First 5.25" 1GB drive?
JW
Was wondering what 5.25" hard drive broke the 1GB barrier? Is it possible
that it was IBM's 0663 Corsair?
Thanks.
Rod Speed
> Was wondering what 5.25" hard drive broke the 1GB barrier?
> Is it possible that it was IBM's 0663 Corsair?
Nope, that was a 3.5" drive, 8 platters.
Quadibloc
That could still be what he is looking for, because 5.25 hard drives
may have become obsolete by the time microcomputer hard drives broke
the 1 GB barrier.
There's an interesting timeline here:
http://www.pcworld.com/article/127105/timeline_50_years_of_hard_drives.html
The Corsair is from 1991, while Quantum's Bigfoot series, which
revived the 5.25" hard drive, came out in the mid-to-late 1990s.
John Savard
Rod Speed
>>> Was wondering what 5.25" hard drive broke the 1GB barrier?
>>> Is it possible that it was IBM's 0663 Corsair?
>> Nope, that was a 3.5" drive, 8 platters.
> That could still be what he is looking for,
Nope, not with the specific model he mentioned.
> because 5.25 hard drives may have become obsolete by
> the time microcomputer hard drives broke the 1 GB barrier.
No they didnt, most obviously with the Quantum Bigfoot.
> There's an interesting timeline here:
> http://www.pcworld.com/article/127105/timeline_50_years_of_hard_drives.html
Yes, but it not specific about whether that drive did break
the 3.5" 1GB barrier, if he just got the size wrong.
> The Corsair is from 1991, while Quantum's Bigfoot series, which
> revived the 5.25" hard drive, came out in the mid-to-late 1990s.
5.25" hard drives didnt die out and then got revived, they continued thru that
time and were just seen in the low cost commodity drives with the Bigfoot.
Quadibloc
> Yes, but it not specific about whether that drive did break
> the 3.5" 1GB barrier, if he just got the size wrong.
It does note that this drive was the first to have _thin-film
magnetoresistive heads_. That was the technology that enabled consumer
hard drives to go up from, say, 85 megabytes, all the way to 540
megabytes - which is what the computers of that day supported without
special driver software.
And yet it took them 8 platters to achieve 1 GB. I would say that
pretty much makes it conclusive that making a 1 GB hard drive - even a
full-height 5 1/2" one - with 1 GB capacity that would fit in a PC-
compatible computer would have been physically impossible prior to
that technology becoming available.
The year before that drive came out, here's what the hard disk scene
looked like:
http://harrysaxon.wordpress.com/2008/11/10/computer-prices-circa-may-1990/
John Savard
Rod Speed
>>> There's an interesting timeline here:
>>> http://www.pcworld.com/article/127105/timeline_50_years_of_hard_drives.html
>> Yes, but it not specific about whether that drive did break
>> the 3.5" 1GB barrier, if he just got the size wrong.
> It does note that this drive was the first to have _thin-film magnetoresistive heads_.
Yes.
> That was the technology that enabled consumer hard drives to
> go up from, say, 85 megabytes, all the way to 540 megabytes
Nope, like it says that was the first drive to have them, and it was a 1GB drive, as it says.
That drive wasnt a consumer hard drive either, its a SCSI drive.
> - which is what the computers of that day supported without special driver software.
It didnt need special driver software to get past 540MB
either, its a SCSI drive which doesnt have that problem.
> And yet it took them 8 platters to achieve 1 GB.
Yes.
> I would say that pretty much makes it conclusive that making a
> 1 GB hard drive - even a full-height 5 1/2" one - with 1 GB capacity
> that would fit in a PC- compatible computer would have been
> physically impossible prior to that technology becoming available.
Thats not right. With the bigger platters it would have been perfectly possible.
> The year before that drive came out, here's what the hard disk scene looked like:
> http://harrysaxon.wordpress.com/2008/11/10/computer-prices-circa-may-1990/
That doesnt look right at all.
http://www.hitachigst.com/tech/techlib.nsf/techdocs/D49C63ECD1D8207487256C62006B797A/$file/0663-e1x_ps.pdf
P7 has a list of ISA SCSI adapters it works fine with.
Quadibloc
> Quadibloc wrote
> > That was the technology that enabled consumer hard drives to
> > go up from, say, 85 megabytes, all the way to 540 megabytes
>
> Nope, like it says that was the first drive to have them, and it was a 1GB drive, as it says.
>
> That drive wasnt a consumer hard drive either, its a SCSI drive.
Yes, it wasn't, but it _was_ true that when consumer hard drives with
540 megabytes started becoming available, they also used
magnetoresistive heads. The IBM technology that first appeared in that
specific drive increased HD capacities by about an order of magnitude
in one sudden jump.
So, while I'm not exactly sure about all the hard drives available
before this specific drive came out, I strongly suspect that even
with, say, twenty-four 5 1/2" platters, nobody could have built a 1 GB
drive that could fit in a normal personal computer. It only became
possible for PC hard drives to have that kind of capacity after this
technology was available.
John Savard
Rod Speed
>>> That was the technology that enabled consumer hard drives to
>>> go up from, say, 85 megabytes, all the way to 540 megabytes
>> Nope, like it says that was the first drive to have them, and it was a 1GB drive, as it says.
>> That drive wasnt a consumer hard drive either, its a SCSI drive.
> Yes, it wasn't, but it _was_ true that when consumer hard drives
> with 540 megabytes started becoming available, they also used
> magnetoresistive heads.
Yes, after all, thats only half the number of platters.
> The IBM technology that first appeared in that specific drive increased
> HD capacities by about an order of magnitude in one sudden jump.
Thats overstating it.
> So, while I'm not exactly sure about all the hard drives available
> before this specific drive came out, I strongly suspect that even
> with, say, twenty-four 5 1/2" platters, nobody could have built a
> 1 GB drive that could fit in a normal personal computer.
Corse they could.
> It only became possible for PC hard drives to have
> that kind of capacity after this technology was available.
Yes, but your line about 5.25" drives is just plain wrong.
Quadibloc
> Yes, but your line about 5.25" drives is just plain wrong.
I'm afraid that I have to admit that you're right. I've done some
further digging, and I find that in the September 10, 1990 edition of
InfoWorld, for example, computers are advertised with a 650 megabyte
ESDI drive as an option. That does make it plausible that someone
could have got to 1 GB before the Corsair came out.
