Intel debuted its groundbreaking 4004 on November 15th, 1971

OBW0549

Joined Mar 2, 2015
3,566
Fond memories. The very first micro I ever worked with was the 4004's big brother, the 4040. I learned a lot from that little beast.
 
I can't find any mention of the 4004 price, to convert to today's dollars.
Computing in the 1970's was very very expensive for all of us that go back that far.
 

JohnInTX

Joined Jun 26, 2012
4,787
Memories for me too, my first micro. Does anyone remember the great software/hardware turf wars that the little 4004 started?

When the microprocessor started to get traction and be marketed it was, broadly, of interest to two distinct groups. The software/minicomputer guys who had dabbled with the hardware add-ons to their HP3000s and PDP-8 and 11 machines. They saw the micro as a way to expand their turf into hardware engineering. The hardware guys (like me) who were adept at wire-wrapping big systems with TTL and fusible link PROMS saw the 4004 as just another 16 pin DIP to add to their box of tools. It was slow, expensive and *wow* hard to understand that datasheet not being computer guys. Magazines like EDN would have splashy graphics showing the CPU's internals with things that a hardware guy would understand - stepping switches for the program counter and the like. It didn't help much. In our company, the turf wars got heated in meetings with managers from both sides vying for control of various projects.

Intel wanted to sell the 4004 as a general purpose computer - and also sell everyone a $10K Intellec development system, assemblers, high level languages (PL/M) and lots of memory chips which at the time was their main business. But to counter that, a couple of engineers in Monterey, CA Ed Lee and Matt Biewer formed Pro-Log corporation which preached a simpler approach to the software. They taught programming by hand-coding, the human as assembler. You would write the program in tables on paper, translate the opcodes and poke in operands like register and goto addresses. You edited the program cut and paste (literally!) and block copy i.e. Xerox your programming sheet, cut and paste new code on the appropriate line then copy again. Insane by today's standards but we got a lot of useful work done that way and by the time we moved on to assemblers and such, we had a deep understanding of the underlying architecture.

Aside: we also found out that the Xerox copier did not copy exactly 1:1 so after a few edit cycles, nothing on the programming sheet lined up any more. Fun.

The resistance to the micro went on longer than you'd expect looking back from today. A buddy and I went to a video game seminar (think PONG, Asteroids etc). given by the designers of such wonders. Maybe 5 years after the 4004 intro, most poo-poohed the micro - too slow, too expensive, not needed etc. Only the last guy, from RAMTEK, demurred. During his presentation of their 8008-based game he described the challenges and advantages they'd encountered in its development but one thing he said sticks out to this day - they made changes with code, the others made changes with yellow wire-wrap wire. By that time, I'd gone through MILES of yellow wire-wrap wire. If we could do changes with a keyboard and terminal.... That sold it for us. My buddy had heard that Mike Quinn Electronics at the Oakland Airport had cosmetic 8008s for $50. 'We NEED one!' So we got one and forced it to work. We eventually wrote a self-hosted assembler for the thing. It would emit the object code on paper tape from an ASR33 that the PDP-11 would read. By that time, our PDP buddies had a 1702A programmer board on the UNIBUS and we all struck an accord, drank lots of beer while the ASR did its 10cps thing and fancied ourselves just something else :cool:.

Still, guys like Ed and Matt were heroes to us lowly hardware guys. My start was taking the course described at the end of the catalog. They also offered clip-on analyzers which took advantage of debug info that the 4004/4040 put on the bus as well as a bunch of LEDs for address, data, stops etc. Just like those PDP guys with their fancy console!

EDIT: note that they call their microprocessor boards "Logic Processors". Turf war terms for us hardware guys.. Comment added after nsaspook's like..
 

Attachments

Last edited:

OBW0549

Joined Mar 2, 2015
3,566
I vividly remember everything John mentioned above, including hand-coding on "assembler worksheets" and punching in hex code on the Prolog box.

Good times.
 

sparky 1

Joined Nov 3, 2018
756
The Intel company put a team together and made it happen for the 4004 CPU chip set that was used in personal computers.
As a start up company, not the biggest at that time landed an order for chip set. A trail with many paths. Eventually
a Japanese business calculator company collaborated in order to improve it's calculator. The customer relationship grew with
Busicom, Bob Noyce of Fairchild and Tadashi Sasaki of Sharp who understood the code and hardware to get the board to work.

