home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.intel
- Path: sparky!uunet!iWarp.intel.com|ssd.intel.com!ssd.intel.com!hays
- From: hays@ssd.intel.com (Kirk Hays)
- Subject: Re: Future of i860 line
- Message-ID: <1992Aug31.170211.18916@SSD.intel.com>
- Sender: usenet@SSD.intel.com
- Nntp-Posting-Host: boston
- Organization: Intel Supercomputer Systems Division
- References: <1992Aug28.182116.4396@megatek.com> <1992Aug29.031847.28867@nas.nasa.gov> <Btr4Iz.GB9@pgroup.com> <1992Aug31.151640.16824@megatek.com>
- Date: Mon, 31 Aug 1992 17:02:11 GMT
- Lines: 44
-
- In article <1992Aug31.151640.16824@megatek.com>, rstewart@ganglia.megatek.uucp (Rich Stewart) writes:
- |> In article <Btr4Iz.GB9@pgroup.com> lfm@pgroup.com (Larry Meadows) writes:
- |> >In article <1992Aug29.031847.28867@nas.nasa.gov> klee@wilbur.nas.nasa.gov (King M. Lee) writes:
- |> >>
- |> >>One reason why the i860(tm) did do better may be that good compilers
- |> >>were not available when it was first introduced. It would have been
- |> >>better for the i860 if they had developed compilers while designing
- |> >>the chip. The compilers have improved, but it may be too late.
- |> >
- |> >Are you listening, Intel?
- |> >
- |> >>
- |> >>Of all the modern micros, the i860(tm) is the only one I know of
- |> >>that allows one to bypass cache. There are two instructions to
- |> >
- |> >Yes! Very good point. If you can't bypass cache then you have
- |> >to hope/arrange for re-use, which is much harder.
- |> >--
- |> >Larry Meadows The Portland Group
- |> >lfm@pgroup.com
- |>
- |> Stop this. You can bypass cache on lots of chips. Most caching cpu's
- |> can have cache bypassed in the page tables. A second way, if your board
- |> is designed to do it, is a software toggle of some sort. I think, most
- |> modern cpu's, have a cache enable/disable pin.
-
- Ah, but tha i860 is the only one that allows you to do it on a load-by-load
- basis, without a lot of OS and page table crufting about, and without flipping
- mode bits, which was Larry's point.
-
- |> And finally, you do not want to use pfld's that might access memory
- |> that is in a cachable address range. (Right Larry?).
-
- Nope, wrong. The pfld instruction works just fine on something which is in cache
- - the main memory access gets ignored, and the value from the cache is returned.
-
- All in all, the i860 is rather strange, but when you hit the "sweet spot", watch
- out.
-
- --
- Kirk Hays - NRA Life.
- Start your day with a little civil disobedience. Call BATF at 1-800-ATF-GUNS or
- 1-800-RUN-GUNS, tell them what you think of gun grabbers and government employees
- that encourage people to narc on their neighbors. Good for the soul.
-