home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!auspex-gw!guy
- From: guy@Auspex.COM (Guy Harris)
- Newsgroups: comp.sys.sun.hardware
- Subject: Re: An Xy450/Xy451 in a Sun4/110
- Message-ID: <15396@auspex-gw.auspex.com>
- Date: 9 Nov 92 00:12:40 GMT
- References: <Bx8rvx.F8x@lysator.liu.se> <1992Nov5.200343.2170@ukw.uucp> <1936@lysator.liu.se>
- Sender: news@auspex-gw.auspex.com
- Organization: Auspex Systems, Santa Clara
- Lines: 17
- Nntp-Posting-Host: auspex.auspex.com
-
- >The board also
- >lacks some ICs that I think may be neccessary for it to function correctly. The
- >question is if getting those ICs will remove the error reported above. The
- >ICs missing is the big square IC that is supposed to sit below the SPARC
- >CPU chip and the two smaller square ICs right besides it (in another 4/100
- >CPU board one is labelled "Weitek", the other I don't know).
-
- In the case of at least some of those chips, I think the symptom you
- will see if they are missing is absolutely crappy floating-point
- performance.
-
- I.e., I suspect the chips are a Fujitsu MB86910 floating-point
- controller chip and two Weitek floating-point arithmetic chips; if
- they're missing, floating-point instructions will get illegal
- instruction traps, and the floating-point emulation code in the kernel
- will emulate them - slowly. Those chips were, at least for a while,
- optional on the 4/1xx machines.
-