Sunday, December 27, 2015

First PC Video Standard

Microsoft in March, 1987 was very much a work in progress creating a multimedia strategy.   Bill Gates joined in a standing ovation for Art Kaiman and a team from RCA  as they demonstrated the first video standard for PC called Digital Video Interactive at a second annual CD-ROM conference.  The team's demo was on a "full height" CD-ROM, using a 6-MHz Intel 286, IBM AT running MS-DOS.
According to John Dvorak, Gates was annoyed by being frozen out of the release date of the first CDi products.  Kaiman told the press, rather diplomatically,  "I can't say this is a CDi killer."  Eventually, the product Intel called Actionmedia went on to win Best of Show at Comdex 1991 on both Windows and OS/2.
In the 1940's RCA introduced the famous RCA jack as a way of connecting a phonograph to their radios.  In 1942 RCA opened what was to become the RCA Sarnoff Labs named after company and NBC founder David Sarnoff - inventor of color TV, VCR and even the first CMOS CPU, COSMAC for the U.S. Space program.  Kaiman was Director of Digital Products for the lab.
dvi_team87
 Kaiman (closest to bottom) before Intel bought DVI and moved the team to the new Princeton Operation.  
collectionkaiman
 Art Kaiman's contribution lives on in the MMX instruction sets on billions of IA-32 compatible CPU's unofficially named for MultiMedia eXtension, Multiple Math eXtension, or Matrix Math eXtension.
  • Mr. Kaiman's diplomas: Rutgers - The State University Newark College of Arts & Sciences, BA, 1963 and Stevens Institute of Technology, Master of Sciences (Mathematics), 1966
  • There are 10 artifacts in this collection:  All were acquired as a group from Kaiman's estate.
      1. RCA Laboratories Outstanding Achievement Award, 1973, Arthur Kaiman, For contributions to a team effort in the design, development, and application of new commercial minicomputer software, signed Wm Webster, VP, RCA Laboratories.  Plate and RCA badge mounted in a mottled brass frame, 11-5/8" x 11-1/8", custom framed by A Gallery, Salt Lake City, Utah, good condition but the plate is slightly crooked
      2. RCA Laboratories Outstanding Achievement Award, 1974, Arthur Kaiman, For contributions to a team effort in the development of a multiprocessor software system for television broadcast automation, signed Wm Webster, VP, RCA Laboratories.  Plate and RCA badge mounted in a mottled brass frame, 11-5/8" x 11-1/8", custom framed by A Gallery, Salt Lake City, Utah, good condition
      3. RCA Technology Symposium, This certificate is awarded to Arthur Kaiman for planning, organizing and conducting the Computers in Mfg. Symposium.  Signed Wm. J. Underwood, Director, Engineering Professional Programs and H.K. Jenny, Manager, Engineering Information.  Acrylic plaque with black background, 6" x 4-1/2", good condition with light scratches.
      4. RCA Certificate of Appreciation, This certificate is awarded to Arthur Kaiman for the colloquium presentation Manufacturing and Productivity - Do It Right the First Time !!, May 11, 1982, Colloquium Committee, David Sarnoff Research Center.  Acrylic plaque with black background, 5-1/2" x 3-1/2", good condition with light scratches.
      5. Ceramic Coaster: Printed on front is a mock newspaper article with the headline: Intel Corp. Breaks Billion Dollar Barrier (this coaster is in the Computer History Museum - View).  Intel's revenue first passed $1 billion in 1983.  4-1/4" square with cork backing, good condition.
      6. RCA Certificate of Appreciation, This certificate is awarded to Arthur Kaiman for the colloquium presentation Digital Video Interactive Technology, March 31, 1987, Colloquium Committee, David Sarnoff Research Center.  Acrylic plaque with black background, 5-1/2" x 3-1/2", good condition with light scratches.
      7. Intel Corporation's 1989 Microsoft CD-ROM Conference Team Award presented to Art Kaiman.  Intel recognizes your outstanding contribution on the 1989 Microsoft Conference.  It was the most successful ever.  Signed Dave House, Bob Brannon, and Rick Stauffer.  Framed certificate, 11-5/8" x 9-1/8" overall, good condition.
      8. Intel Princeton Operation, Intel A82750PA and Intel A82750DA, PRO 750 ADP i750 F.C.S., July 21, 1989.  2 computer chips suspended in an acrylic block, 5" x 3-1/2" x 1-3/8", light scratches and the left chip is slightly crooked in the block.  This item is pictured in Manifest Technology's DVI Technology Products Intel PRO750 Product Family (1989) online page (view).  This item commemorates the first customer shipment of this product.
      9. Intel PRO DVI Technology, Happy Holidays, 1990.  Clear glass mug, etched, 5-3/4" high, good condition.
      10. Intel AVK, In recognition of your contribution to the development of AVK and the ActionMedia II Products, Art Kaiman.  Clock, black finished brass face with etched to brass numerals, logo, and pictorial representation of the product; separate black finished brass plaque with silver decoration, etched to brass; black felt background, deep brown wood frame, 12" x 18" x 2", missing the minute hand and cap, it also needs cleaning, good otherwise.  This clock commemorates the introduction of AVK and ActionMedia II as part of the DVI project between about 1990 and 1992.  CLICK HERE to see details.

