Slashdot Mirror
Ted Hoff Talks About The Invention Of The Intel 4004
An AC sends us this interesting piece - "I recently came across this not-so-new interview with Ted Hoff, the inventor of the first CPU in the world - Intel4004. It's fascinating reading: the birth of the chip, the dispute over credibility, patent filing and his later life with Atari."
The Microprocessor as a concept had been around for a long time; as multiple chips on a single circuit board. The old Prime computers for on example had a plug-in board labeled "Microprocessor" way before the 4004. He just took the next logical step and packaged it into a single chip. Why he gets credit for something so trivial I can't understand.
Very informative link. It'd be nice if it were on the front page, so you can sort of compare the two stories.
There was one other store in Manhattan at around the same time. IIRc, they had set up a Dazzler-like display on a machine with its monitor pointed out at the sidewalk. Story goes that so many people were attracted to it that the police politely asked the store not to continue showing it. However, I do not trust my memory of that.
Stan went on to become the moving spirit of Computer Shopper; was a genial (and slightly subliterate) guy.
I was a midnight hacker in 1960 (on the BWEWS DIP in Colo. springs, before the COC went under the mountain), so I learned my basics extremely well. When I heard of the Homebrew Computer Club (?), it sounded very interesting, but I was in a psychological bind back then and didn't follow up on it. These guys were building their own machines *before* the 4004 came out, and the AL1 chipset was probably not available, or if it was, was 'way too expensive. Some homebrew machines might have been all-discrete.
Enby in Waltham
Yes, this is timely. *cough*
This also just in. Kennedy was shot.
For a different but nethertheless even more interesting perspective please also have a look at Mr. Shima's interview http://www.ieee.org/organizations/history_center/o ral_histories/transcripts/shima.html
He was Busicom's (the first customer) and designed the 4004 with Intel.
No...I'm having problems as well, reading it from my Win98-equipped work machine. I'm guessing it's Macintosh smart quotes that no one bothered to strip.
-lee
Dude - go read the article, it was a japaneese company.
Have you compiled your kernel today??
The early uProcessor market almost died due to
massive failures in consumer electronics such
as overproduction of hand calculators and early
computer games. A few intrepid companies like
Intel kept the faith.
Maybe if you had read the interview, you would know exactly how close to yet far away from the truth you sound :)
Your right to not believe: Americans United for Separation of Church and
Part of it at least is the fact that it was a four bit processor, if I remember right
as far as I know, they moved most of their stuff out of lubbock--they are in dallas now, I think--the buildings that they used to be in is now (partly) a southwestern bell tech support call center.
Slackware: old school feel, new school gear.
I and a buddy built a couple of MOS Technology 6502A's with a cubic architecture in 1974-75. Actually, they were alternating 6502As and 6512As (they accessed memory on opposite clock values). We put 1K of memory shared between each pair of processors, leaving 2K for the OS (executive) in each processor's memory map (8K only). We didn't have enough money for more memory or real keyboards, so we used a bunch of microswitches glued together as a hex keypad, and 8 nixie tubes for output.
My memory may be suspect, but I thought it had an eight bit accumulator. I wrote an executive for it in 1974. Biggest problem was to create a multiplier in less than a page of 256 nibbles, so I could calculate Fourier Transforms with it.
"Texas Instruments invented the integrated circuit, microprocessor and microcomputer. Being first is our tradition."
I read about TI's integrated circuit invention, hand wired components. They totally missed the point about integrated circuits. Others were advancing the technology in ways that we still use today and TI gets the patent using technology that was abandoned before it even got started. Makes sense to me.
'Intellectual Properties' are uncontrollable in the wild. To base an economy on them is just stupid.
Maybe their 4004 is screwed up?