Looking at the ads in the back, typical consumer hard drives were
still in the range of 20 MB to 120 MB; but the Maxtor XT-8760E had 677
MB formatted and 769 MB unformatted capacity... it was ESDI, and
_recertified_ specimens of it sold for $1799 from one advertisement.
John Savard
Quadibloc
The January 28, 1991 issue has a 1.2 GB drive from Fujitsu, and this
may well predate the introduction of IBM's Corsair drive that year.
John Savard
Rod Speed
> Quadibloc <jsav...@ecn.ab.ca> wrote
>> Rod Speed <rod.speed....@gmail.com> wrote
Yeah, I couldnt find much on the net about the sequence of drives of that size.
Before the net got up much of a head of steam basically.
Its certainly not my recollection that the drive the OP asked about was
the first over 1GB, even if he just got the physical size of the drive wrong.
Hard to check tho.
JW
wrote in Message id:
<02714054-2bae-46f2...@b13g2000prf.googlegroups.com>:
>On Apr 12, 11:28 am, "Rod Speed" <rod.speed....@gmail.com> wrote:
>> JW wrote
>>
>> > Was wondering what 5.25" hard drive broke the 1GB barrier?
>> > Is it possible that it was IBM's 0663 Corsair?
>>
>> Nope, that was a 3.5" drive, 8 platters.
I had thought that a 3.5" platter would need a 5.25" enclosure, but
looking at the pictures again, you're right - it was a 3.5" drive.
>That could still be what he is looking for, because 5.25 hard drives
>may have become obsolete by the time microcomputer hard drives broke
>the 1 GB barrier.
And I think you're also right. There may never have been a 5.25" 1GB drive
at that point in time, which is probably why I had no luck using Google.
Thanks all.
Ahem A Rivet's Shot
"Rod Speed" <rod.sp...@gmail.com> wrote:
> Its certainly not my recollection that the drive the OP asked about was
> the first over 1GB, even if he just got the physical size of the drive
> wrong.
We fitted 1GB SCSI drives to a bunch of boxes in 1990, I think they
were 5.25" full height drives. There were 4GB drives just being offered but
we went with the 1GB drives for better performance.
--
Steve O'Hara-Smith | Directable Mirror Arrays
C:>WIN | A better way to focus the sun
The computer obeys and wins. | licences available see
You lose and Bill collects. | http://www.sohara.org/
Lawrence Statton
>
> And I think you're also right. There may never have been a 5.25" 1GB drive
> at that point in time, which is probably why I had no luck using Google.
What about the Seagate ST4120xx drives? That was the 1.2G unformatted
(formatted out to ~1008 megabytes, so just a blonde c*nthair short of
1G) disk that I had a shelf full of. It came in SCSI and SMD that I
know of - ST41200N was the SCSI variant and ST41201J was the SMD.
The exact year I can't nail down.
--L
Ben Pfaff
> What about the Seagate ST4120xx drives? That was the 1.2G unformatted
> (formatted out to ~1008 megabytes, so just a blonde c*nthair short of
> 1G) disk that I had a shelf full of. It came in SCSI and SMD that I
> know of - ST41200N was the SCSI variant and ST41201J was the SMD.
A 1008 megabyte drive would be advertised as a 1 GB drive.
--
Ben Pfaff
http://benpfaff.org
Patrick Scheible
> On Apr 12, 6:48=A0pm, "Rod Speed" <rod.speed....@gmail.com> wrote:
> > Quadibloc wrote
>
> > > That was the technology that enabled consumer hard drives to
> > > go up from, say, 85 megabytes, all the way to 540 megabytes
> >
> > Nope, like it says that was the first drive to have them, and it was a 1G=
> B drive, as it says.
> >
> > That drive wasnt a consumer hard drive either, its a SCSI drive.
>
> Yes, it wasn't...
In that time period, higher-end consumer computers used SCSI. Macs
all used SCSI, and many high-end PCs did too.
-- Patrick
Charlie Gibbs
<93ef6f39-2a0d-4fc8...@f15g2000pro.googlegroups.com>,
jsa...@ecn.ab.ca (Quadibloc) writes:
> On Apr 12, 2:31 pm, "Rod Speed" <rod.speed....@gmail.com> wrote:
>
>> Yes, but it not specific about whether that drive did break
>> the 3.5" 1GB barrier, if he just got the size wrong.
>
> It does note that this drive was the first to have _thin-film
> magnetoresistive heads_. That was the technology that enabled consumer
> hard drives to go up from, say, 85 megabytes, all the way to 540
> megabytes - which is what the computers of that day supported without
> special driver software.
s/the computers/MS-DOS and Windows/
(...he said as he installed a 1GB drive in his Amiga...)
--
/~\ cgi...@kltpzyxm.invalid (Charlie Gibbs)
\ / I'm really at ac.dekanfrus if you read it the right way.
X Top-posted messages will probably be ignored. See RFC1855.
/ \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign!
Rod Speed
> Quadibloc <jsa...@ecn.ab.ca> wrote
>>>> That was the technology that enabled consumer hard drives to
>>>> go up from, say, 85 megabytes, all the way to 540 megabytes
>>> Nope, like it says that was the first drive to have them,
>>> and it was a 1G= B drive, as it says.
>>> That drive wasnt a consumer hard drive either, its a SCSI drive.
>> Yes, it wasn't...
> In that time period, higher-end consumer computers used SCSI.
Only the Mac did, and that wasnt higher end.
> Macs all used SCSI,
Yes.
> and many high-end PCs did too.
Nope, few of them did. They mostly used ESDI.
Scott Lurndal
>Patrick Scheible wrote
>> Quadibloc <jsa...@ecn.ab.ca> wrote
>>> Rod Speed <rod.speed....@gmail.com> wrote
>>>> Quadibloc wrote
>
>>>>> That was the technology that enabled consumer hard drives to
>>>>> go up from, say, 85 megabytes, all the way to 540 megabytes
>
>>>> Nope, like it says that was the first drive to have them,
>>>> and it was a 1G= B drive, as it says.
>
>>>> That drive wasnt a consumer hard drive either, its a SCSI drive.
>
>>> Yes, it wasn't...
>
>> In that time period, higher-end consumer computers used SCSI.
>
>Only the Mac did, and that wasnt higher end.
