ralphc@tekcae.CAX.TEK.COM (Ralph Carpenter) (09/11/90)
An interesting reminder of my youth hangs on the wall of my cubicle: a 26" platter from a CDC 6600 disk drive that was decommissoned about 1973. I think the unit was a Bryan(t). As I remember, the heads were hydraulically driven. The story was that one job would periodically show up that put the mini-van sized unit into resonance and, unless the operator rolled that job out until some other jobs came along to randomize the head movement, the whole unit would hop until it moved off the holes in the raised floor. Everyone that sees the huge platter asks me how much data it can (could) hold. Does anyone out there know the specs of this platter and the unit it came from? I would enjoy learning more about it. Thanks Ralph Carpenter PO Box 500, Mail Sta. 19-075 Tektronix, Inc. Beaverton, OR Phone (503) 627-4004 FAX (503) 627-5339
3ksnn64@cadlab.ecn.purdue.edu (Joe Cychosz) (09/11/90)
In article <6536@tekgen.BV.TEK.COM> ralphc@tekcae.CAX.TEK.COM (Ralph Carpenter) writes: > >An interesting reminder of my youth hangs on the wall of my >cubicle: a 26" platter from a CDC 6600 disk drive that was >decommissoned about 1973. I think the unit was a Bryan(t). As I >remember, the heads were hydraulically driven. The story was >that one job would periodically show up that put the mini-van >sized unit into resonance and, unless the operator rolled that >job out until some other jobs came along to randomize the head >movement, the whole unit would hop until it moved off the holes >in the raised floor. > You are right. It was from a Bryant disk, otherwise known as the 6603 disk. CDC did make a drive that had even larger platters approximately 3'-4' in diameter. The University of Illinios use to have one on there G20 system. I remember rolling the platter around in the hallways of the EE building. They also made nice coffee tables. One such table use to be at FermiLab. Anyway, back to the Bryant: Capacity: 56Mbyte Weight: 4,300 lbs Surfaces: 32 (4*9 - 4) The unit had 4 quadrants. The quadrants were organized into a upper half and a lower half. Each half used equal and oppsite positioning to conteract arm movement. The first unit was delivered to Livermore. My friend Don Crouse who installed a good many of these drives has quite a few stories about these drives. The most notable are: 1. One drive that was on its way to France fell out of the loading bay of a 707 to the ground. Upon arivial in France, the drive worked even though it was shaped more like a parrallelogram instead of a box. The heads were wrapped and locked in there retracted positions and survived the drop. 2. While installing a 6600 system in Mexico, one of the Bryants fell through the floor. The Bryant was the last of the big disk technology. Shortly after it 14" technology arived on the scene, which lasted up until a few years ago (yes I know there are some 14" drives still in production, but the majority are much smaller now). If you get some of that magnetic partical material that they use to evaluate magnetic tape media and spray it on a portion of your disk you should be able to read the bits. Joe Bytes/lb is left as an exercise of the reader.
