Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!ub!acsu.buffalo.edu
From: jones@acsu.buffalo.edu (terry a jones)
Newsgroups: comp.sys.ibm.pc.hardware
Subject: Re: AMD386-DX
Message-ID: <69454@eerie.acsu.Buffalo.EDU>
Date: 7 Apr 91 05:19:51 GMT
References: <1991Apr2.022516.26153@news.iastate.edu> <68305@eerie.acsu.Buffalo.EDU> <1032@stewart.UUCP>
Sender: news@acsu.Buffalo.EDU
Organization: SUNY Buffalo
Lines: 33
Nntp-Posting-Host: beatrix.eng.buffalo.edu

In article <1032@stewart.UUCP> jerry@stewart.UUCP (Jerry Shekhel) writes:
>jones@acsu.buffalo.edu (terry a jones) writes:
>>
>>The only problem I see with the elevated clock rate 386 is the fact
>>that some of us need the FPU.  I don't see any 40Mhz 80387 parts as yet, and
>>I'd expect costs there to also be out of sight.
>>
>
>In the latest PC-Week, practically every non-Intel manufacturer of numeric
>coprocessors has a full-page advertisement for their new 40MHz 387.  Ain't
>this industry great?
>

	Yeah, but you didn't mention cost.  I'd still much rather have the
FPU on the same die as the CPU.  The performance gain is considerable.
I never meant to imply that there would not be FPUs available at the higher
clock rate.  I do maintain that the benefit of having them is diminished by
the fact that a single die solution at the same clock rate (or slightly less)
will perform better in floating point intensive applications, and possibly
be more cost effective.

	I'm waiting to see if someone other than Intel offers us such a
solution since it seems as though the legal precedence has been established.
Refer to the second portion of my original article.  All AMD has to do is
be careful what they name the device. ;)



Terry Jones   				{rutgers,uunet}!acsu.buffalo.edu!jones
SUNY at Buffalo ECE Dept.	  or: rutgers!ub!jones, jones@acsu.buffalo.edu