Actually, my Amiga 1000 had a (aftermarket) SCSI adapter and
an 80MB quantum fireball (iirc) drive, circa 1989.
soctt
Rod Speed
> Rod Speed <rod.sp...@gmail.com> wrote
>> Patrick Scheible wrote
>>> Quadibloc <jsa...@ecn.ab.ca> wrote
>>>> Rod Speed <rod.speed....@gmail.com> wrote
>>>>> Quadibloc wrote
>>>>>> That was the technology that enabled consumer hard drives to
>>>>>> go up from, say, 85 megabytes, all the way to 540 megabytes
>>>>> Nope, like it says that was the first drive to have them,
>>>>> and it was a 1G= B drive, as it says.
>>>>> That drive wasnt a consumer hard drive either, its a SCSI drive.
>>>> Yes, it wasn't...
>>> In that time period, higher-end consumer computers used SCSI.
>> Only the Mac did, and that wasnt higher end.
> Actually, my Amiga 1000 had a (aftermarket) SCSI adapter
There were lots of those for the PCs.
Thats a separate matter to his claim about higher end consumer computers,
with after market adapters you'd have to say COULD use not used.
And the Amiga 1000 isnt a higher end consumer computer anyway.
Rod Speed
> JW <no...@dev.null> writes
That was one of the drives I was thinking about.
> The exact year I can't nail down.
The product manual is dated June 1990
http://www.seagate.com/ww/v/index.jsp?name=Wren_7&vgnextoid=085f5a802efbd010VgnVCM100000dd04090aRCRD&locale=en-US&pf=1&reqPage=Legacy#
Rod Speed
> Rod Speed <rod.sp...@gmail.com> wrote
>> Its certainly not my recollection that the drive the OP asked about was
>> the first over 1GB, even if he just got the physical size of the drive wrong.
> We fitted 1GB SCSI drives to a bunch of boxes in 1990, I think they were 5.25" full height drives.
Yeah, the Seagate ST41200N certainly was
http://www.seagate.com/ww/v/index.jsp?name=Wren_7&vgnextoid=085f5a802efbd010VgnVCM100000dd04090aRCRD&locale=en-US&pf=1&reqPage=Legacy#
Rod Speed
>>> So, while I'm not exactly sure about all the hard drives available
>>> before this specific drive came out, I strongly suspect that even
>>> with, say, twenty-four 5 1/2" platters, nobody could have built a
>>> 1 GB drive that could fit in a normal personal computer.
Turns out that the Seagate ST41200N did it with 8 platters.
>>> It only became possible for PC hard drives to have
>>> that kind of capacity after this technology was available.
Thats not right, most obviously with the Seagate ST41200N.
>> Yes, but your line about 5.25" drives is just plain wrong.
> I'm afraid that I have to admit that you're right. I've done some
> further digging, and I find that in the September 10, 1990 edition
> of InfoWorld, for example, computers are advertised with a 650
> megabyte ESDI drive as an option.
And the Seagate ST41200N was indeed a full height 5.25" drive,
June 1990 vintage too from the product manual date
http://www.seagate.com/ww/v/index.jsp?name=Wren_7&vgnextoid=085f5a802efbd010VgnVCM100000dd04090aRCRD&locale=en-US&pf=1&reqPage=Legacy#
> That does make it plausible that someone could have got to 1 GB before the Corsair came out.
Looks like it was a real tossup time wise and the Seagate does fit his original
drive physical size requirement, tho that turns out to have been wrong.
Alfred Falk
news:02714054-2bae-46f2...@b13g2000prf.googlegroups.com:
How about something in the CDC (later Seagate) Sabre series? In 1992
worked with some 2GB disks which were in the 5.25" full-height form
factor.
Lawrence Statton
> How about something in the CDC (later Seagate) Sabre series? In 1992
> worked with some 2GB disks which were in the 5.25" full-height form
> factor.
They kept making those disks for years - at one point I had fifteen
trays of six disks each of the 9G variant (ST410800xx) all in HVD
SCSI, connected to a couple of Fiberchannel controllers.
The disks filled most of two racks and the controllers filled up about
a third of a third rack.
At another point, I bought at auction an early EMC storage array -- A
pair of six-foot cabinets stuffed to the gills with the 2G variant.
Since in most of my applications, physical space was no concern, I
stuck with 5.25 inch full-height disks for as long as I could -- I
never found the smaller disks to be as physically reliable.
In twenty-plus-years of sysadminning, I've only had one 5.25" disk
crash, and it made its impending demise very apparent -- you could
hear the grinding in the thrust bearing, and finally radial runout
grew to the point that it could not reliably track the data.
I took it off line, but kept the spindle going just to see how long
until the heads met the surface - and it was another year at least.
Actually - the thing I remember MOST vividly about that one - it was a
41200N, I set up an extra SCSI bus JUST for that disk and its
replacement, and did a dd with the block-size set to the exact number
of blocks in a cylinder. "click. . . click . . . click . . . " it was
copying a whole cylinder per second -- so, the entire copy took just a
few minutes.
--L
Patrick Scheible
Next came with SCSI. Most workstations were SCSI, and high-end PCs
often used SCSI add-on cards.
-- Patrick
Rod Speed
> sc...@slp53.sl.home (Scott Lurndal) wrote
>> Rod Speed <rod.sp...@gmail.com> wrote
>>> Patrick Scheible wrote
>>>> Quadibloc <jsa...@ecn.ab.ca> wrote
>>>>> Rod Speed <rod.speed....@gmail.com> wrote
>>>>>> Quadibloc wrote
>>>>>>> That was the technology that enabled consumer hard drives to
>>>>>>> go up from, say, 85 megabytes, all the way to 540 megabytes
>>>>>> Nope, like it says that was the first drive to have them,
>>>>>> and it was a 1G= B drive, as it says.
>>>>>> That drive wasnt a consumer hard drive either, its a SCSI drive.
>>>>> Yes, it wasn't...
>>>> In that time period, higher-end consumer computers used SCSI.
>>> Only the Mac did, and that wasnt higher end.
>> Actually, my Amiga 1000 had a (aftermarket) SCSI adapter
>> and an 80MB quantum fireball (iirc) drive, circa 1989.
> Next came with SCSI.
That wasnt a consumer computer.