</i>
According to thiso r/busicom/busicom.html">link</A>, the designer of the 4004 was Masatoshi Shima.<BR><BR>
o c/aoc.icons3.html">this</A>.<BR>< ;BR>
<A href="http://member.nifty.ne.jp/handheld/calculat
To see the Intel/Busicom agreement, click
<A href="http://www.busicom-corp.com/intel.html">h ere</A>.<BR><BR>
To read about the Busicom calculator, read
<A href="http://www.asiaweek.com/asiaweek/features/a
<A href="http://www.busicom-corp.com/4004g.gif">Im age</A> of the 4004.
From "view source"
<META NAME="GENERATOR" CONTENT="User-Agent: Mozilla/3.0Gold (Macintosh; U; PPC)">
Netscape did this?
"that's not encryption - it's a new perl script that I'm working on..." - from some Matrix parody
now, if only patent lawyers these days still thought like that.
have we emulated this yet ???
Early to rise and early to bed makes a male healthy and wealthy and dead.
All you classic gamers:
Good inside-look at Atari's collapse down toward the end of the interview.
Four people are credited with designing the 4004: Ted Hoff, Stan Mazor, Federico Faggin and Masatoshi Shima.
There are evidently bad feelings between Faggin and Hoff because Faggin feels he did all of the real work, and Hoff got much of the credit. Many accounts do not give Shima any credit, only giving credit to the three Intel engineers (Shima was an engineer at Busicom, a Japanese calculator company at the time, and later became an Intel engineer).
Interview with Shima (extremely interesting and detailed)
Another interview with Shima
Interview with Hoff
An e-mail from Mazor, and nice pictures of the 4004
A really nice picture of the 4004
A picture of three of the engineers (no Shima) years later
A picture of all four engineers
Federico Faggin's initials on the 4004 -- the only initials on the chip
This is his list, re-formatted in HTML:
- Shima of Busicom specified ~8 custom chips, but there was some
micro-coding for floating point operations using 16-digit primitive
operations like shift left, add, subtract, etc.
- Hoff believed there were too many complex special chips and that
the micro engine was too high a level to be practical for Intel
to design and build and proposed an alternative architecture.
- Hoff proposed a simpler architecture using 4-bit data path and
simple instructions based upon the PDP-8 model we were using at
Intel to run 'Focal' programs.
- Shima and Busicom opposed the Intel architecture at first.
When I joined in Sept 1969 I was a computer designer for
more than 3 years and just worked on Symbol at Fairchild, and
did logic design on decimal floating point unit (like calculator).
- Hoff and I did program 'snipets' to prove how we could code the
'4004' to do calculator functions. I suggested enhancements to
ISA including; JIN, FIN, and FIM, and some other minor stuff.
The 4001- 4004 architecture was refined by Hoff and Mazor thru
Dec. 1969, and by Jan 1970 Busicom agreed to our architecture.
- The highlights of the architecture are explained elsewhere but
include: on chip dynamic RAM for stack and registers, 4-bit data
path, time multiplexing 16-pins, and I/O ports integrated on
memory chips with local instruction decoding.
- Faggin joined Intel after architecture was completed. Faggin did
all logic design, circuit design, layout, production supervision,
testing, etc. Faggin also 'invented' how to do capacitor with
silicon gate which had very low gate/diffusion capacitance for
bootstrapping amplifier for dynamic logic circuits.
- Shima did all final calculator code, layout checking, logic checking
and greatly contributed to the success of project; he also changed
2 instructions BBL and ADDC.
- Intel filed patents on unique features of 4004, and the holders
are Hoff, Faggin, Mazor.
In summary split the credit as follows;I think you mean first single chip cpu : )
Destrago Z. Scudiero -Noize Incorporated -Void42
a motorized abacus?
Great Microprocessors of the Past and Present (V 12.0.2)
its a cool read.. starts with the 4004.. covers a bunch of jazz.. odd processors, all kinds of crap
"Fact: Most TI stuff is built to almost military specs: the home computer's cards were cased in cast aluminum."
Didnt the military (in the states) put a block on ordering any more of their chips after one screwed up and took out a pretty expensive plane/satellite or something?