I recall for years many magazine articles on the latest garage made computers. The Pro-Log was an eprom programmer.
You could'nt just buy a computer you needed to assemble and code it from scratch, debugging and crashing was common.
The personal computer like the Kenbak-1 was very basic. The PDP-8, TRS-80 builders wanted more robust system.
The floppy disk did not have enough through-put by itself. Today it is cloud.

link: is a picture of the Busicom board that used that chip set from Intel from what I understand.
http://www.vintagecalculators.com/html/busicom_141-pf_and_intel_4004.html
 
Last edited:

WBahn

Joined Mar 31, 2012
29,976
I can't find any mention of the 4004 price, to convert to today's dollars.
Computing in the 1970's was very very expensive for all of us that go back that far.
I looked into this some time back and, in current dollars, the 4004 was about $375. A small sampling of chips and prices over time showed that processors have tended to stay in that $400 to $800 price range, including several of the entry-level AMD Epyc-Rome versions. I suspect this reflects a marketing reality that a CPU can sell for roughly so many hours of labor and if it's much over that then sales are going to plummet.
 

DickCappels

Joined Aug 21, 2008
10,153
From my perspective it was Chuck Peddle at MOS Technology that. with their MCS6500 series processors gave a 16 bit address space and 8 bit datapath for only $25 in single unit quantity, made integrated CPU's turn the corner toward processors for lower cost products. Had it up and running in a few days, while the 8008 build was never finished.
 

MrChips

Joined Oct 2, 2009
30,708
By that time I was favouring the MC6809. My build included graphics CRT screen and 8” floppy drive. The next build was MC68008. I still have both.
 

ZCochran98

Joined Jul 24, 2018
303
The 4004 is 27 years older than I am. I find it quite impressive how far computing and microprocessor technology has come since then (especially with modern ARM microcontrollers). The first processor for which I learned assembly was the 68HC12S chip from NXP (or Freescale) - still old, in processor years, but much "newer" than the 4004. I did have to learn some basics of the 8048 recently for a (not-yet-complete) project (redesigning and modernizing an 80s digital-analog hybrid synthesizer). That was an adventure....
Those of you who worked with these old processors have a great deal of respect from me - the amount you could accomplish and do with, comparatively-speaking, much less storage and flexibility and smaller word sizes is astounding.
 
4004 (0V,-15V), 4040 (0V,-15V), 8008 (+5V,-9V) were PMOS, quite slow clock speeds and awkward power supply voltages. That was also a cost driver. 8080A is NMOS but still needed +5V, -5V, +12V.

Next generation CPU's are NMOS single +5V supply 6502 6800 Z-80 etc. - much cheaper and faster, so that IC technology really allowed things to take off.

The killer app for the early CPU IC's - calculators then pinball and later, video games. But I remember a 4004 on a card for a PDP-8 I think, so it might have been a support CPU.

For the 4004 price $60 (in today's dollars of $375) plus clock gen and support 2KB ROM 4001, I/O+register RAM 4002, SIO 4003...
Median US income in 1971 was $9,030 so I can't see anything affordable to a hobbyist until the late 1970's with the Altair and IMSAI.
 

DickCappels

Joined Aug 21, 2008
10,153
It was when my chemistry teacher realized that I was taking notes from an 8008 brochure during a lecture that he pulled me into his office and scolded me that I should be concentrating on the importance of chemistry in electronics. At that time, I was just starting realize how to use a microprocessor to actually do things.
 

ErnieM

Joined Apr 24, 2011
8,377
My first micro was about '77 when I got a SBC kit based on the 1802. Came with a hex keypad you could bang code into the 256 bytes of RAM. Used it in a comp sci final project but I forget what that was. Sadly I pitched it some 20-30 years later to make room as one of the chips had died.
 

MrChips

Joined Oct 2, 2009
30,708
My first microprocessor project was hand coding an RCA1802 in machine code to convert a Micro Switch wordprocessing keyboard and send esc codes to CP/M Wordstar.
 
Yep, the RCA 1802. I got the 1802 kit from Popular Electronics or radio Electronics. You could disconnect the keyboard when your done. I did a crane for a model railroad. Hand coded. Later I did it in BASIC using a Vitrax 9 SBC.

The first routines I wrote were to emulate gosub and return. I think I can find my paper copies of those.

Prior to that I had written programs for the PDP-8 and PDP-11 in assembly and FOCAL and BASIC respectively.
 
Top