shj@ultra.com (Steve Jay) (09/12/90)
In <1990Sep11.034940.16212@ecn.purdue.edu> 3ksnn64@cadlab.ecn.purdue.edu (Joe Cychosz) writes: >The unit had 4 quadrants. The quadrants were organized into a upper >half and a lower half. Each half used equal and oppsite positioning >to conteract arm movement. I think this is confusing the 6603 with the slightly later 6638. The 6603 had just one positioner which moved all the heads in the same direction at the same time. I seem to remember the term "quadrants" associated with the 6638. The rest of the info on the 6603 matches my recollection. The 6603's claim to fame was its very fast (by 1966 standards) transfer rate: 1.4 Million 12 bit words a second (I think). It did this by reading/writing 12 heads in parallel. Average access time: ~200 milliseconds, average latency: 32 milliseconds (900 RPM), so you had to wait a while before you saw the high transfer rate. To this day, every time I write a line of code which will cause a disk access, I can hear the CLUNK, CLUNK of the 6603 as it moved the heads. >My friend Don Crouse who installed a good many of these drives has quite >a few stories about these drives. Don would come to our site (University of Arizona), and work on his Super-P diagnostic system while the local engineers took the disk apart. Where is Don these days? I haven't seen him in about 20 years. One of my favorite 6603 stories is the time the engineers left a hydraulic hose loose while they went to lunch, and came back to find the entire contents of the hydraulic system (5 or 10 gallons) dumped under our false floor. The folks in the offices in the basement wondered about the yellow stuff leaking onto their desks. We just missed emulsifying the entire building when the expanding puddle of fluid stopped short of a very large air conditioning duct. Ah, they don't make 'em like they used to. Steve Jay shj@ultra.com ...ames!ultra!shj Ultra Network Technologies / 101 Dagget Drive / San Jose, CA 95134 / USA (408) 922-0100 x130 "Home of the 1 Gigabit/Second network"
rrr@u02.svl.cdc.com (Rich Ragan) (09/12/90)
My Kronos 2.0 instant says: 6603 and 6603-Mod-1 Disk Files Equipment type: DA Sectors/track: 64 in outer zone 50 in inner zone Tracks/device: 2048 Words/device : 7,471,104 (60 bit words) Max Data rate: 61.1K wds/sec outer zone 48.5K wds/sec inner zone Transfer rate: 1.4 usec/word (12 bits) outer : 1.8 usec/word (12 bits) inner I don't have a 6603 platter but my recollection is that they were more like 1 meter in diameter and the 6638 platter could have been around 26". I have some 6638 platters at home so I might remember to check their size. There is a story (possibly apocryphal) that at some site the heads crashed on a 6603 and the motors kept turning. The friction is supposed to have ignited a fire on the disk platters which (story has it) contained aluminum and magnesium. Supposedly the result was a computer room equivalent of a meltdown. -- Richard R. Ragan (408) 496-4340 Control Data Corp., Silicon Valley Operations
jitze@u02.svl.cdc.com (Jitze Couperus) (09/13/90)
I believe Steve Jay is right - the 6603 disk was bigger than 26 inches - I thought I heard 'em referred to as 1 metre disks... I first saw that particular peripheral attached to an RCA301 computer (early solid state - contemporaneous with IBM 1401) and this was in turn derived from a specialised military application referred to as "mobidic" which consisted of aforementioned RCA301 with a Bryant Excello mounted in back of a 5 ton truck - the name mobidic derives from "mobile disk drive". RCA got the contract and made a civilian version with a Cobol compiler which is where I came in... -- Jitze Couperus Control Data - Silicon Valley Operations jitze@u02.svl.cdc.com Voice (408) 496-4334 FAX (408) 496-4106
steele@ohstpy.mps.ohio-state.edu (09/13/90)