But perhaps the worlds first microprocessor. The worlds first all solid state CPU, made from descrete logic, was built in 1960 for early fighter jet navigation. The unit had an 18 bit instruction/data path and 2Kb (thats bits!) of memory which at the time cost ~$1.00 a bit!
Actually, I was poking around the other interviews (tweak the URL to get there - someone knows good web design!) and ended up with Moore's interview (You know, Gorden E Moore? From whose business ventures damn near all of our IC technology sprung?) where he talks about the same stuff with Rob Walker.
Holy shit, how's that for a killer of DRAM patents?
Another fine clip:
Just thought you might like to see it.
--
ALL YOUR KARMA ARE BELONG TO US
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
During college (ten years ago), I met a guy who was stockpiling 4004's. He used to work for BC Tel (a telco in Canada), and figured that they'd be worth something someday. I doubted that they would be.
Slashdot: come for the pedantry, stay for the condescension.
i don't think this is the first CPU. as far as i know, i was taught some japanese companies were working at the same time on a competitive CPU. and some company selected the 4004 for its calculator or something. no this is not a fairy tale. this is how i was taught.
Banu
apparently, the transcriber initially used the abbreviation "up" for "micro"-"processor" (there being no "mu" key on the keyboard). Then, a slip of the finger during the query-replace and we get "popped microprocessor"
Anyone know if the video for the interview is available anywhere?
Perhaps I am a little late here, for which I apologise...but just last year the nobel prize(physics) fo r the invention of the integrated circuit went to...
A physicist at Texas Instruments...
I am affraid that the nobel commitee will be most distressed that they have given the prize to the wrong person...
PS... first commercially available microprocessor is about all you can say for the intel 4004... plus whatever the spin doctors want to add... sort of like that guff about ENIAC being the first computer.
have a better one
qbed
imagination is more important than knowledge --Albert Einstein-
Does that include biological CPUs?
First CPU was developed by the US Government for use in the Guidance system of their ICBM's. Yes thats right, the entire microelectronic movement was pushed forward by the US governments need for smaller more accurate weapons of mass destruction.
------
------
Post Your Inappropriate Responses!
Dude, do me a favor.....Shut the Fuck up.
Thank You
Maybe their 4004 is screwed up?
that nobody likes a smartass? ;) I mean, born in '37 and graduated from RPI in '54?
3000+ comments meta-modded. 0 mod points awarded.
Lesson for other meta-suckers: Don't believe the hype!
If I recall correctly, Intel built an early version of what became the 4004 under contract to DataPoint Corp, but the chip failed to meet performance spec's, and Datapoint walked. The Intel guys thought they had a possible winner anyway, and went ahead.
Not to knock Datapoint; they had the earliest commercial ethernet system around. Very much an innovator.
I see tons of funny characters in places of punctuation marks such as quotes and apostrophes, which detracts from the reading.
Is this some proprietary Microsoft encoding that's not available in Netscape on Linux or what? I tried switching various encodings to no avail.
After the first 8008-based personal computers within the range of a hobbyist came out, a friend and I took the train from Long Beach (N.Y.) into Manhattan, with the goal of visiting all of the computer stores that day. There were three. The biggest was in the back of Polk's hobby shop. There, a salesman toggled a program into an Altair to make the lights go back and forth (this might even have been before the Kansas City tape interface was developed, because the computer clearly didn't have any nonvolatile media). I think one of the stores was closed or out of business, and the other store had an Altair in a box and they opened the top flap just so that we could see the top of an Altair in a box - that's all the proof they had that they were actually selling a computer. So, that one in the back of Polks was the only working personal computer in a store in Manhattan that day.
Bruce
Bruce Perens.
..a beowulf cluster of these!!
On a more serious note, I really am curious; can anyone make any claims for a producing a cluster-type machine using the oldest CPU possible ?
Donte Alistair Anderson Roberts - hi son!
Karma: Chameleon
What is the state of Occam? I recall wanting a bunch of T9000s (or am I mixing up the name with the terminator?) because my amiga 500 blew at 3d rendering.
