Re: Using REP with INx and OUTx?

Ed Beroset wrote:
robertwessel2@xxxxxxxxx wrote:
Jim Leonard wrote:

Certainly no PC HDs of the XT era were zoned (and remember that there
was no such thing as a "SCSI" HD at the time either - you had a
separate SCSI controller running MFM or ESDI drives), so there should
not be different data rates for inside and outside cylinders.

It's been almost 30 years, but I still remember it pretty well.
Actually, SCSI is older than the PC XT. The first SCSI standard was
approved in '86 or so, but it was a Shugart offering before that. I
don't recall the original name, but it was advertised in Byte in the
early 1980s.

SASI. As long as we're regurgitating bits of ancient history, before
SASI, there was another standard called SMD.

Interestingly enough, even though I don't know of any hard drives using
zones at the time, Commodore's floppy disks DID use zones. For anybody
who cares, its zones are covered in a table here:

http://www.6502.org/users/andre/petindex/drives/roms/

[ ... ]

FYI, the original XT used DMA for both floppy and hard drive access. (I
have the BIOS listing in front of me now.)

If nothing else, it's comforting to know that my memory isn't
completely gone! :-)

Another interesting bit of strangeness about the PC and XT was that they
used DMA channel 0 for DRAM refresh to save the cost of incorporating
dedicated DRAM refresh circuitry. That seems kind of silly these days,
but things were a LOT more expensive back then -- it wasn't unusual to
pay over US$200 for an 8K expansion board back in the late 1970s, and
that's unadjusted for inflation. By comparison, today my local grocery
store sells SD Card memory devices and 64M is the smallest size the
carry. It sells for around $16. Using 1977 pricing, that much memory
(had it even been available!) would have been worth around $1.6 million.

I'm not sure that's really a particularly fair comparison. On one hand,
it's true that the SD card probably has as fast (or faster) access time
than most normal memory of the day. At the same time, the SD card is
used as mass storage, which would more reasonably be compared to
something like a disk drive at that time. At the same time, the SD card
also includes a micro-controller to interface between the Flash chip
and the SD bus. This is usually a 32-bit processor running at something
like 16 MHz or so.

Altogether, what you have is much closer to a complete 1977 mainframe
computer system than to just some memory. Around that time, an IBM
System/370, Model 148 would have been a reasonable choice in this
range. IBM has a short profile of this system at:

http://www-03.ibm.com/ibm/history/exhibits/mainframe/mainframe_PP3138.html

Depending on exactly how much main memory you wanted, the price was
only a half million dollars or so -- but that was basically equivalent
to the micro-controller of the SD card. The storage is slower and
cheaper than main memory, but smaller/faster/more expensive than disk.
At that time, a rough equivalent would have been the IBM 2305 Fixed
Head Storage Unit -- though even the larger option (the Model 2) only
held 22.4 megabytes per dual-drive unit, so you'd have needed 3 of them
to get (roughly) 64 megabytes of capacity. IBM has some stuff about it
here:

http://www-03.ibm.com/ibm/history/exhibits/storage/storage_2305.html

Unfortunately, that doesn't include a price. At a wild guess, cost per
byte around 10x that of disk storage might be reasonable. The IBM 3330
disk drive cost about $80,000 for 200 megabytes. That comes to about
$400/megabyte for the disk, so at a wild guess, the 2305's would come
to about $400x10x64, or just a tad over a quarter of a million dollars.

Of course, there's also the minor detail that the 2305 wouldn't
actually work with a model 138. The closest model it would work with
was a model 155, but I can't find a price on that -- but probably
higher, since it was somewhat older.

--
Later,
Jerry.

.

Relevant Pages

  • Re: separate hard drive for scratch disks for two different programs?
    ... measures out at a higher speed than the 750GB SATA ... much faster using the middle cylinders than the SCSI ... This is a function of the way the OS deals with the disk, ... of sharing the load between multiple drives, ...
    (rec.photo.digital.slr-systems)
  • Re: Not able to boot from SCSI disk on SYM21002 but all other works
    ... It is a two channel SCSI hba with 16 bit SCSI and PCI interface. ... as I tried to boot the new system from the SCSI disk it presents only "PRESS ANY KEY TO REBOOT' and stops. ... So if any of the drives in the SCSI system could be found this would be enabled as C: ...
    (comp.periphs.scsi)
  • Re: New system using AHA 2940U2W SCSI???
    ... I would like to build a new system centered around an Asus/Intel chipset and possibly a Q6600 CPU but want to keep my my old AHA 2940u2w SCSI. ... I hate to lose my Plextor SCSI drives or my SCSI scanner at this time. ... Know of any benchmarks pitting UW SCSI2 against SATA, and possibly even Firewire in the mix?? ... The cache on the disk controller board ...
    (alt.comp.periphs.mainboard.asus)
  • Re: A single SCSI disk scanned multiple times
    ... We have a Solaris machine with a 36GB Seagate ST336607LC SCSI boot ... boot disk will be scanned 16 times! ... are the address pins, ... kind of problem more often with external drives, ...
    (comp.sys.sun.hardware)
  • Re: Ram Drive
    ... > program had to wait for data to be written to disk drive. ... Data is literally left in memory ... Hard drives do have a certain amount of cache - mine is ... I am talking about effectively creating a hard disk in memory, ...
    (microsoft.public.windowsxp.general)