 + p&h. Guaranteed 100% compatible. Check out recent issues of PC Weak. _______________________________________________________________________________ Jim Goes |BITNET: GOES@UMNACVX or JGOES@UMINN1 Health Services Administration |InterNet: GOES@VX.ACS.UMN.EDU School of Public Health |UUCP: {fav backbone}!uxacs!vxacs!goes University of Minnesota |MaBell: (612) 624-3118 ________________________________________________________________________________ QuayleWatch Quote of the Week: "Quite frankly, teachers are the only profession that teach our family" (recent address to members of the Hudson Institute)
smsmith@hpuxa.ircc.ohio-state.edu (Stephen M. Smith) (12/09/90)
In article <1990Dec8.122259.20246@ugle.unit.no> janeri@lise.unit.no writes: >So much for the complete compatibility of the Cyrix math coprocessors: >We have already (first attempt) found a 386 clone that the chip doesn't >work with. The fact was acknowledged by the dealer. >So please, everybody, try before you buy.... This might be misleading since you are not clear about who the dealer is. Are you talking about the dealer of the Cyrix chip, or the dealer of the 386 clone? The *tone* of this posting makes it look like Cyrix is at fault, but the way you *worded* it makes it look like the dealer refers to the 386 clone maker, and therefore *he* might be at fault for providing an incompatible motherboard (i.e., not using good industry standards). From all the literature I have read, I have not till now even heard a whisper about incompatibility with Cyrix's chips. In fact just last night a was reading a technical article in Byte about the 100% compatibility of these chips; the article went into detail about how many instruction sets were used, etc... Please clarify! S. "Stevie" Smith \ + / <smsmith@hpuxa. \+++++/ " #*&<-[89s]*(k#$@-_=//a2$]'+=.(2_&*%>,,@ ircc.ohio-state. \ + / {7%*@,..":27g)-=,#*:.#,/6&1*.4-,l@#9:-) " edu> \ + / BTW, WYSInaWYG \ + / --witty.saying.ARC
sichermn@beach.csulb.edu (Jeff Sicherman) (12/09/90)
In article <1990Dec8.183756.17223@magnus.ircc.ohio-state.edu> smsmith@hpuxa.ircc.ohio-state.edu (Stephen M. Smith) writes: >In article <1990Dec8.122259.20246@ugle.unit.no> janeri@lise.unit.no writes: >>So much for the complete compatibility of the Cyrix math coprocessors: >>We have already (first attempt) found a 386 clone that the chip doesn't >>work with. The fact was acknowledged by the dealer. >>So please, everybody, try before you buy.... > >This might be misleading since you are not clear about who the >dealer is. Are you talking about the dealer of the Cyrix chip, >or the dealer of the 386 clone? The *tone* of this posting makes >it look like Cyrix is at fault, but the way you *worded* it makes >it look like the dealer refers to the 386 clone maker, and therefore >*he* might be at fault for providing an incompatible motherboard >(i.e., not using good industry standards). > Also, has it been verified that a *real* intel math coprocessor will work in this system. It may be the system's fault, not the chip. Jeff Sicherman
cs161fhn@sdcc10.ucsd.edu (Dennis Lou) (12/09/90)
I've got a 20 Mhz 386DX that used to be a 16 Mhz 386DX until I changed the oscillator. (the CPU and all the C&T chips were stamped '20mhz' so why not... memory is at 80ns anyway). Unfortunately, the documentation I got with the motherboard is rather minimal and I bought it used. What speed 287 can I drop in? I'd very much like to get AMD's 10Mhz $99 special so I can play Falcon 3.0 at higher speed (when it comes out!) The motherboard is supposedly from Everest, but there's no markings on it. -- Dennis Lou | "But Yossarian, what if everyone thought that way?" dlou@ucsd.edu | "Then I'd be crazy to think any other way!" [backbone]!ucsd!dlou +---------------------------------------------------- dlou@ucsd.BITNET cs161fhn@sdcc10.ucsd.edu | Woz went to my high school.
janeri@Lise.Unit.NO (Jan Eri) (12/09/90)
In article <1990Dec8.122259.20246@ugle.unit.no> I wrote: >So much for the complete compatibility of the Cyrix math coprocessors: >We have already (first attempt) found a 386 clone that the chip doesn't >work with. The fact was acknowledged by the dealer. >So please, everybody, try before you buy.... I am sorry that my first posting was misleading. In the above article I'm referring to the dealer of the system, who knew from previous experience that the chip didn't work in this particular system, AND that the Intel chips work OK. (this is confirmed!) This lead me to the same conclusion as Stephen M. Smith : it IS peculiarities with this particular board! (although as far as we have seen the board works fine with DOS and OS/2 in all other respects) The intention with my initial posting was to warn you: It is possible to find systems in which the industry standard Intel chips work OK, but the chip from Cyrix do NOT! for further details please Email, Jan Eri jan.eri@sintef.no janeri@lise.unit.no
davidsen@sixhub.UUCP (Wm E. Davidsen Jr) (12/10/90)
In article <27265.275d7c84@kuhub.cc.ukans.edu> huff@kuhub.cc.ukans.edu (Steve Huff, University of Kansas, Lawrence) writes: | 1. I have heard of other companies besides Intel | that produce and sell them. Any recommendations? | Costs? Compatibility? The vendors are Cyrix and IIT. I don't know about IIT, but Cyrix wants the timing to within spec, while Intel will definitely run with some out of spec timing, such as a 40/60 duty cycle instead of 50/50 on the clock. Sorry, I don't remember which is specified and which will still work... Anyway, with a good motherboard any of them will work. I have used a few of the Cyrix clone of the 387SX (part # not handy). It was a hard choice, the Cyrix is less expensive, lower power (CMOS), and and faster at the same clock speed. I didn't have to think very hard about it. Cost $231 for the last one. | 2. Is there any way to fool a program (e.g. CAD) | into thinking a math coprocessor exists when it | really doesn't? Besides the extreme delay in | running everything through the 386, would it | cause any other problems? There are a few emulators around, I think on simtel20. I don't have their list here. It will be painfully, unbearably slow. -- bill davidsen - davidsen@sixhub.uucp (uunet!crdgw1!sixhub!davidsen) sysop *IX BBS and Public Access UNIX moderator of comp.binaries.ibm.pc and 80386 mailing list "Stupidity, like virtue, is its own reward" -me