Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!usc!apple!agate!shelby!portia.stanford.edu!elaine41.stanford.edu!fangchin
From: fangchin@elaine41.stanford.edu (Chin Fang)
Newsgroups: comp.unix.sysv386
Subject: ESIX and IDE drives
Summary: ESIX Rev.D should not have problems with IDE drives
Keywords: IDE ESIX
Message-ID: <1990Dec10.021335.5221@portia.Stanford.EDU>
Date: 10 Dec 90 02:13:35 GMT
Sender: news@portia.Stanford.EDU
Organization: Stanford University, California, USA
Lines: 42

In an earlier posting of mine, I repeatedly misspelled Connor as
Cornor.  I hope this short message will clear up any confusion that
the misspelled name may cause.

I think even Interactive 386/ix can use IDE as well.  After all, IDE
is basically a "transparent" extension of the ISA bus, so I dare even
to guess most 386 unices can use this drive interface too. I
personally know some micro manufacturers in San Jose area using IDE
drives and running Interactive 386/ix in their development labs.

One additional info I forgot to mention:  Connor drives offer 15 Mhz 
data transfer rate.  In DOS environment, I used the latest Coretest
from Core International to test that claim.  It came very close
indeed. (the drive comes with 64k data buffer, which I could not turn
off, but these days who doesn't?  Lots non-caching controllers like WD1007
comes with 32k - 64k or so so called look ahead buffer anyway)

15 Mhz is fairly good, consider just two years back 10 Mhz was the
state of the art.  Now SCSI II and new ESDI drives all come in 15 Mhz
range.  Some pricy ones even offer 20 Mhz or even 25 Mhz, encroching 
the SMD interface range not long ago reserved only for workstations
and/or minis. You can buy controllers for such really fast drives,
but the $$$ involved would be hard to swallow.

Two years back, when I tested my Miniscribe 3130, coretest
consistently showed a 780Kb/s data transfer rate, much lower than
the claimed 10Mhz in the spec. sheet.

The above results should not be taken as a serious hardware 
performance measure however.  I think most readers of this group know
that such results just a biased view of certain respects of the hardware
being tested, they are NOT the whole picture, for reference only.


Chin Fang
Mechanical Engineering Department
Stanford University
fangchin@portia.stanford.edu
fang@rocket.cadcam.rok.com
{there is no pc anymore. If sparc1 1+ is slower than an EISA 33 Mhz
486, why call the later pc, the former workstations, a more prestigious
catagory?  henceforth I will call Sun's sparcs pc too. Laugh.....}