> Most workstations were SCSI,
Those werent consumer computers either.
> and high-end PCs often used SCSI add-on cards.
Not that often, actually, they mostly used ESDI.
Anne & Lynn Wheeler
from long ago and far away ...
Date: 01 Dec 88 17:42:54
To: wheeler
Subject: Trip Report COMDEX/FALL'88
I spent 2 days at the COMDEX/FALL 1988, primarily was interested in
the magnetic/optical disks/interface status in the industry.
1. Embedded SCSI drives:
(a) 5 1/4" SCSI magnetic disks:
There are some new announcements but really no surprises. Most venders
have disks in the range of 14-16 ms seek time, 1.875-2.4 MB/S raw disk
data rate and 8.33 ms latency. Most disks capacity are in the range
of 380MB/760MB, unformatted. These are the Maxtor 8380S/8760S class of
disks, perhaps with a little better SCSI overheads and command decode
time. No company announces Redwing-class disk (with 12 ms seek, 6 ms
latency, 1.5 ms track-to-track seek, 3MB/S raw data rate and 1GB
unformatted capacity). The venders are still working on the
higher RPM and better seek (servo) mechanism.
CAPACITY AVG SEEK/TRK-TRK LATENCY RAW DATA RATE
(MB) FMT (MS) (MS) (MB/S)
HP97540 663 17/3 7.47 2.5
CDC WREN 6 676 16/3 8.33 1.875
CDC RUNNER 330 10.7/4 8.33 1.875
(SHORT STROKE)
NEWBURY 9820 690 16/2.6 8.33 1.875
MICROPOLIS 660-890 14/3 8.33 2.5
1590's
HITACHI 661 16/4 8.33 2.4
DK515C-78
TOSHIBA 330 18/5 8.33 1.875
MK-250F
SIEMENS 660 16/5 8.43 1.875
4400
FUJITSU 660 16/4 8.33 1.875
M2263
FUJITSU 660 16/4 8.33 1.875
****
REDWING 857 12/1.5 6.02 3
MAXTOR 670 18/3 8.33 1.875
(b) 3 1/2" SCSI magnectic disks:
Most announcements of 3 1/2" disks are not in the high performance
area, like Lighting file. Seagate, Fujitsu, Miniscribe, and Toshiba
are in this area (seek time > 20 ms, trk to trk > 6 ms, capacity less
than 120 MB). Only two venders announcements have higher performance:
CAPACITY AVG SEEK/TRK-TRK LATENCY RAW DATA RATE
(MB) FMT (MS) (MS) (MB/S)
CONNER 210 19/5 8.33 1.5
CP-3200
MICROPOLIS 200 15/4 8.33 1-1.75
1770 (Zone Bit Recording)
****
LIGHTING 320 12.5/1.5 6.89 2
SUMMARY:
Most SCSI drives announced in the exhibit have a readahead buffer in
their embedded controller. Some venders have less SCSI overheads than
current Maxtor's drive (827 us), but probably still in the range of 600
us. Maxtor will use 16-bit National HPC controller in their next
generation controller. We can expect that the total SCSI overheads will
drop to the range of 200 us by then. The disk industry will follow this
direction to fix current SCSI number 1 problem, i.e. excessive bus
overheads and command decode overhead **.
Most high-performance drives announced in Comdex do not use ESDI below
SCSI interface. As a result, a much better command decode time ** than
Maxtor's 8380S/8760S (1.6 ms) is realized (400-600 us).
** Command decode overhead is the time from receiving the last
command parameter to the time the seek starts or in the case of
readahead buffer hit, from the last command parameter to the
time to locate the data in the buffer.
The Redwing and Lightning are the best breed of drives in 1989. It is
important that we can ship these files in Rel. 1 to offer better
price/performance and compete with the high-performance IPI2 disks.
2. IPI2 drives:
Fujitsu and NEC announced 3MB/S IPI-2 disks last year in 8" or 9" form
factor (avg. seek 15 ms, trk to trk 4 ms, 8.33 ms latency). CDC shows a
8" 6MB/S IPI-2 disk (1GB), priced at $8700 in a single unit. This data
rate is achieved by pairing heads and twin read/write channels that
operated in parallel. The NEC and Fujitsu will announce similar disks
in the near future perhaps with even higher data rate. The IPI-2 bus
operates (2 byte) at 10 MB/S. With parallel disk technology (pioneered
by Fujitsu), IPI-2 disks can easily go up to 10 MB/S.
There is no IPI-2 disk in 5 1/4" form factor yet. Xylogics and
Interphase each have announced a IPI-2 controller for the VME bus. It
is not a industry secret that IPI2 is the heir to SMD interface. The
conversion from SMD to IPI is a natural path for higher DASD subsystem
performance. How quickly that happens depends largely on the price of
IPI drives. The press predicts that the first conversions from SMD to
IPI2 will occur in mid-1989.
IBM is working a IPI2 disk from the Sutter technology (4.5 MB/S), priced
at $7000 in OEM quantities (2 actuators). The OEM group in San Jose
already discloses Sutter to some 20 companies, including Sun
Microsystems. Sun is working on a VME IPI2 controller (modify a
vender's controller, so that it will be difficult to be cloned by other
controller house). The information from San Jose OEM group suggests
that Sun will ship the new 20-MIPS workstation with much more powerful
bus than the VME they currently used. Sun is also looking for parallel
disks for their high-end applications.
3. Parallel disks/striping:
Most parallel disks (multiple heads within a disk) are limited to the
8"-and-above form factor. The typical users are supercomputer,
minisupercomputer and graphics/image processing. It is still a niche
market in 1988. It may change in next year as more powerful
workstations and IPI2 disks available. Apollo Computer has announced a
disk striping product by ganging 4 ESDI disks together (5 1/4") for
their 1000 series superworkstation.
In October, 1988, Micropolis scrubbed its parallel disk array. It was
not related to the technology according to the press. There was no
large enough market to sustain the program. The real reason was that
most systems makers would rather build the drive arrays themselves than
buy a finished subsystem on an OEM basis. This explains that I don't
see too many 'drawers' or 'disk arrays' products in the Camdex show.
