home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!darwin.sura.net!Sirius.dfn.de!fauern!LRZnews!reseq!regent.e-technik.tu-muenchen.de!mch
- From: mch@regent.e-technik.tu-muenchen.dbp.de (Michael Hermann)
- Newsgroups: comp.sys.ibm.pc.hardware
- Subject: 486/50 chipsets
- Summary: What chipset to avoid?
- Keywords: chipsets, performance
- Message-ID: <mch.711708544@regent.e-technik.tu-muenchen.de>
- Date: 21 Jul 92 08:49:04 GMT
- Sender: news@reseq.regent.e-technik.tu-muenchen.de
- Lines: 39
-
- I'm going to buy a 486/50 that should last the next 2 years
- (as my old 386sx lasted 3 years). At the moment I think
- EISA would be overkill as I cannot afford a complete EISA-
- peripheral-set (HD,,ETHERNET,VIDEO). Also LocalBus is somewhat
- uncertain but let's see in 2 years ...
-
- So I'm stuck with an ISA-Bus. What I definitely want to avoid
- is a chip-set that has problems with driving the ISA-BUS.
- Particularly it should:
-
- let me allow to slow down the bus or insert wait-states
- provide A16-A23 reasonably fast (or decode CS16 reasonably late)
-
- A second problem is the cache:
- I've read reports about various '386-chipsets and they vary
- widely in their behaviour once the cache misses. I'd like to
- avoid a chipset that has problems with filling the cache
- (like no burst or waiting until entire line is filled before
- the 486 can go on).
- It should do at least staggered refresh and support multiple
- pages of DRAM (performing a page-cycle only if it detects a hit)
-
- I don't need support for mixed-size DRAMS. I could live with
- a relatively small cache (64k) if this means smaller line
- size.
-
- The real thing would have:
-
- selectable associativity (1/2/4)
- selectable line size (letting some CACHE-RAM be unused if necesssary)
- 0 wait states at CACHE-HIT (but at 50 MHz I doubt it)
- 4 or more pages of DRAM supported for direcct hit
- static column or nibble-mode supported
-
- If anyone has good/bad experiences with a particular 486/50-chipset,
- please mail me.
-
- Michael Hermann
- mch@regent.e-technik.tu-muenchen.de
-