lerici@super.ORG (Peter W. Brewer) (06/14/90)
In article <6662@amelia.nas.nasa.gov> eugene@wilbur.nas.nasa.gov (Eugene N. Miya) writes: >In article <39284@mips.mips.COM> mash@mips.COM (John Mashey) writes: > >>1) I/O will always be a problem, for all of the usual design reasons, > > "When you can give me disks which rotate significantly > faster than 3600 RPM, then we can start talking." > --Don Senzig > Time for wide, short, dense, and durable tapes.. tape media holds more than optical disks, and winchesters and can be made faster and virtually random... I think Seymour spoke about the future of tapes once upon a time.. CDs.. nope the holes can only be so small, better tape technology beats 'em. Where is Nakimichi when you need them ;-) Peter -- Peter Brewer |||| ||||| ||||||||| |||||| //|||||\ |||||| lerici@super.org || ||__ || || || || || THE Supercomputing || || ||^^^^^^\\ || || || Research Center ~~~ |||||||| ||||| || || ||||| \\|||||/ ||||||
eugene@wilbur.nas.nasa.gov (Eugene N. Miya) (06/14/90)
In article <27502@metropolis.super.ORG> lerici@metropolis.UUCP (Peter W. Brewer) writes: >Time for wide, short, dense, and durable tapes.. tape media holds more than >optical disks, The ILLIAC IV had such a optical tape unit in 1975. There are scattered samples of the material around. The company went out of business before delivering. The tape were optical strips about 4 inches wide and maybe 3 feet long. Problems were mechanical, handling. Does any one really need the name of the company? I can ask Ken Stevens. --e. nobuo miya, NASA Ames Research Center, eugene@orville.nas.nasa.gov {uunet,mailrus,other gateways}!ames!eugene
doc@tera.com (Dan Cummings) (06/15/90)
Advances in I/O performance will lag behind advances in CPU speed for some time to come but there is rather a lot that can be done with current technology and current 3600 RPM disks to improve I/O performance. Just as a start I would refer interested parties to the papers coming from John Ousterhout and his group at UC Berkeley on Log Structured File Systems and RAIDs (Redundant Array of Inexpensive Disks). After digesting all of this information I invite you to come up to Seattle and see what we are doing at Tera Computer for high speed I/O. Another good stab at increasing I/O speed is found in the LD cache approach used in Unicos on the Cray YMP. Don't give up yet! Dan Cummings doc@tera.com
hinton@netcom.UUCP (Greg Hinton) (06/16/90)
In article <1990June14.645408090@tera.com> doc@tera.com (Dan Cummings) writes: > [ . . . . ] >Just as a start I would refer interested parties to the papers coming >from John Ousterhout and his group at UC Berkeley on Log Structured File >Systems and RAIDs (Redundant Array of Inexpensive Disks). > [ . . . . ] Can someone post a bibliography of papers on these topics? Thanks. -- Greg Hinton DOMAIN: hinton@netcom.uucp UUCP: uunet!apple!netcom!hinton