Parallel disk (parallel heads) venders:
CAPACITY AVG SEEK/TRK-TRK LATENCY RAW DATA RATE
(MB) FMT (MS) (MS) (MB/S)
Century Data: 600 15/3.5 8.33 12.3
(5 modified SMD channels, 8")
Fujitsu Eagle 1000 16/4 8.25 18
(6 modified ESMD channels, 10.5")
IBIS 2800 16/2.5 8.33 12
(2 proprietary channels, 14", VME bus connected)
4. Optical disks
3M/Sony agreed to have a ISO standard (130 MM) read/write medium. Most
other R/W optical manufactures probably will join this standard. This
definitely has very positive impact to r/w optical disk industry. The
WORM (write once) disk industry does not have a standard now. The r/w
optical disk form factor probably will change to 86 mm in 1991.
3M has announced a 650MB (two sides) optical r/w cartridge. It is
priced at $250 in single unit. The price of one disk medium is expected
to drop to $50 or even $20 in the future, depends on the volume of the
market, but is always more expensive then a CD-ROM medium. 3M claims
that the medium is reliable. The archival and shelf life are more than
ten years. The Erase/Write/Read cycle can sustain more than one million
times.
Several venders offers the optical r/w (rewritable) disk drives. The
price tage is still high (in the range of $4000). Several venders
offers jukebox (WORM) for large on-line storage. The concept will apply
to rewritable disks soon, may be even has a mixture of different optical
disk technologies within a jukebox.
One interesting application for r/w optical disks comes from a start-up
Epoch Systems. It offered a LAN file server, called Epoch-1 Infinite
Storage Server, which is a general-purpose file server that implements a
hierarchy of solid state, magnetic disk, and optical disks.
The access time of rewritable optical disks is in the range of 50-100
ms, the best data rate is about 1MB/S for a READ and about half of that
for a WRITE. This data rate is better than a tape. In addition,
optical disk has a random-access capability. In the seminar, most
people agree that rewritable optical disks are not just a promise. The
emerging products from system hardware and application software houses
will speed up the acceptance of this technology. The price will drop
and the performance will improve. The CD-ROM probably will survive
forever. The WORM probably only serves as a niche as in medical, CD-ROM
master mask market, etc.
5. SCSI 2
I talked to NCR and Western Digital about their plans of emerging SCSI
2 chip. They both indicate they have a prototype in the lab and is
waiting for ANSI SCSI II standard. The SCSI 2 chip introduction
(command queueing, higher data rate, etc) will within one quarter when
ANSI announces the standard. The first product probably will be
one-byte SCSI II (9-10 MB/S). However, the two-byte SCSI II will be
easy to implement.
Before the SCSI II chips announcement, Emulex will announce a 6-7 MB/S
SCSI I chip in 1989.
... snip ...
--
virtualization experience starting Jan1968, online at home since Mar1970
Christian Brunschen
It seems that depends on the precise timeframe. As wikipedia summarizes as
<http://en.wikipedia.org/wiki/Enhanced_Small_Disk_Interface>,
"ESDI was popular in the mid-to-late 1980s, when SCSI and ATA were young
and immature, and ST-506 was not fast or flexible enough. ESDI could
handle data rates of 10, 15, or 20 megabits per second (as opposed to
ST-506's top speed of 7.5 megabits), and many high-end SCSI drives of the
era were actually high-end ESDI drives with SCSI bridges integrated on the
drive.
By 1990, SCSI had matured enough to handle high data rates and multiple
types of drives, and ATA was quickly overtaking ST-506 in the desktop
market. These two events made ESDI less and less important over time, and
by the mid-1990s, ESDI was no longer in common use."
So while ESDI was common for a while for high-end-drives, it was replaced
by (E)IDE/ATA (on the lower end) and SCSI (on the higher end) once those
matured and improved.
Certainly by the time I was involved in building PCs, SCSI cards were a
common addition to PCs to be able to connect SCSI hard drives, because
SCSI offered higher performance than IDE/ATA for hard drives: much more of
the processing could be offloaded to the controller, for one thing; but
all while still remaining in a reasonable price bracket and while
retaining the same form factors as IDE hard drives. Also, SCSI allowed
many more drives to be connected at the same time: even on SCSI version 1
this would allow up to 7 drives, where frequently PCs would have a single
(E)IDE bus allowing for just 2 hard drives.
Also, for external hard drives, SCSI was for a long time the only really
viable solution - which was another reason to put a SCSI card into a PC,
even if its internal hard drives were connected using (E)IDE/ATA.
SCSI was also used to connect devices such as scanners and CD-ROM drives:
the most common early ways to connect CD-ROM drives to PCs were through
either SCSI or through sound cards with bespoke interfaces, until ATAPI
entered the scene - using the ATA electrical and physical bits, but SCSI
commands and responses. The wikipedia article
<http://en.wikipedia.org/wiki/Parallel_ATA> is quite helpful.
Best wishes,
// Christian Brunschen
Rod Speed
>>>>>>> Yes, it wasn't...
>>> Next came with SCSI.
>>> Most workstations were SCSI,
Its clear we are discussing 1990.
> As wikipedia summarizes as
> <http://en.wikipedia.org/wiki/Enhanced_Small_Disk_Interface>,
> "ESDI was popular in the mid-to-late 1980s, when SCSI and ATA
> were young and immature, and ST-506 was not fast or flexible
> enough. ESDI could handle data rates of 10, 15, or 20 megabits
> per second (as opposed to ST-506's top speed of 7.5 megabits),
> and many high-end SCSI drives of the era were actually high-end
> ESDI drives with SCSI bridges integrated on the drive.
> By 1990, SCSI had matured enough to handle high data rates and
> multiple types of drives,
But wasnt used that much in 'high end comsumer computers' at that
time, let alone the 1GB drive being discussed.
> and ATA was quickly overtaking ST-506 in the desktop market.
> These two events made ESDI less and less important over time,
> and by the mid-1990s, ESDI was no longer in common use."
Thats well after the 1990 being discussed.
> So while ESDI was common for a while for high-end-drives,
> it was replaced by (E)IDE/ATA (on the lower end) and
And with consumer computers except for the Mac.
> SCSI (on the higher end) once those matured and improved.
But not that much in consumer computers in 1990 except the Mac.
> Certainly by the time I was involved in building PCs, SCSI cards were
> a common addition to PCs to be able to connect SCSI hard drives,
They were never all that common in PCs. In spades with 1GB SCSI drives in 1990.