But that was a long time ago. Still, it woudl be cool to revive these old computers. Can anyone confirm or deny the rumor that Yale sold a CM-1 a few years back for $500, becaues they needed the floor space and it was a hassle getting parts?
Semiconductor mask works are protected under Title 17 of the U.S. code, the same title that contains copyright law and the DMCA. They are protected for between ten and eleven years after registration or first demonstration (e.g. at a trade show).
Patents, as usual, last for 20 years after filing.
All your hallucinogen are belong to us.
Will I retire or break 10K?
It was never patented
Amber Yuan 2k A.D
"and dear god does this website suck now." -- CmdrTaco
One of the many rather informative site on camputer history can be found here, complete with pictures, and something of a context of the start of the industry. I'm sure that a search using your favorite search engine will pop up more.
"It is a greater offense to steal men's labor, than their clothes"
Fact: BigBlockMopar works for Texas Instruments.
Ha! I wish!
Resumes are available, I'd move to Lubbock in a second.
Fire and Meat. Yummy.
Hah! I had a TI 99/4A and that giganto "military spec" expansion housing was fucking ridiculous. (I wouldn't be shocked if it was a converted minicomputer part.)
It was. TI has often said that the whole system architecture of the TI-99/4A was a scaled-down TI-990 minicomputer system.
There *are* a few problems with it, though - and they're all based on the fact that when the computer was designed and introduced in 1979, more than half of the cost of the system would have been the 16k of RAM that it had, had the RAM been a full 16 bits wide.
For cost reasons, sadly, they multiplexed the bus down to 8 bits outside the system, and put all the RAM onto the TMS9918 video processor. All RAM was then called through the video chip. Slow.
There was some cache RAM on the 16 bit bus, and if I recall correctly (been a while since I fired up Editor/Assembler on one of those things), there was also some 16 bit wide scratchpad built into the TMS9900 CPU chip.
One of the more popular recent TI-99/4A hacks has been to stick static RAM chips on top of the cache RAM, build some address decode logic, and actually move the 32K RAM expansion onto the 16 bit bus. 30% speed increase with only a few wire-wrapped connections, it's very nice.
Amazingly overbuilt. But most of the TI systems that ended up with collectors still work to this day. You can't say that about Commodore 64s, with their *lovely* aluminized cardboard RF shields and high-performance serial disk drives.
Anyway, nobody wanted military spec expansion at the twice the cost of the computer. Users wanted cheap slots and cards like Apple and IBM had.That problem was more TI's marketing department's fault.
Their idea was to saturate the market with consoles, which were built (relatively) cheaply. And then, TI was going to make their money as people lined up to buy disk systems, memory expansions, speech synthesizers, etc.
TI agressively tried to stop other companies from making hardware or software for their systems. They suppressed technical information on the system and went so far as to design a "Version 2.2" QI-console, which ignored the aftermarket cartridges that hadn't been made by TI. So, while MunchMan was a good game in its own right, you couldn't play PacMan on a V2.2 TI-99/4A - that was an Atari-made cartridge.
Unfortunately, they completely overestimated the interest that most people would have in their computers. On the other side of the coin, they underestimated things, too: as shipped, the TI-99/4A was pretty useless. It was assumed that people would program in BASIC as a hobby, but that no one would ever want to go beyond that.
It wasn't until 1981 - two years after the TI first came out - that the Editor/Assembler, MiniMemory and P-Code Pascal Development Systems came out. After all, in TI's view, no one wanted to learn a complicated programming language.
Marketing also has to be blamed for their disastrous advertising. While Vic-20 boxes were screaming "Vic-20 - the FRIENDLY computer, with COLOR and MUSIC", TI's advertising was Bill Cosby looking lonely. Of course, the TI blew the Vic-20 - and arguably the C-64 - out of the water in *every* respect, but consumers still ran to the Commies.
Towards the end, while TI-99/4As were selling in K-Mart for $99 each, it's estimated that TI was losing $50 on the sale of each one. They pulled the plug October 19, 1983.
