home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.ibm.pc.hardware
- Path: sparky!uunet!nic.unh.edu!kepler.unh.edu!pss1
- From: pss1@kepler.unh.edu (Paul S Secinaro)
- Subject: Re: Hardware Cache
- Message-ID: <1992Aug14.144300.19632@newshost.unh.edu>
- Sender: news@newshost.unh.edu (USENET News System)
- Organization: University of New Hampshire - Durham, NH
- References: <Bsuosv.6MA@cs.dal.ca> <1992Aug14.064957.20034@athena.mit.edu> <1992Aug14.072931.11693@u.washington.edu>
- Distribution: na
- Date: Fri, 14 Aug 1992 14:43:00 GMT
- Lines: 35
-
- In article <1992Aug14.072931.11693@u.washington.edu> newsman@milton.u.washington.edu (Chris Ricci) writes:
-
- >I recently saw an ad for a caching IDE controler card. It was cheap!
- >($99.00) Does anyone know the advantages(Disadvantages) of these. Are
- >they any better than software caches?
-
- I've recently had some experience with a 4MB cache on a DTC 3290HD
- EISA SCSI controller. From what I can see, it's pretty much useless.
- For DOS, Smartdrv 4.0 seems to work much better. For example - take a
- large uuencoded file and try to uudecode it. With the DTC's cache,
- even with writeback enabled, the disk thrashes like crazy. With
- Smartdrv, there's very little disk activity and the operation is much
- faster. And any real OS like UNIX or OS/2 will have built-in caching.
-
- (Does anyone out there have any experience with this controller?
- Maybe I'm doing something wrong?)
-
- Based on my experience, I'd try-before-you-buy to see if you get any
- performance improvement. One advantage is that you might save a
- little DOS memory by not having to load Smartdrv. If you're running a
- 32-bit OS, though, I think you'd be better off buying a non-caching
- controller, and then using the money you save to buy more RAM for your
- motherboard.
-
- Paul
-
-
-
-
-
- --
- Niven's Law #14: There exist minds that think as well as you do, but
- differently.
- Niven's Corollary: The gene-tampered turkey you're talking to isn't
- necessarily one of them.
-