Museum of Information Technology at Arlington

[Gil Carrick]

A reply to a post in another forum asked me to post information about our museum. Most of the history and other details are on our web site. The link is in my sig.

[wepwawet]

Hi,

interesting stuff:-)

do you have more information about the Phase IV "CPU"?
Here raised the question about it to be a pre-4004 CPU and besides this being a welcome (for some) and odd (for others) thought not really somebody here KNOWS.

Do you have info about exact manufacturing time and maybe a picture of the single CPU board?

Thanks in advance, Michael

[6a6ar09a]

I would like to know that too. I recently did small research and come up with this list;

1. AL1 | Four Phase Systems, (April) 1969

2. MP944 | CADC F14 , 1970

3. 4004 | Intel, (November 15) 1971
[three more days and happy birthday dear 4004]

4. TMX 1795 | Texas Instruments, 1972

Would like to get some comments from you and... all others.

Regards

[Gil Carrick]

http://www.interesting-people.org/archives/interesting-people/200205/msg00072.html

This email explains the situation. I had seen a posting on the "For Sale" forum and responded to it.

I don't have time to add the picture to the web pages right now, but I have uploaded a picture of the CPU card with this chip. The image is here:

http://amit-tx.org/images/Four%20Phase%20IV-70%20CPU%20Card.JPG

The 3 chips on the right side are AL4 chips. This is a remask of the AL1 chip, changed to improve the yield, according to Boysel. The IV/70 system that this chip went in was a 24 bit machine. The AL1 is an 8 bit slice, so this card has 3 of them. The next 3 chips are RL (Random Logic) chips that glue things together. The 3 on the left are ROMs. The two in the middle were not on the original CPU. They were developed later and extend the instruction set with decimal arithmetic. This was designed to support COBOL and to make it faster to do byte addressing (needed by COBOL) since the system was word addressable and converting to byte addressing required a divide by 3 - a relatively slow operation.

[6a6ar09a]

Thank you very much, specially for this nice picture of the CPU card and nice explanation.

I found this e-mail during my digging over the web and I use them as one of the references. Also I found this documents;

http://www.google.com/url?sa=t&source=web&ct=res&cd=2&ved=0CAsQFjAB&url=http%3A%2F%2Fcorphist.computerhistory.org%2Fcorphist%2Fdocuments%2Fdoc-4946dbc7a541f.pdf%3FPHPSESSID%3D144228c0da006280a4c1b222c8004737&ei=zEHzSpGlA5fWmgPLm4miAw&usg=AFQjCNEU2WXRoYzu4RGlEKretwaPslf0Kg&sig2=9pbExGTHlkQM9K4sH66rbg

and

http://books.google.com/books?id=Qge1DUt7qDUC&lpg=PA258&ots=oyPLIJoxnT&dq=AL1%20chip&pg=PA253#v=onepage&q=AL1%20chip&f=false

and I am thinking... is it too much if I kindly ask you for your comment.

Best regards.

[Gil Carrick]

Well, I am hardly unbiased. I was Four Phase employee 473 (IIRC) and worked for them for about 6 years, ending up as a manager in Software Development and responsible for all the compilers and all the word processing software. After I left I had my own firm for about 10 years, consulting and writing software that ran on their hardware. When Motorola canceled the Four Phase hardware my sales went from $10K per month to $0 in two months. I had to get an honest job again. <GRIN>

Here is a book that contains another account of the Four Phase/ AL1 development history:

http://www.amazon.com/Exposing-Electronics-Artefacts-Bernard-Finn/dp/9058230562

It is well researched and has many footnotes. I will let it speak for me.

[wepwawet]

pretty cool information!
Absolutely worth it to folow up!
Hey Chris, isn't this something for your heart?

[Gil Carrick]

I hesitate to add it in this forum, but I have several of these boards available. They contain the AL-4 variation of the chip. I would probably consider selling some for an outrageous amount. They were used in a refurb facility to test the newer IV/90 systems. We could not have picked a worse time to start a new non-profit museum. We could really use some money. That or a commercial building here in Arlington, TX. If you are interested, make me an offer I can't refuse. <GRIN>

[doccybrown]

Many thanks for the picture and informations!
I`ve got two Four Phase-chips in my collection
I am not sure about what they do. They are
both big ones with 64pins, maybe you know?
91130121 I/O-12
91150052 MC-5

[Gil Carrick]

Well, as you can tell, I am a software guy, not hardware. But I can give you an idea off the top of my head. The I/O-12 chip was involved in the architecture of the I/O system. Today we might call it an ASIC. I am a bit more vague about the MC-5 chip, but I am fairly certain it was involved in the Master Clock circuit. I have some circuit diagrams and can probably dig out the sheets where these chips were used. You can email me through the museum site if you are interested and I can (eventually) find those sheets and copy them for you. I could also take a picture of the boards they go in.

[wepwawet]

As we are close to the everlasting question about "What is a CPU and therefor what was the first CPU and we are pretty nice linking actually here is a link to some old posts I did related to the invention of the phrase/acronym "CPU" (esp. the last post on that page - it goes back to 1955.

http://www.cpu-world.com/forum/viewtopic.php?t=6560

Absolutely fascinating though here comes something up that seriously might compete the 4004 (as a first day thought without digging too deep;-)

[wepwawet]

[doccybrown]

He already told Micha, read carefully!
Looks like you are more excited than me
"The two in the middle were not on the original CPU. They were developed later and extend the instruction set with decimal arithmetic."

[wepwawet]

[Gil Carrick]

Interestingly, the slots for those chips were on the original CPU card, but the chips were not available for some time - a year or so it seems now. But this was last century, remember, the late 60s and early 70s, so I may be wrong about the time.

In general the machine was used to support multiple terminals (up to 32) running a single application like key-to-disk (keypunch emulation), word processing, IBM terminal emulation or custom applications written in assembler or COBOL.