🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.sys.ibm.pc.hardware.storage
Article #125

Re: Quantum Fireball SCSI slower than its EIDE pendant.

#125
From: Stefan Esser
Date: Mon, 06 Nov 1995 14:27
37 lines
2169 bytes
In article <472s5j$qta@worf.qntm.com>, Ralf-Peter Rohbeck <rrohbeck@qntm.com> writes:
|> >Quantum? What about a Fireball Pro[tm] with 256k or 512k cache?
|> Well, as long as there's no reason to increase the RAM size why should
|> we?
|> 
|> BTW, "cache" is misleading. You'll never get any decent hit rates from
|> anything less than a few Megabytes. It is a multi(or rather semi?)line
|> track buffer. When you're running a few MB's worth of SMARTDRV or
|> other cache any smaller amount of drive cache is useless. So it's just
|> track buffering with read ahead and write back.

Don't think so ...

Tagged command queues can increase performance 
and require a few hundred KByte of disk cache ...

Seems enough of a reason to have it, especially
on drives that have slower mechanisms and might
have more of an advantage from command reordering.

But then, most Fireball customers probably won't
use an operating system that is capable of more
than one outstanding I/O ... :)

Anyway. If I was to buy a 1GB drive, I'd choose
the IBM 31080 because it has a decent cache and 
is not a "DOS only" drive (for the price of the
Fireball, too !) ...

Regards, STefan

PS: I like my Atlas a lot, and its large cache :)
-- 
 Stefan Esser, Zentrum fuer Paralleles Rechnen		Tel:	+49 221 4706021
 Universitaet zu Koeln, Weyertal 80, 50931 Koeln	FAX:	+49 221 4705160
 ==============================================================================
 http://www.zpr.uni-koeln.de/~se			  <se@ZPR.Uni-Koeln.DE>
Technical Details
Message-ID: <47l60o$6au@news.rrz.uni-koeln.de>
Path: arc1-ca.pugleaf.net!RRZ.Uni-Koeln.DE!se
References: <46la9g$mh8@majestix.uni-muenster.de> <46lv82$btn@natasha.rmii.com> <46nn94$7hk@wn1.sci.kun.nl> <46o888$jef@decaxp.harvard.edu> <46qhkb$7av@wn1.sci.kun.nl> <472s5j$qta@worf.qntm.com>

Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!RRZ.Uni-Koeln.DE!se
From: se@MI.Uni-Koeln.DE (Stefan Esser)
Newsgroups: comp.sys.ibm.pc.hardware.storage,comp.periphs.scsi
Subject: Re: Quantum Fireball SCSI slower than its EIDE pendant.
Date: 6 Nov 1995 14:27:36 GMT
Organization: Institute for Mathematics, University of Cologne, Germany
Message-ID: <47l60o$6au@news.rrz.uni-koeln.de>
References: <46la9g$mh8@majestix.uni-muenster.de> <46lv82$btn@natasha.rmii.com> <46nn94$7hk@wn1.sci.kun.nl> <46o888$jef@decaxp.harvard.edu> <46qhkb$7av@wn1.sci.kun.nl> <472s5j$qta@worf.qntm.com>
NNTP-Posting-Host: sysiphos.mi.uni-koeln.de
To: Ralf-Peter Rohbeck <rrohbeck@qntm.com>