home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.sys.intel:2675 comp.arch:11613
- Newsgroups: comp.sys.intel,comp.arch
- Path: sparky!uunet!europa.asd.contel.com!howland.reston.ans.net!wupost!spool.mu.edu!umn.edu!csus.edu!netcom.com!norm
- From: norm@netcom.com (Norman Hardy)
- Subject: Re: Superscalar vs. multiple CPUs ?
- Message-ID: <1992Dec13.203624.12416@netcom.com>
- Organization: Netcom - Online Communication Services (408 241-9760 guest)
- References: <PRENER.92Dec12010123@prener.watson.ibm.com> <1992Dec13.063930.8687@netcom.com> <PRENER.92Dec13030928@prener.watson.ibm.com>
- Date: Sun, 13 Dec 1992 20:36:24 GMT
- Lines: 17
-
- In article <PRENER.92Dec13030928@prener.watson.ibm.com> prener@watson.ibm.com (Dan Prener) writes:
- >In article <1992Dec13.063930.8687@netcom.com> norm@netcom.com (Norman Hardy) writes:
- >
- >> In article <PRENER.92Dec12010123@prener.watson.ibm.com> prener@watson.ibm.com (Dan Prener) writes:
- >
- >I was thinking of situations such as a floating-point functional unit
- >having state that includes the current rounding mode.
- >--
- > Dan Prener (prener@watson.ibm.com)
-
- Good point. Thank you. The 88K's systolic array carries
- such state independently for each stage of the array. This
- avoids draining the array upon changing the state. The same
- works for a shared functional unit.
-
- By the way this was either John Cocke's idea or he would
- know whose idea it was.
-