  • DVI team image from Manifest Technology.
 
rcatdvi
In the same month as the DVI Demo for Gates, Kaiman gets an award for DVI merit.
 rcasoftware
Stamped signature of Dr. William Webster.  Webster joined RCA in 1946 after getting a degree from Union College, next door to the G.E. Homestead in Schenectady, N.Y.   He went on to be a thorn in his neighbor's side holding patents in television, vacuum tubes, gas tubes, circuitry, and semiconductor devices.
IMG_3525
Personally Signed by Mr. "Intel Inside" Dave House.  Rick Stauffer, marketing manager at PRO, was loosing the standards battle to Quicktime and MPEG by 1992.
i750chiplucite
First i750 chip-sets "W #12".  These packages look empty but the MMX they inspired packs a lasting punch.
IMG_3524
Time ticks down for Actionmedia II ostensibly because of high costs for both consuming and producing DVI video .
 
proswag
Intel SWAG!
In 1993 the Intel  Princeton Operation (PRO) was finally closed relocating some but ending the role of Art Kaiman .  Although DVI was ultimately not successful it paved the way for other formats like AVI that have made the PC a multimedia powerhouse.
Collector's Insight
IBM PS/2 Ultimedia Model 57 SLC and 77s are based on DVI technology.  These originally included the $2K Actionmedia II board. They can be identified by front audio jacks and an early PS/2 CD-ROM drive.  Also look for DVI and Ultimedia logos.

Tuesday, November 17, 2015

1-Bit Datapoint Logic Board Revealed

“Faced with the facts that the logic design of the 8008 was made by Datapoint and its initial chip implementation was covered by a Texas Instrument patent application, Intel conferred on 4004 the status of the first microprocessor.”  – Dan Alroy  chairman of the 1975 IEEE conference.


The very first x86, Intel compatible logic board used a Datapoint 1201 chipset not made by Intel.   This is the TTL implementation done for the world’s first microcomputer the Datapoint 2200. It has four early production Intel 3101 memory chips that provide an interconnect for simple shift register memory also supplied by Intel in large quantities.

Recently arrived in the lab is a Datapoint CPU board with 6947 date code on four blue memory slots each flanked by an Intel 3101 static memory chip – Intel’s first product from 1969.   Fleetwood Mac was just climbing the charts when these babies were made.




photo(1)w3

Datapoint 1-Bit, Intel Compatible, Rev A. Logic Board


In the first year of Intel’s operation memory chips were a priority over microprocessor design.   The 3101 you see in the picture is Intel’s very first product having gotten done eight months after the company’s founding.  Intel then contracted with Computer Terminal Corp. to design a microprocessor called the 1201.