This is my personal favorite 6603 story. I wasn't there when it happened, though. By the time I got there a year or so later, it had been replaced with a 6638. (Every time I smell alcohol in the air, I think of the hours and hours spent by C.E.'s buffing those platters!) As most of you (some of you? ANY of you?) will remember, the 6603 platters were mounted with the axis of rotation horizontal, in a narrow cabinet with doors on the front and back. Well, to the carefully observant, there was ONE difference between the front and back doors: one had a metal pocket built into it for the manuals! As the story has it, an (unnamed) C.E., had just finished with a repair of the 6603, and was replacing the doors (with it running, of course), and got them transposed. When the door with the manual pocket was closed (on the wrong end), the edges of the platters made contact with it, causing a rather obvious failure of the unit (screeching, smoke, etc.!). A rather long period of down-time supposedly followed... :-) And yes, Steve, you're right. They MOST CERTAINLY don't make 'em like they used to! - - - - - - - - - - - - - - - - - - - - - - - - - Gregory (Greg) E. Steele (Internet):ges+@osu.edu Network Engineer The Ohio State University (BITNET): STEELE@OHSTPY Instruction & Research Computer Center 1971 Neil Avenue - Rm 406 Voice: 614-292-4843 Columbus, OHio 43210-1210 FAX: 614-292-7081
3ksnn64@cadlab.ecn.purdue.edu (Joe Cychosz) (09/14/90)
In article <1990Sep11.211524.2956@ultra.com> shj@ultra.com (Steve Jay) writes: >In <1990Sep11.034940.16212@ecn.purdue.edu> 3ksnn64@cadlab.ecn.purdue.edu (Joe Cychosz) writes: > >>The unit had 4 quadrants. The quadrants were organized into a upper >>half and a lower half. Each half used equal and oppsite positioning >>to conteract arm movement. > >I think this is confusing the 6603 with the slightly later 6638. The 6603 >had just one positioner which moved all the heads in the same direction >at the same time. I seem to remember the term "quadrants" associated >with the 6638. The rest of the info on the 6603 matches my recollection. Yep! A few more update facts and corrections. As time goes by, the numbers are all just a blur. 6603, had 39"x1/2" magnisium platters. In each half there were 7 platters (1 clock and 6 data/12 sides) rotating vertically. A later drive, the 808 had 72 26" alluminum platters. We had some of these at Illinois also and I have a platter out of this. This drive like the later 6638 was devided into quadrants. I believe there were about 10 of these drives made. 6638, don't know what size platters, was 800 Mbit drive and had 4x18 platters. The disks spun at 1200 rpm and unlike the 808 used equal and oppisite positioning as noted above. > >The 6603's claim to fame was its very fast (by 1966 standards) transfer >rate: 1.4 Million 12 bit words a second (I think). It did this by >reading/writing 12 heads in parallel. Average access time: ~200 >milliseconds, average latency: 32 milliseconds (900 RPM), so you had >to wait a while before you saw the high transfer rate.
rrr@u02.svl.cdc.com (Rich Ragan) (09/14/90)
I measured my 6638 platter last night. It is indeed 26" in diameter. Therefore, the original poster must have a platter from a 6638 (808) or the related 821 (serial). I heard a 6603 story (possibly apocryphal) that at some site the heads crashed and the motors kept on turning. Supposedly the magnesium platters soon ignited and the equivalent of "meltdown" occurred in the computer room. 6603 Tech specs from my Kronos 2.0 instant: Equipment Type: DA Sectors/track : 64 outer zone : 50 inner zone Tracks/device: 2048 Words/device : 7,471,104 (60 bit words) Max data rate: 61.1K wds/sec. outer 48.5K wds/sec inner -- Richard R. Ragan (408) 496-4340 Control Data Corp., Silicon Valley Operations
rchrd@well.sf.ca.us (Richard Friedman) (09/15/90)
shj@ultra.com (Steve Jay) writes: >One of my favorite 6603 stories is the time the engineers left a hydraulic >hose loose while they went to lunch, and came back to find ... UH OH HERE COMES THE NOSTALGIA AGAIN! I DONT THINK I'M READY FOR ANOTHER WAVE. CDC's been dead a long time now... dont you think its time to bury the corpse?? -- /\=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=/\ \/Richard Friedman (415)540-5216 | rchrd@well.sf.ca.us \/ /\Pacific-Sierra Rsrch (Berkeley) | or well!rchrd@apple.com /\ \/=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=\/
3ksnn64@cadlab.ecn.purdue.edu (Joe Cychosz) (09/15/90)
In article <25833@shamash.cdc.com> rrr@u02.svl.cdc.com (Rich Ragan) writes: >I heard a 6603 story (possibly apocryphal) that at some site >the heads crashed and the motors kept on turning. Supposedly >the magnesium platters soon ignited and the equivalent of >"meltdown" occurred in the computer room. Livermore threatend to ignite there disk after it was decomissioned. I don't know if they ever did. Joe