> because SCSI offered higher performance than IDE/ATA for
> hard drives: much more of the processing could be offloaded
> to the controller, for one thing; but all while still remaining in a
> reasonable price bracket and while retaining the same form
> factors as IDE hard drives. Also, SCSI allowed many more
> drives to be connected at the same time: even on SCSI
> version 1 this would allow up to 7 drives, where frequently PCs
> would have a single (E)IDE bus allowing for just 2 hard drives.
There werent that many consumer computers with 7 1GB drives in 1990.
> Also, for external hard drives, SCSI was for a long time the only really
> viable solution - which was another reason to put a SCSI card into a
> PC, even if its internal hard drives were connected using (E)IDE/ATA.
Yes, but again, that wasnt common with CONSUMER COMPUTERS in 1990.
> SCSI was also used to connect devices such as scanners and CD-ROM
> drives: the most common early ways to connect CD-ROM drives to PCs
> were through either SCSI or through sound cards with bespoke
> interfaces, until ATAPI entered the scene - using the ATA electrical
> and physical bits, but SCSI commands and responses. The wikipedia
> article <http://en.wikipedia.org/wiki/Parallel_ATA> is quite helpful.
Sure, but thats an entirely separate issue to whether 1GB SCSI
drives were used much at all in CONSUMER COMPUTERS in 1990.
ArarghMai...@not.at.arargh.com
<jsa...@ecn.ab.ca> wrote:
>On Apr 12, 7:56 pm, "Rod Speed" <rod.speed....@gmail.com> wrote:
>
>> Yes, but your line about 5.25" drives is just plain wrong.
>
>I'm afraid that I have to admit that you're right. I've done some
>further digging, and I find that in the September 10, 1990 edition of
>InfoWorld, for example, computers are advertised with a 650 megabyte
>ESDI drive as an option. That does make it plausible that someone
>could have got to 1 GB before the Corsair came out.
I bought some 670 meg ESDI's back about then - the one I checked had
1989 date codes on the chips. Micropolis brand.
>Looking at the ads in the back, typical consumer hard drives were
>still in the range of 20 MB to 120 MB; but the Maxtor XT-8760E had 677
>MB formatted and 769 MB unformatted capacity... it was ESDI, and
>_recertified_ specimens of it sold for $1799 from one advertisement.
The first ones I got were like $2,500, new. The last ones I got
were like $500, but that was maybe 5 years later.
>
>John Savard
--
ArarghMail104 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html
To reply by email, remove the extra stuff from the reply address.
ArarghMai...@not.at.arargh.com
<jsa...@ecn.ab.ca> wrote:
<snip>
>The January 28, 1991 issue has a 1.2 GB drive from Fujitsu, and this
>may well predate the introduction of IBM's Corsair drive that year.
I have some 1.2 GB drives that have dates of 1991, and a 3.6 GB
drive dated 1993, all full height 5.25" SCSI. I wonder if they
still work?
Don't think I am going to bother to find out, though. :-)
ArarghMai...@not.at.arargh.com
If you still have them, look at the date codes on the chips on the
bottom.
Anne & Lynn Wheeler
re:
http://www.garlic.com/~lynn/2011f.html#22 First 5.25" 1GB drive?
found in old email log ... also just did search and found thread in
google usenet archive
Newsgroups: comp.periphs
Subject: Re: Need info about CDC Wren hard disks
Summary: a summary of interesting imprimis offerings
Message-ID: <10...@alice.UUCP>
Date: 16 Nov 89 07:46:10 GMT
References: <4...@cs.columbia.edu>
Distribution: usa
Organization: AT&T Bell Laboratories, Murray Hill NJ
Lines: 49
this is a brief summary of the latest offerings from imprimis.
as always, i take care but no responsibility for details; call
your imprimis rep. IN PARTICULAR, the prices are for me as an at&t
person; we get huge volume discounts, you should probably add 50-75%
to get a more available price.
the data comes from data sheets and salespeople. caveat empor.
wren vii:
the latest in a long line of disks. 5.25" 1.2GB SCSI disk.
average seek 16.5ms, 40KH MTBF. sustained 1.7MB/s.
available now as evaluation units at $3294, probable eventual cost
of ~$2500 ($2/MB).
elite:
new range of 5.25" disks (eventually replacing the wren's).
comes in SMD, IPI-2 and SCSI-2 interfaces. capacity is
1.2GB (1.5GB scsi). latency is <6ms, average seek 12ms.
sustained transfer of 3MB/s. 100KH MTBF.
smd evaluation units in jan, scsi production in apr/may
($4.5-5K).
sabre 2hp:
new version of the regular 8" sabre; 1.2GB and 50KH MTBF.
IPI-2 interface, sustained 6MB/s (twice regular sabres).
latency 8.3ms, ave seek 15ms.
these are shipping now, $7.4K.
sabre 2500:
2.5GB, evaluation jan/feb, seek 13ms, MTBF 100KH,
3MB/s, $8K.
arraymaster 9058:
(base for imprimis's raid).
this controller ($15K, at beta sites now) connects to drives
(any kind, speed) via IPI-2 and connects to a host via IPI-3.
assuming fast drives like sabre 2hp, host data rates are
25MB/s peak, 22MB/s sustained. imprimis will be selling a couple
of packages based on this controller; a small pseduo-disk of
5GB, 20MB/s sustained transfer, and a larger disk 16GB, with
two 18MB/s sustained i/o ports. both these packages have a lot
of internal error correction, a mean time to data loss of ~114yrs.
P.S. i note in passing that the WREN V and WREN VI were plagued with early firmw
problems regarding bus timeouts on long I/O transfers. these have been fixed
(my drives were fixed under warranty) and new drives should be okay.
but be wary of older drives.
Anne & Lynn Wheeler
Date: Fri, 21 Sep 90 11:46:19 EST
From: wheeler
Subject: 18Sep90 Commercial Analysis NewsNotes
MICROSCIENCE INTERNATIONAL CORP BUYS SIEMENS DRIVE LINES
MICROSCIENCE has entered the high-capacity end of the 5.25-inch disk
drive market by acquiring manufacturing and marketing rights for the
Megafile product line from SIEMENS AG of Germany. Siemens had
abandoned its 5.25-inch OEM drive business in late May, citing price
competition and high operating costs as reasons. However, even before
that decision was made, Siemens was negotiating for Microscience to
build its 777-MByte and 1.2-GByte Megafile drives under an OEM
contract, setting the stage for this agreement. The deal includes
most of the manufacturing line equipment Siemens used to build the
files in Germany, which Microscience will relocate to their facilities
in Taiwan. Remaining parts inventory is said to be minimal and
Microscience does not intend to hire any of the Siemens employees who
worked on the project. However, Siemens is expected to work closely
with Microscience on product development through June of next year.