Two ex-NASA engineers at Computer Terminal Corp were fresh off a success with their 3300 terminal and wanted something with an “improved control unit” according to their business plan.

Enter Intel and Texas Instruments pitted in a battle to develop the world’s first 8-bit microprocessor.   Both built the 1201 CPU for the Texans but in one of the biggest business blunders in history CTC did not secure the rights to the CPU for a mere fifty thousand dollars.  They could have easily owned Intel at that point.

Instead the company rushed out our board in the pictures using TTL (transistor to transistor logic).   It was actually quite fast.  But thanks to the design’s inability to access memory directly through the blue “69” slots we have this stunning  gold 3101 four-banger.   The first x86 compatible in history.

Intel sent its main salesperson out into the field in the summer of 1971.  He reported back that there was some tepid interest in the 1201 which Intel renamed 8008 to fit with the entirely incompatible 4004.

IMG_3307stunner

Datapoint technology evolved very quickly going through four major archetecture iterations between 1969 and 1974.   3300 base (3360), Datapoint 1 (1-bit Serial), Datapoint 2 (8-bit Parallel) and 5500 (16-bit).  With the exception of 3300, Intel stumbled through these same iterations after a delay in 1976 with i432 (8086 was an emergency replacement).

raytheon

Intel's CPU voltages continued to be hamstrung by Datapoint's original 15V design creating a major opportunity for ex-Intel designer Faggin to create the Z80, a superior 8-bit design that required only 5V.  Zilog rocketed to the top of the 8-bit heap but was quickly at a dead end with the 16-bit Z8000 introduced in 1979. 

1973revisions

1973 Revisions, 2200 (H) and 5500 (Sample)

The 2200 and 5500 use the same form factor however  the 5500 uses a 16-bit architecture that is  incompatible with Intel 8088.

Victor Poor completed one of the key milestones in computer evolution over Thanksgiving weekend 1969 - 8008 logic on his living room floor!

prototype2200

Datapoint breadboard logic

The 15V leads and proprietary Datapoint connector (3300) tie it to Victor Poor's early R&D department.  The spelling of "Buss" is old school Navy electrical buss not the more modern spelling referring to a modern Data "Bus" although it is impossible to tell which engineer did this unless the handwriting is recognized.

Friday, November 6, 2015

'66 Death Star

During one of the most frigid moments of the Cold War, LBJ authorized a classified $1.5 billion, military orbiting laboratory called Project-MOL.  It could photograph ICBM sites in the Soviet Union at high resolution. Douglas spearheaded intercontinental warfare from space designing a station to conduct planetary warfare on a massive scale.

Previously, only three pieces  of the cancelled project have ever been discovered. The Gemini-B, a highly modified military mock-up using a previously flown capsule and two blue space-suits found in a locked museum room at Kennedy Space Center.

NEAImage
"NEA Reference" marked on back dated 12/66.

Add to the list a fourth, custom modified, ground computer unit used to issue the spacecraft digital commands!  I call it Astro-Ray #1, packed with 56 Astrodata computer boards, and weighing in at over 100 lbs.  It is Raytheon Serial #1. Astro-dated 6/66.

Only 20 years old when the MOL test flight released three satellites and successfully returned to earth,  a moon-sized planet killer called the Death Star was written into George Lucas' first draft of  "Adventures of Luke Starkiller - The Star Wars".  Fox legal had a provision in the contract that Lucas could not use "Luke Starkiller" in any of his own movies.

From Tucker Electronics, I have a seven part installation, operating and parts manual produced by Raytheon for this one-off military system.   It contains a fascinating look at confirmed Project-MOL hardware in practical use.



