Michael_J_Dedina@cup.portal.com (06/29/88)
Would a 30ms hard drive give you significantly better performance than a 60 or 80ms drive on an Amiga (I'd be using it with an A500 and the -MAX- interface)? I remember hearing that it wouldn't make much difference on a PC/XT, because disk access time was insignificant compared to other bottlenecks in the system. ------------------------- Mike Dedina Indiana University Speech Research Lab dedina@gold.bacs.indiana.edu dedina@iubacs.bitnet Michael_J_Dedina@cup.portal.com
steveb@cbmvax.UUCP (Steve Beats) (06/30/88)
In article <6957@cup.portal.com> Michael_J_Dedina@cup.portal.com writes: >Would a 30ms hard drive give you significantly better performance than a 60 or >80ms drive on an Amiga (I'd be using it with an A500 and the -MAX- interface)? >I remember hearing that it wouldn't make much difference on a PC/XT, because >disk access time was insignificant compared to other bottlenecks in the system. >------------------------- >Mike Dedina Definitely yes! I've been working on a Quantum drive with an average 19ms seek time for about a month. That sucker really flies. In the course of doing some testing and diagnostic work, I had to use a couple of drives that were rated at 65ms and 80ms average access times respectively. The difference was spectacular (or apalling, whatever your point of view is). Basically, once you start getting into SCSI transfer rates, a long seek time becomes a major detriment to the overall performance and "feel" of the drive. Bear in mind, if you are getting 500-600K burst transfer rates from your drive, that evens out to about 1ms per 512 byte block. Therefore, stepping to the next track on an 80ms seek time drive just cost you 40K/s of throughput, regardless of interleaving. A point to ponder, no ? Steve
dedina@cup.portal.com (07/01/88)
How much difference does hard disk access time make on the Amiga? I remember hearing that on some systems (like the PC/XT) a 60ms drive would perform almost as well as a 30ms drive, because access time was minor compared to other factors limiting drive performance. Is this also true of the Amiga? Please send email. ------------------- Mike Dedina dedina@gold.bacs.indiana.edu dedina@iubacs.bitnet dedina@cup.portal.com
cjp@antique.UUCP (Charles Poirier) (07/01/88)
In article <4156@cbmvax.UUCP> steveb@cbmvax.UUCP (Steve Beats) writes: >In article <6957@cup.portal.com> Michael_J_Dedina@cup.portal.com writes: >>Would a 30ms hard drive give you significantly better performance than a 60 or > >if you are getting 500-600K burst transfer rates from your drive, that evens >out to about 1ms per 512 byte block. Therefore, stepping to the next track >on an 80ms seek time drive just cost you 40K/s of throughput, regardless of >interleaving. A point to ponder, no ? Well, to compute this is a bit tricky. First, you want track-to-track time rather than average access time. You need to know how many tracks it does per second. Etc.... Back-of-the-envelope-wise, though, I'd say the slow drive would lose a great deal more than 40K/second versus say a Quantum Q280. -- Charles Poirier (decvax,ucbvax,mcnc,attmail)!vax135!cjp "Docking complete... Docking complete... Docking complete..."
thad@cup.portal.com (07/01/88)
Re: disk speeds, yes, the faster drives ARE faster, even under the existing file system. The Maxtors I'm using are faster than the ST251 drives I'm using, and the ST251-1 (28mS) are noticeably faster than the ST251-0 (39mS). Occasionally I put over 1 GB of HD on one of my systems, and that's 11 drives. FOr those who remember my talk at BADGE several months ago, the 190MB Fujutsi tape unit I showed works fine; am completing utility software and, scheduling permitting, may give a mini-update at BADGE of HDs and tape units. One NEVER has enough disk space, so plan for expansion.
farren@gethen.UUCP (Michael J. Farren) (07/01/88)
In article <6957@cup.portal.com> Michael_J_Dedina@cup.portal.com writes: >Would a 30ms hard drive give you significantly better performance than a 60 or >80ms drive on an Amiga (I'd be using it with an A500 and the -MAX- interface)? >I remember hearing that it wouldn't make much difference on a PC/XT, because >disk access time was insignificant compared to other bottlenecks in the system. Well, as a PC owner, I can assure you that it does make a difference, even in that old and antiquated machine. On an Amiga, you'd notice the difference even earlier. Go for the fast drive. -- Michael J. Farren | "INVESTIGATE your point of view, don't just {ucbvax, uunet, hoptoad}! | dogmatize it! Reflect on it and re-evaluate unisoft!gethen!farren | it. You may want to change your mind someday." gethen!farren@lll-winken.llnl.gov ----- Tom Reingold, from alt.flame
daveh@cbmvax.UUCP (Dave Haynie) (07/02/88)
in article <6957@cup.portal.com>, Michael_J_Dedina@cup.portal.com says: > XPortal-User-Id: 1.1001.4529 > Would a 30ms hard drive give you significantly better performance than a 60 or > 80ms drive on an Amiga (I'd be using it with an A500 and the -MAX- interface)? > I remember hearing that it wouldn't make much difference on a PC/XT, because > disk access time was insignificant compared to other bottlenecks in the system. Under 1.2, probably not much difference. Under 1.3, with Fast Filesystem, the only real limiting factors are hard drive and controller speeds. Given a fast enough controller, you'll get a noticable increase. When I went from a 65-or-so millisecond Seagate to an 18ms Quantum, I definitely noticed a difference. Also, FFS on a plain Amiga seems just as fast as FFS on a 68020 system (to the user, I never actually tried DiskPerf or anything), so there's no software getting in the way like in the old filesystem, which really gets faster with a faster CPU. > Mike Dedina -- Dave Haynie "The 32 Bit Guy" Commodore-Amiga "The Crew That Never Rests" {ihnp4|uunet|rutgers}!cbmvax!daveh PLINK: D-DAVE H BIX: hazy "I can't relax, 'cause I'm a Boinger!"