"We consider this to be an important part of the deal", said Kevin
Nagle, Microscience president and chief executive.
Microscience expects to ship 777-MByte and 1.2-GByte drives in volume
by December. Although prices were not available at this time, Siemens
had priced the drives between $1800 and $2000 in OEM quantities. A
1.6-GByte drive is currently in development with evaluation units
planned for December. This drive is expected to feature a 13-ms seek
time and come with either a SCSI or ESDI interface.
(Electronic News 9/3/90, p. 19)
OEM DASD
COMMENT: Most of the large OEM contracts for the 760-MByte level have
been made, but the 1.2- and 1.6-GByte contracts are largely still
open. If Microscience can maintain the Siemens quality and, with
Taiwanese manufacturing, reduce the cost sufficiently to become price
competitive, they will have a chance at some of the design-in
contracts. IDC estimates that unit shipments of 5.25-inch drives of
greater than one gigabyte will increase to about 45,000 by 1993 from
approximately 20,000 this year.
Christian Brunschen
[ ... snippage ... ]
>> In that time period, higher-end consumer computers used SCSI.
>
>Only the Mac did, and that wasnt higher end.
The Macintosh II (1987) and its successor the Macintosh IIx (1990) were
definitely 'higher-end' personal computers. They were the first ones that
could display full 24-bit RGB without aftermarket upgrades, and supported
multiple monitors in different sizes. And they used SCSI.
[ ... more snippage ... ]
Rod Speed
>>> In that time period, higher-end consumer computers used SCSI.
>> Only the Mac did, and that wasnt higher end.
> The Macintosh II (1987) and its successor the Macintosh IIx (1990)
> were definitely 'higher-end' personal computers.
Nope, they were just the Mac approach to consumer computers.
> They were the first ones that could display full 24-bit RGB without
> aftermarket upgrades, and supported multiple monitors in different sizes.
Thats just feature creep seen with all consumer computers.
> And they used SCSI.
Yes, that was never in dispute.
What was in dispute was whether Macs were higher end consumer computers.
Ahem A Rivet's Shot
"Rod Speed" <rod.sp...@gmail.com> wrote:
> What was in dispute was whether Macs were higher end consumer computers.
They were certainly among the most expensive consumer computers,
which is probably as good a definition of high end as any.
Rod Speed
>> What was in dispute was whether Macs were higher end consumer computers.
> They were certainly among the most expensive consumer computers,
That just makes them overpriced, not higher end performance wise.
> which is probably as good a definition of high end as any.
No its not. With something like consumer computers, its performance that matters, not price.
Quadibloc
> No its not. With something like consumer computers, its performance that matters, not price.
Well, the Macintosh offers more _quality_ as some perceive it.
Eventually they did performance too; thus, they had a dual-processor
G5 Macintosh Pro...
John Savard
Rod Speed
> Rod Speed <rod.speed....@gmail.com> wrote
>>>> What was in dispute was whether Macs were higher end consumer computers.
>>> They were certainly among the most expensive consumer computers,
>> That just makes them overpriced, not higher end performance wise.
>>> which is probably as good a definition of high end as any.
>> No its not. With something like consumer computers,
>> its performance that matters, not price.
> Well, the Macintosh offers more _quality_ as some perceive it.
Sure, but thats a separate matter to whether Macs of the era we
were discussing, 1990, were higher end consumer computers or not.
> Eventually they did performance too; thus,
> they had a dual-processor G5 Macintosh Pro...
Sure, but long after the time we were discussing.
And to get back to what was being discussed, very
few Macs used 1GB hard drives in 1990 anyway.
Roland Hutchinson
What's this "they" jazz -- that's _my_ Linux box you're talking about.
I'm posting from it now.
--
Roland Hutchinson
He calls himself "the Garden State's leading violist da gamba,"
... comparable to being ruler of an exceptionally small duchy.
--Newark (NJ) Star Ledger ( http://tinyurl.com/RolandIsNJ )
Tim Shoppa
> Hi,
>
> Was wondering what 5.25" hard drive broke the 1GB barrier? Is it possible
> that it was IBM's 0663 Corsair?
>
Just going over higher-capacity commodity drives from the late 80's/
early 90's that I'm familiar with:
Maxtor XT-3380 (380 Mbyte 5.25" ESDI) was like 1986/1987.
HP 97536 (380 Mbyte 5.25" SCSI) was 1986.
Fujitsu had SMD 8" drives in 1987/1988.
Seagate Sabre-series 8" drives (SMD and IPI) were 1 Gbyte in 1988 or
so.
Seagate Wren 7 ST41200N was 1 Gbyte SCSI 5.25" in 1990.
HP C2235 (400 Mbyte 3.5" SCSI) was 1990/1991.
HP 97560 (1.4 Gbyte 5.25" SCSI) was 1990.
Hitachi had some Gbyte range ESDI drives in 1990/1991 too.
I'm not sure who was first, but the HP's and Seagate Wren's were
widely deployed at the 1Gbyte range in 1990/1991.
All these hit the surplus market en masse in mid-90's.
Tim Shoppa
The Hitachi 5.25" ESDI details are slowly coming back to me now...
DK515 was 1989 and 760 Mbyte formatted (may have been 1 Gbyte
"unformatted"), DK516 was 1.2 Gbyte formatted and 1990/1991. I was
very highly enamored of the Hitachi ESDI drives for a good while in
the 90's. Micropolis made similar units with similar prices.
Tim.
Anne & Lynn Wheeler
and some '91 history ... from long ago and far away
Date: 14 Feb 91 09:03:07 GMT
To: wheeler
Subject: Fujitsu Announces 2 GB, Ave. Positioning 11ms, 5.25" DASD
On February 14, 1991, Fujitsu announced and began OEM marketing four
models of M2652 series DASD, a new high-capacity 2GB, 5.25" DASD family.
