"TeraStor, in short, is at this point not much more than an idea and the three smart fellows that come with it..."
"...How does TeraStor's technology compare with typical hard drives in other respects? The latter have a seek time - the time it takes to locate a random piece of data on the disk - of 15 to 20 milliseconds. They can feed data into the computer's semiconductor memory at 15 MB per second. Gordon Knight is afraid to commit to TeraStor's specs (or price) so soon, but says his product will be competitive with old-fashioned hard drives. McCoy aims to have TeraStor drives on "on store shelves early next year...."
....but yet they weren't exactly shy about sneaking in that 10ms reference in that LAN Times article, weren't they?
forbes.com
On a related note, here is Rodney Van Meter's initial take, taken from the comp.arch.storage newsgroup. RVM is the Caltech dude who maintains the comp.arch.storage web site.
Article 2 of 2
Subject: Re: Terastor
From: rdv@alumnae.caltech.edu (Rodney D. Van Meter)
Date: 1997/03/10
Message-Id: <5g23kp$in2@gap.cco.caltech.edu>
Newsgroups: comp.arch.storage
[More Headers]
From what I can figure out through the marketting hype, TeraStor has
developed a technique that allows the lens and laser of a
magneto-optical system to be mounted on the kind of sliding/flying
head mechanism used in hard disk drives. The closer spacing will
allow smaller laser spots and give you better MO density.
Of course, one of the advantages of MO is that the head ISN'T in
contact with the disk, reducing wear and the chance of head crashes.
Also, I think it will be more difficult (though not necessarily
impossible) to do removable media with a slider.
I'm concerned about media lifetime with a slider, too -- will it have
an effect?
Note the key words in the press releases: "when implemented". Who
knows how much they really have working, and when products will really
be shipping? "Early 1998" sounds VERY optimistic to me if they don't
already have more than bench experiments.
I'm also skeptical of the "order of magnitude denser than anything
else". May be true today, but double-layer DVD should be within that
order of magnitude, even accounting for 5.25" instead of 3.5". Denser
bits doesn't always mean higher useful capacity; magnetic disks have
done nicely with multiple platters. Don't forget that IBM has
promised us 50 GB 3.5" hard drives within the next three years.
The CEO is James McCoy, former CEO of Quantum, so it's not a
fly-by-night unknown out there.
Been there, done that. Going from great prototype to practical
storage peripheral is HARD: MTBF, repeatability, maintainability,
economical, manufacturable, on time, documented and easy to install is
TOUGH.
Stay tuned. Might be great, might be too little too late.
--Rod |
