jim@tct.uucp (Jim Kunzman) (11/10/90)
>In article <2730420B.34EB@tct.uucp> jim@tct.uucp (Jim Kunzman) writes: >>On the way to the office today a co-worker and I were discussing mass >>storage technology. The thread was prompted by the recent Byte issue >>dedicated to the same subject. I recalled the wave of hysteria brought >>about by the announcement of Digital Paper by Britain's ICL. At the time >>several U.S. companies were rumored to be interested in licensing the >>technology. I believe IOmega was one such company. I fully expected to >>be hearing some product announcements, but I haven't even read any >>articles about Digital Paper for over a year. >> >>Does anyone have any information about what is happening in the digital paper >>arena? Is it a dead product? >> > >Digital paper is alive and well, and living in the development labs. >Most people are now calling it digital optical tape, since many people >seem to think that "digital paper" is actually *made* of paper. In >reality, it's made using the same base material as conventional >magnetic tape. The estimated media life is 20 years. > As I understood it at the time the word "paper" came from the concept that it could be cut and used on hard drives, tape spools, credit cards, etc. >You've got the company name almost right -- it's ICI Imagedata, not >ICL. ICI stands for Imperial Chmical Industries, a British chemical >firm. ICI Imagedata is one of their subsidiaries. Dow Chemical is >also making the tape. Thanks for refreshing my RAM, or I should say "Error Correcting" my RAM. >Products, on the other hand, are mostly in the "development" stages. >Here's what I've been able to find out in the last few months of >hunting around. > >- Creo Systems in Burnaby, B.C. introduced the first digital optical > tape system a few months ago. The CREO 1003 stores one terabyte > (1,000 gigabytes) on a single 800 meter by 35 mm (12-inch reel) > tape. It runs at 3MB/sec SCSI and has an average access time to any > block on the tape of 28 seconds. The drive is about the size of a > refrigerator. Unfortunately, they want $250,000 (yes, $0.25 > million) for the drive.... a little steep for most sites. Creo can > be reached at (604) 437-6879; the contact's name is Lou Misshula. > When I talked to him a few months ago, they didn't have any plans > for making a smaller, more affordable system. > If my calculations are correct, this would imply the capability to store approximately 220 MB on one surface of a 3 1/2" hard disk, minus a few for the spindle, formatting, etc. The figures would be near 450 MB for a 5 1/4" hard disk. Of course these figures go up for two sides and multiple platters. I realize that write times may be much slower than magnetic technology, but this could be partially offset by using standard drives or large RAM arrays as buffers. I would think read times and access times might approach standard magnetic drives. >- I talked to the product manager at Bernoulli about a month ago. They > were working on a digital paper version of the "Bernoulli Box". The > prototypes are basically finished, but according to him, IOmega did > some market research and discovered there was no interest in the > technology, so they have no plans to bring it to market. I have no > idea which idiot marketing research firm they hired -- obviously > they only checked the PC market place, instead of the markets where > people use disks somewhat larger than 40MB... :-) > Too bad they didn't look to the future. I can think of many areas crying out for this technology, CD-ROM mastering, audio mastering, video mastering, digital mapping, archiving of all sorts of large data bases, etc. to name only a few. I just scrapped the idea of buying a Bernoulli Box for archival storage because of the cost/MB. Their drives may be expensive, but their media aren't cheap. :-( List for 3 cartridges ($225) for a 44 MB unit means $1.75 per meg. Hard drives are rapidly approaching that cost. Given that some are now removable or at least transportable means increased competition for IOmega. I'd need to buy a lot of 44 MB cartridges to offset the exhorbitant cost of their drives. In the end I decided to go with tape/hard drives for now and wait for mass storage - the next generation. >- The most interesting product, in my opinion, is from LaserTape > Systems in Campbell, CA. They take a standard IBM 3480 tape drive > (they are using STC's and someone else's [PTSI? something like that > anyway]), rip out the magnetic head, and stick in an optical head, > and then drop in their own electronics. Some of the specs of the > system are: > > Speed: 3MB/sec SCSI sustained now > 6MB/sec SCSI sustained soon > 12MB/sec SCSI (100 Mbits/sec) sustained by Q4 1993 > > > Capacity: 50GB/cartridge now > 100GB/cartridge by Q4 1993, backward compatible > > Size: 8-inch form factor > Cartridges are about the size of old 8-track tapes > > Cost: $20,000 (est.) for the drive > $250 for a 50GB tape ($0.05/megabyte) > > LaserTape is not using the same recording mechanism as Creo. Creo > uses 32 beams (laser diodes) to write a 32-track tape. LaserTape uses > a single beam and "scans" it across the tape using an acousto-optical > mechanism which changes the "angle" of the laser beam via a ~100 MHz > RF signal. Currently they use 2 lasers to write two lines across the > tape; they will increase their speed by simply adding more emitters. > It's basically 1.5MB/sec per laser beam. > > The only problem is, the drives won't be available until mid-1992. > They're in the prototype stage right now. But using the 3480 drives > and cartridges is clever -- all those stackers, jukeboxes, automated > storage systems (5,000-tape changers), storage racks, and so on will > all work unmodified with these drives. So there's no need to wait > for someone to invent them. When I was talking to the company > president last week, he was describing some of these automated tape > libraries, once they use his 50GB cartridges, in terms of PETABYTES > (1,000 Terabytes). LaserTape Systems can be reached at (408) > 370-9064; ask for Allan Connover. > Great idea. The price may still be a little steep, but once they amortize the engineering costs they could really be competitive. Eight inch tapes would fit in nicely into the majority of todays UNIX platforms too. >That seems to be it. I called up ICI Imagedata (can't seem to find >their number right now, or I'd post it here) and asked them who was >developing systems based on the media, and the only names I got were >Creo and LaserTape. So if there's someone else out there doing it, >either ICI doesn't know about it, or won't tell. Supposedly they are >trying to market the technology in Japan; maybe we'll see something >from Fujitsu or Hitachi or something soon. > >--Dave Curry >SRI International Yes, unfortunately it seems that no one in the U.S. is really interested in developing and marketing new technology today. I was disappointed that IOmega dropped the ball on this one. They were one of the few companies that have the technical expertise and industry respect to pull this off. I used to be very impressed with the company, but they seem to have stagnated as far as developing new technologies. Perhaps though they are on to something better. :-) Thanks Dave for the info. I'm posting this reply since this is a rather low volume news group. Perhaps your information will spark others to provide additional comment. -Jim Kunzman -- Jim Kunzman at Teltronics/TCT <jim@tct.uucp>, <uunet!pdn!tct!jim> !(This space intentionally left blank.)