Great machine. Terrible execution.
Fire and Meat. Yummy.
I read about TI's integrated circuit invention, hand wired components. They totally missed the point about integrated circuits.
For sure, Jack Kilby's invention was more about fitting two transistors into one package with reduced manufacturing costs, than it was about connecting the two transistors to each other.
That followed very quickly, before TI made the IC public.
As for the hand-wired, yeah, at the time, almost all transistors were what is called a "point contact" transistor. They weren't the familiar robust N and P sandwich that we know now. Back then, most transistors included at least one hand-wired connection. Logically, therefore, two transistors on one piece of silicon will require two hand-wired connections, and that's how it was. While mass-production is one of the most sacred features of our perceptions of transistors and ICs, back then, the one hand-made connection on each one wasn't considered to be a big deal: the alternative was still vacuum tubes, which often have a lot of hand-made connections. Take a close look at the inside of a vacuum tube and you might even see pencil-marks from the QC department on the plate.
Point-contact transistors basically died out in the early 1960s. (Fragile, expensive, low beta, low power capability, noisy, inconsistent, etc...)
Fire and Meat. Yummy.
Didnt the military (in the states) put a block on ordering any more of their chips after one screwed up and took out a pretty expensive plane/satellite or something?
Not that I've ever heard of. Working for a defence contractor, I've personally sold the United States Navy, Marines and Coast Guard several systems which had loads of TI parts in them. In fact, I often spec TI parts where possible, because their stuff is tough as nails.
I think it would be rather tough to do that, anyway: open up *anything* that doesn't have highly integrated chipsets, and you'll probably see an array of SN74xx chips, all with the little TI logo on them.
TI also makes ICs for a lot of other companies, too. I understand they fab for AMD, among others.
As for the cast aluminum accessory cards, take a look at this. Almost halfway down, you'll find a picture of an open "PEB". From left to right, the cards appear to be the "firehose" flex cable interface card and the 32k RAM expansion card (both in cast aluminum cases), a few empty slots, and then what appears to be a CorComp (aftermarket) RS-232 card and an unknown aftermarket diskette controller card. (You'll note that the aftermarket realized that TI was into overbuilding things.)
That's their *home* computer stuff. Cast aluminum cards. You should see their industrial electronics.
Fire and Meat. Yummy.
Here's some more hopelessly overbuilt TI stuff.
...Misty, water-colored memories...
Fire and Meat. Yummy.
I spent the last hour reading the link. It's way cool.
Brackets contain world's first nanosig, highly magnified:[.]
Very Interesting. I'd never seen the resolution of this.
If you go to the interview and search on "Hyatt" you come to the point where Hoff belittles Hyatt's patent, gives some lame implied demurrer about the design not being the implementation, blames the PTO for doing its job, and then admits that "royalties are being paid" to Hyatt, presumably by Intel and everyone else who constructs microcomputers.
It's pretty easy to believe that if Hyatt's patent had no merit, or limited scope, or even if it had a disqualifying claim, then Intel, Compaq, IBM, Dell, Apple, and all those other 8,000-ton gorillas would have fought it in court, successfully, and they would not be paying Hyatt anything nor citing his numbers on their plastic.
--Blair
"In your patent application for The Universe, it is not necessary to provide a working model."
It's been a really long time since I've seen one of those...
There was an old joke in the industry that TI had grade A engineering and grade D marketing. They had all sorts of good products that no one used, and they suffered as a result. I think a lot of their lifeblood came from their early patents, which they gladly licensed to other companies.
Of course that was the old TI. Recently they've totally recast themselves as a DSP company. They sold off product lines in most of their old markets, and focused on what they saw as an emerging market with the potential to be huge. It was a bold move, and it seems to have paid off for them. TI is doing pretty well these days, and it looks like they were right about the DSP market.
Live and learn.
--Lenny
Prototypes are not required. I believe there used to be a requirement along these lines, but it was dropped as inventions became too complex to justify this.