RAYTHEON Model 15379 Documentation

 NAVSHIPS 0967-277-1010

Transistorized space modem



mol-remote
r-trans

Dual 10K transmitter version of modified Gemini remote #1 for MOL - added US-Band frequency and lots of redundancy.
x_open
x-open
x-exper
Experimental unit combines Astrodata (NASA data acquisition) and Raytheon (Missile type board)
use ray
mol2
Paper trail for unit is classified in giant weapons contract called N600.  N600 has progress reports on thousands of different parts from Allied to Zinc metallurgy.
x-tone
Navy radio expert told me just one 10KW transmitter weighed 3 tons.
xspec666
Astrodata command module remote is modified for Project-MOL on 6/66.   Only white hulled ship in Navy, S.S. LaSalle is retrofitted for special mission in September 1966 in Norfolk. 
xlink
Petty Officer 3rd Class on La Salle to photograph mission  identified Astrodata cards and frequency counters as project MOL.  Companion Astrodata countdown box using nixie tubes was also present.
xconnect

Stay tuned to this blog as I will update it as more information is uncovered.  

Sunday, October 25, 2015

Alvy Ray's Pixar Image Computer


Animator and CEO John Lasseter made an amazing dimpled design for the Image Computer that perfectly captures the transformation into Pixar the studio.  Silver diamonds and a flecked Zolatone finish distinguish the remarkable metal bezel. 
According to Alvy Ray Smith, the pioneering Pixar Image Computer had three sugar daddies.   These included George Lucas and Steve Jobs as well as a founding educator in the New York Institute of Technology.  The school was initially housed in a former Long Island garage and chauffeur's quarters.
Smith coined Pixar from a Spanish verb “to make pictures” which Jobs used upon buying the company in 1986 from Star Wars creator George Lucas.    Lucas was in a divorce battle in range of his valuable movie franchise so he readily dropped his price from $100 to $10 million for a quick sale.  Smith and his partner Edwin Catmull had a 5% ownership stake that they sought to protect as Jobs grew increasingly desperate waiting for their breakout hit - Toy Story.
closex2
MacosxlogoX1
"I have been shameless about stealing great ideas"
Alvy Ray Smith sent me an email clarifying the role of Steve Jobs in the design process.   Jobs insisted on using frog design (small f and d) but didn't fully realize the rack mount nature of the computer as a large 80's video card (16 AMD ALUs in 4 Channels) that required another computer to work.   The first version was like a death star  in design and scope selling  only a  few dozen units.  The second version, called PII, broadened its appeal, landing sales from Disney Studios and even medical imaging contracts.
The Pixar Image Computer was not purple. It was shades of gray. The outer box was dark gray, approaching black. The front of the box was Zolatone, a fleck painting process, with flecks of different shades of gray. The photo is of the product itself, not a prototype.
The role Jobs had in this was that he insisted that we use his favorite design company, Frog Designs, for the design of the box. We went along with the demand. Just part of the care and feeding of sugar daddies, as Ed (Catmull, my partner) and I used to say. Steve was the third of three sugar daddies on the way to Pixar, so we had experience in this arena!
Our engineering/manufacturing guys were furious because the design cost $150,000 as I recall and was a total luxury since nobody would see the box (it was meant to be rack mounted in a back room somewhere). This cost was totally unnecessary in other words, and just further jacked up the cost of an already expensive machine. 
The dimpled square was a John Lasseter design, and so was the font choice for the name, and so were the little diamonds between the letters. So it's not clear what Frog contributed, other than a huge bill. I suppose the suggestion of Zolatone was worth the $150K!? In retrospect it was another Jobism, show not substance, that cost us not him. - Alvy Ray Smith - June 18,2015

