home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cis.ohio-state.edu!zaphod.mps.ohio-state.edu!news.acns.nwu.edu!network.ucsd.edu!galaxy!watnxt06.ucr.edu
- From: insom@watnxt06.ucr.edu (Chris Ulrich)
- Newsgroups: comp.os.os2.misc
- Subject: Re: Does VESA Local Bus put a huge strain on a CPU?
- Message-ID: <23700@galaxy.ucr.edu>
- Date: 6 Nov 92 04:33:04 GMT
- References: <82994@ut-emx.uucp>
- Sender: news@galaxy.ucr.edu
- Lines: 26
- Nntp-Posting-Host: watnxt06.ucr.edu
-
-
- I see this local bus/EISA debate as quite silly. They are
- useful for two entirely different kinds of cards.
- Local bus is useful for only dumb frame buffers. Frame buffers
- such as SVGA cards require all the bandwidth they can get, and
- also require the CPU to do anything. I suppose this would also
- hold true for sound cards and digitized sound.
- EISA is much better for those cards which have lots of
- intelligence. This is because EISA (and MCA) are designed to
- share the bus between multiple busmasters. There is no ethernet
- or SCSI card in existence which could saturate a 33mb/sec bus.
- Furthermore, even if they could, they would be saturating the bus,
- not CPU cycles, because they would be conducting the transfer
- themselves. The CPU in EISA systems is still able to function
- out of its cache or access non locked memory (memory not accessed
- by the busmaster).
- Local Bus is not even a decent solution to video. A real video
- standard would have the video memory located at the top of the
- memory map (all of it, not in 64k chunks), a smart coprocessor,
- and DMA.
- Unless Local Bus standardizes interrupt sharing and multiple
- busmasters, it is not an acceptable replacement for EISA.
- Of course, that doesnt mean you shouldnt have both...
- --
- insom@ac.ucr.edu You wouldnt treat me this
- insom@ucrvms.bitnet way if I were a palindrome
-