This is the highest capacity available on a 5.25" devices from any
vendor.
The average access time is 11 msec. The average latency is 5.6 msec
(5,600 RPM). This was achieved by using thin film heads, thin film
disks and a light-weight actuator. The data transfer rates are 4.758
MB/sec (IPI-2), 10 MB/sec (SCSI Synchronous, W/256KB buffer) and 3
MB/sec (SCSI Asynchronous).
The M2652 will be shipped from June 1991. Samples will be priced at
1,000 KYen. Fujitsu plans to sell 500,000 units of these four models
(total) over three years.
Fujitsu 5.25-Inch HDD Specifications
+-----------------------+-----------------------------+
| Model | M2652 Series |
|-----------------------+-----------------------------|
| Sample Price (Yen) | 1,000 KYen |
| (dollar @ 130Y/$) | $7,692 |
| Interface | IPI-2, SCSI |
|-----------------------+-----------------------------|
| Capacity | 2 GB |
| Data Rate | 4.758 MB/sec |
| - SCSI Synch(256KB BF)| 10 MB/sec |
| - SCSI Asynch | 3 MB/sec |
| Avg Positioning Time | 11 msec |
| RPM | 5,400 |
| Latency | 5.6 ms |
|-----------------------+-----------------------------|
| Bytes per Track | 52,864 |
| Cylinders | 1,893 |
| No of Disks | 12 |
| No of Heads R/W + SR | 20 + 1 |
| Dimension(IPI, SCSI-D)| 146mm x 220mm x 83mm |
| - WxDxH (Others) | 146mm x 203mm x 83mm |
| Weight | 4 Kg |
| Power | +12V, +5V |
+-----------------------+-----------------------------+
... snip ...
for random other piece of info ...
Date: 19 Mar 91 17:59:57 GMT
To: wheeler
Subject: PC OPERATING SYSTEMS, 1990
According to 3/15 NY Times "Business Day" (Source: Dataquest)
1990 unit shipments of personal computer operating systems in U.S.
were as follows:
SYSTEM QTY
MS-DOS 14,021,000 (75%?)
Windows 2,039,000
MacIntosh 1,946,000
UNIX 399,000
OS/2 300,000 (1.6%?)
... snip ...
a little more disk:
Date: 07 Aug 91 15:33:44 GMT
To: wheeler
Subject: Toshiba 3.5" 1-Gigabyte drive (Price=$2000)
"Toshiba begins US Output of one-Gigabyte Disk Drive: (Price = $2000)
"Toshiba Corp. said it has begun US Production of the first 3.5"
disk drives with memory capacity of one gigabyte, or roughly one
billion characters of information.
... snip ...
and then there is
Date: 03 Sep 91 07:38:55 GMT
To: wheeler
Subject: Glass coated ceramic disks from KYOCERA, Japan
KYOCERA DEVELOPS GLASS COATED CERAMIC DISKS
-------------------------------------------
According to Nikkei Sangyo Shimbun (one of industrial newspaper
published by NIKKEI), 30th of August, Kyocera announced glass coated
ceramic substrates for hard disks.
This new substrate makes possible to record 60MB on 2.5 inch disk,
since less roughness than aluminum/glass substrates are now being
marketed. Sumitomo Tokushu Kinzoku Co. (Sumitomo Special Metal Co.),
and others are now developing the ceramic substrates also, however
Kyocera is the first manufacturer who made the sample disks ready to
ship.
Kyocera will ship these ceramic disks to the HDD manufacturer of
U.S. for evaluation, mass production will be started based upon the
evaluation results.
(|Yuki Note| No date reported when scheduled to start)
The construction of substrate is coated 25 micron thickness of glass
on both surfaces, and total thickness of disk is 0.635 mm (0.025
inches).
The mechanical strength of Alumina-Ceramic is four(4) times of
Aluminum or Glass disks, and easy to be polished. This unbreakable
disks would also contribute to reduce the defective at the assembly
processes. Ready to ship samples at 1.8 and 3.5 inches size. The
price of sample is 2,000 yen to 3,000 yen ($14.81 to $22.22 US, 135
yen/Dollar), and it'll be approx. 1,000 yen ($7.41 for piece) beyond
starting of mass-production.
With Aluminum substrates existing, 0.1 micron flying height and 40
MB of storage capacity at 2.5 inches disk, however with this new
substrate 0.05 micron flying height would be possible. At the
present, Aluminum substrate is the majority of HDD. And, Asahi Glass
Co., and HOYA are producing the glass substrates.
Rob Doyle
> Patrick Scheible wrote
>>> Rod Speed<rod.speed....@gmail.com> wrote
>>>> Quadibloc wrote
>
>>>>> That was the technology that enabled consumer hard drives to
>>>>> go up from, say, 85 megabytes, all the way to 540 megabytes
>
>>>> Nope, like it says that was the first drive to have them,
>>>> and it was a 1G= B drive, as it says.
>
>>>> That drive wasnt a consumer hard drive either, its a SCSI drive.
>
>>> Yes, it wasn't...
>
>
> Only the Mac did, and that wasnt higher end.
>
>
> Yes.
>
>> and many high-end PCs did too.
>
> Nope, few of them did. They mostly used ESDI.
>
SCSI for Linux also. I finally figured out what SCSI disk
controller Linus used and all my disk problems magically
stopped. ESDI support was iffy at best. BIOSs were buggy.
Rob.
Rod Speed
> Rod Speed wrote
>> Patrick Scheible wrote
>>> Quadibloc<jsa...@ecn.ab.ca> wrote
>>>> Rod Speed<rod.speed....@gmail.com> wrote
>>>>> Quadibloc wrote
>>>>>> That was the technology that enabled consumer hard drives to
>>>>>> go up from, say, 85 megabytes, all the way to 540 megabytes
>>>>> Nope, like it says that was the first drive to have them,
>>>>> and it was a 1GB drive, as it says.
>>>>> That drive wasnt a consumer hard drive either, its a SCSI drive.
>>>> Yes, it wasn't...
>>> In that time period, higher-end consumer computers used SCSI.
>> Only the Mac did, and that wasnt higher end.
>>> Macs all used SCSI,
>> Yes.
>>> and many high-end PCs did too.
>> Nope, few of them did. They mostly used ESDI.
> SCSI for Linux also.
Not much in 1990, the time being discussed.