Monday, October 19, 2015

Armstrong's Giant Leap for Technology


As the Apollo guidance computer was landing Neil Armstrong on the moon two ex-NASA engineers were landing the PC on user's desktops.   Ex-Navy radioman Victor Poor did the logic for what was to be renamed the Intel 8008 over Thanksgiving weekend.  A pre-PC product called the Datapoint 3300 (100 times better than the model 33 Teletype) was shipping with early programmability features.
"The cursor can optionally be programmed through the computer." - Modern Data magazine review of Datapoint 3300.  The unit has two compact "friends" in a 3300P printer and 3300T cassette tape. 
Mad Men style marketing was in fashion complete with mini dress clad models looking attentively at smart bosses.  The price was also going down rapidly with the 3300 selling for $5K.  Ads  promoted the local functions as well as traditional time sharing.
The "improved control unit" got a new name in 1969 but the  microprocessor would have to survive its first spectacular failure.  A pre-PC called Viatron showed that government adapted  technology made a poor transition to low cost, mass produced computers.
Viatron System 21 - Painful Face Plant
According to its wiki Viatron Corp. gets credit for coining the term microprocessor.   On my wish list was to someday look past the Datapoint 2200 - to someday collect parts of absurdly rare and eccentric Viatron System 21XX. Thanks to some record Boston snow that day came much sooner than l expected!
The discovery of Dr. An Wang's  "7-board" personal Viatron System 21 was  by accident sitting at Applebee's Restaurant checking texts. I made contact with a former Wang employee during one of Boston's snowiest winters on record.  He claimed to have possession of a dissected Viatron given to him by Dr. An Wang.
Wang was a MA neighbor of little known Viatron Corp.  Working from an Air Force AESOP concept for small interconnected devices that would let an end user take control of a smart terminal complete with built in software anyone could understand.  According to a published roll out plan the company promised they would start handing out in Boston in September 1969 for an amazingly low rental price of $39 per month.
After scanning An Wang's autobiography for clues on how this device shaped his thinking on computers I can conclude that he initially wanted nothing to do with the word "computer" but everything to do with revolutionizing office productivity.
Viatron was up against nearly impossible odds for making this machine work.   MOS semiconductor maker AMI convinced the company president to send them research dollars.   The company also spent money to make a "Printing Robot" keeping true to the initial Air Force spec of having a typewriter as part of the system.    Wang's System 21 has the optional  printing robot!
According to our source from Wang, Viatron made only 25 working units - losing over one million dollars per unit in investor funds.  A raging lawsuit followed the ousting of the CEO over dinner at his home and ensuing corporate bankruptcy. AMI was listed in the complaint - American Microsystems International not the AMI builder of modern PC BIOS.
Our Viatron has dual tape controller boards.  Not even twins even though they have the same assembly codes.   The boards have what I consider a struggled design.  The bus function is appended on crude daughter boards.  Another illustration of the massive product problems the company faced.  Also a Viatape drive and tape cassette produced by an outside company for Viatron seems to be of a much higher quality design.
While Viatron's terminal product got done for the glossy brochures the computer option was a complete failure.
Close-up of 2-bit memory board.
IMG_1566ab
TI TMS-1723 memory chips from early 1970
Datapoint 3300 "Glass Teletype" - The Giant Leap
The 3300 is an early adopter of the inexpensive 7400-series control logic from TI introduced in 1966.   The technical reference manual is one of the few small device references to include a compete logic diagram of the system control unit.
The modern microcomputer owes its existence to this 1969 great grandaddy of tech.    Once the data communications side had been ported to a "green screen" the industry broke wide open for Intel and others to perfect and expand.
NAND gates would find their way into 8008 logic done by Victor Poor in that same year these were shipped. IBM's unofficially supported Conversational Programming System supplemented local functionality running in Model 33 emulation mode.
Our typewriter footprint 3300 is in top condition  and even lights a red toggle for "local mode".  With preliminary text handling capability desired most by TTY operators this unit can cut but not yet paste.
In a popular Modern Data rundown of interactive terminals we find "the cursor can optionally be programmed from the computer" on the 3300.  PC-like 3300 has a small impact printer and Phillips made tape cassette.