All that is required in a patent application is enough information to allow people "of ordinary skill in the field" to make the invention.
Doesn't applying for a patent require that the applicant be able to show some evidence that they've made progress toward using the patented techniques? Shouldn't the patent office require a prototype in order to grant such a patent? That way, TI wouldn't have been able to patent Intel's processor, because it only has the specs but no silicon. If they were able to get Intel's design docs and create first silicon before Intel did, then they would be showing evidence that they have improved Intel's ideas and might possibly have something worthy of a patent.
So whatever became of these patents filed by TI and others? I'd imagine that they'd have expired by now, and Hoff says that the royalties were minimal because Intel had strong evidence that TI had stolen their patents - but even so, the patents did remain valid, correct?
-- Imagine how much more advanced our technology would be if we had eight fingers per hand.
Easy. 4 bit bus, 4 bit architecture, and 4x4 bit registers (command, decoder, decoder control(mask) and interim). It had 45 opcodes, all hardwired into just 2,300 transistors.
I used to have one, but gave it to a museum. Now on ebay they are fetching about US$100.
the AC
Hemos is like...sci-fi fans;he thinks technology is cool, but he hasn't bothered to understand the science it's based on
It looks like TI had a habit of Patenting Intel's stuff, at least according to this guy. A lot of their design specs on the 8008 ended up showing up in Ti's patents.
Another interesting note is that Intel didn't even bother to patent the 4004 because they thought the idea of putting a computer on a chip was 'obvious', and their patent guy thought it would be a pain in the ass
Amber Yuan 2k A.D
"and dear god does this website suck now." -- CmdrTaco
4004 not found.
The CPU, per se, took up 12 slots in a 3 foot by 8 foot cabinet. . .and it drove analog devices. It was "updated" with an auxiliary computer in the late 1970's, which used the early PC hobbyist's friend, the 8-inch hard-sectored floppy.
The entire purpose of the system, was to drive a simulator for USAF B-52 Electronic Warfare Officers. Even in the 1980's, we had a heck of a time getting parts for it, and were screaming for a IC-based replacement. . . . 4004's would have increased performance several orders of magnitude...
Also, the Transputer was an early machine designed for clustering massively parallel systems. Each Transputer chip had four interfaces, which you could use to connect to neighboring machines in a big mesh, or build more hypercubish things with.
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
right here
He's the guy who bolted from Intel and started up Zilog (in a nutshell - detailed versions welcome).
Excerpt: Three weeks after that disappointment, a new run came. My hands were trembling as I loaded the 2-inch wafer into the probe station. It was late at night, and I was alone in the lab. I was praying for it to work well enough that I could find all the bugs so the next run could yield shippable devices. My excitement grew as I found various areas of the circuit working. By 3:00 a.m., I went home in a strange state of exhaustion and excitement.
try { do() || do_not(); } catch (JediException err) { yoda(err); }
Intel has often claimed that the 4004 was the first CPU chip. And it's generally accepted as fact.
However, it's not.
TI unveiled one in 1970. I can't even remember the part number because it didn't get any popularity, but itwas basically the entire CPU board from a TI minicomputer compressed onto one chip.
The patent wasn't issued until 1973.
"Texas Instruments invented the integrated circuit, microprocessor and microcomputer. Being first is our tradition."
- TI Product Manual
Fact: Texas Instruments makes more chips every day than Frito-Lay.
Fact: Texas Instruments made the first 16-bit CPU chip, too - the TMS9900. It was used in TI-99/4A home computers and Patriot guided missiles.
Fact: Most TI stuff is built to almost military specs: the home computer's cards were cased in cast aluminum.
Intel is just an annoying little upstart, and the Pentium 4 is merely the continued evolution of the 4004, which was merely a hand calculator chip.
Oh, yeah, and TI did that, too, also in 1971. Only, I'd submit that Intel didn't complete the job, the 4004 required support ICs. TI's didn't.
Fire and Meat. Yummy.