Slashdot Mirror
Intel 4004 Turns 30
fm6 writes: "Just the thing to remind an aging geek of his mortality: this week marks the 30th anniversary of the Intel 4004, the very first microprocessor. Another historical page here, and a column bemoaning the absence of dancing in the streets here. Trivia -- why 4004? Because it was the fourth component in a 4-bit chipset." You might want to read the interview with Ted Hoff from a few months ago, it's pretty informative about the origins of the 4004.
The best way to celebrate would be the release of a new athlon ;)
"...you can steal my woman, but you ain't done nuthin' smart."
Is it still in production anywhere and what's the current record for overclocking one of these babies?
A feeling of having made the same mistake before: Deja Foobar
Unless you can port linux to this, why do we care? This is slashdot, we have standards!
Trapped in Time... Surrounded by Evil... Low on Gas.
...pretty soon, it will start experiencing its midlife crisis.
Maybe some day, when I get tired of making small electronic curcuits explode, I'll get one of these and build an SAP (simple as posssible) computer out of one, just for jollies, assuming I still have eyes left.
A feeling of having made the same mistake before: Deja Foobar
4004 Not found.
--
The Cap is nigh. Time to get a fresh new account.
"this week marks the 30th anniversary of the Intel 4004, the very first microprocessor. "
What?
I thought Microsoft made the first microprocessor after purchasing the idea from Al Gore.
But, well, if they say so on Slashdot, it MUST be true.
-Shaunak.
Perhaps no overclocking and Linux, but -vice versa- there exists a 4004 software emulator for Linux (e.g., i4004em).
Every day is the 5nth anniversary of .055% of everything that ever happened, and as a rule we celebrate very little of it, or it would occupy all our time.
Do we really think the 4004 might be offended by the oversight, or that microprocessors in general aren't getting enough attention in the press? I think the computer industry as a whole could be modded down a point as it is.
The 4004 was certainly a significant milestone, but I think the 8080 launched in 1974 was truly the "Model T" of the computer industry. That was the chip that was general enough to really run everything. It was the basis for all the microcomputers and the CP/M operating system.
In fact, I believe Zilog Z80s (an 8080 clone with some extra instructions -- around 1977?) are still being manufactured as controllers in various products.
Sometimes it's best to just let stupid people be stupid.
The 8008 was twice as powerful as the 4004.
If only naming conventions could make that much sense today . . .
Can anyone name the first true single-chip microprocessor? It has to have integrated RAM, ROM and I/O.
I remember stopping by the Intel booth at the National Computer
Convention in New York in 1971-1973 timeframe (can't remember exact
date).
My Dad had put me on a train to New York to expand my teenage
horizons. I returned with 4004 and 8008 data sheets and some chip
samples. I spent the next few months dreaming up what I was going to
do with the chips and drawing schematics.
I never did build anything with them, because owning a terminal and a
modem was more important to me at that time than a having a uP - if I
had had my priorities straight, I might be famous now [grin]. I did end up
designing and building 3 different video terminals, though.
Thanks for the memories.
-Rick
I had an Odyssey video console for my old B&W TV back in the day. It had a membrane style keyboard, which you could program assembly code in, and it ran on the 4004.
I want to delete my account but Slashdot doesn't allow it.
about the 4004 development, right here - they were Intel's customer at the time.
try { do() || do_not(); } catch (JediException err) { yoda(err); }
... the first microprocessor was older than UNIX
It seems a bit strange for me to think that first unix didn't run on a machine with microprosessor.
According to The Chronology of Personal Computers (1969-1971):
The first production run of the 4004 was in December 1970. Admittably the production run had to be tossed due to mask errors, but 2nd and 3rd production runs in Jan and Feb of 71 were more sucessful (the 2nd run still had errors). Sample calculator designs were shipped to Busicom in March 71 - comprising 4 4001s, 2 4002s, 2 4003s and 1 4001.
The only relevance of November 71 that I can find, was that the MCS-4 microcomputer based on the 400x series was released. But thats not the microprocessor itself.
One thing that stands out, is that Intel have had production problems and bugs since day 1 :)
HAS ANYONE PORTED LINUX TO IT YET?
4 bit data registers, 640 bytes (yes, bytes) of addressable memory - I think I can safely say "no".
<obligatory_MS_bash>
But I hear Win 3.11 is coming along nicely. Thrashes like a bitch, though.
</obligatory_MS_bash>
According to this site, the first personal computer was Simon, c. 1950, a relay and paper-tape affair. You can argue with their definitions, but it has a lot of interesting historical machines.
MITS Altair really started the PC revolution, in that it was readily available, had a decent amount of compute power, and was affordable.CS and other hobbyists get all sniffly for the good old days. I was having a huge amount of fun a couple weeks ago hacking a 6502. That we even recall such an occasion should suggest to you that so long as some of us are alive, remembering and playing around with such artifacts defines who we are.
A feeling of having made the same mistake before: Deja Foobar
It is lesser known because the designer, Ray Holt, only received clearance to publish information about it in 1998.
I recall the 6502 instruction set as being mingy when compared to the Z80 (for instance); I regarded that as a handicap at the time (this was 1981 or so) but it was probably a good thing for a beginning assembly hobbyist to cut his teeth on.
:)
This is true. The 6502 was a competitor to Intel's 8080. The latter was much more powerful (and the later Z80 by Zilog was like an 8080 on steroids), but the former was cheaper... IIRC something like $400 for the 8080 and $250 for the 6502 (and these are mid-70s dollars). Also the support circuitry for the 6502 was simpler, further reducing cost.
Pardon if I got some facts wrong, age and alcohol have taken a good number of brain cells
I'm sure i'll miss a few but here goes,
Sorry the formatting is poor due to the lameness filter.
4004
4040
8008 8080 Z80 (Zilog) Z8000 (16-bit)
8086 8085 Z800 (Z80 extension)
80186
80286
386SX also IA468 (still born new archi)
386DX
486SX
486DX
486DX-2
486DX-4
Pentium, AMD K5, 586 (cyrix)
P-MMX P-PRO K6 686 Win chip
P-2 Celeron K6-2 686MX Win Chip II
P-III Cel(2) K6-3 ?
Coppermine Athlon Cyrix III
T-bird
P4 Tualatin Athlon XP
I've missed out the Xeons, and of course all the
microprocessors that didn't have some lineage
to the orignal 4004. Although the instruction
sets changed a lot particular from the 4004 to
8080 and from the 8080 to 8086, there is enough
similarity in there style and content to claim
that your Pentium 4 or Athlon XP is directly
descended from the 4004. It makes you wonder
if Intel can really expect to shift people from
the x86 arch to a totally new one.
Still utterly unused, in anti-static foam, three Intel 4004s. My roommate decided to start collecting old CPUs, and I managed to find these, free. I still want to make a very simple blinking-lights toy with one of these, and proudly put the "Intel Inside" sticker on the box :)
:)
Goddess, this brings back memories! Hanging out at the library, using their terminal to call (at 300 baud, that was *fast!*) the HP-2000 system at Harper College, and chatting with friends who had serious money (Jeff actually *built* an Imsai 8080 unit, though he got many of the parts free by schmoozing the sales person).
30 years, gads. Back then, having even a floppy disk was a wild dream, now we have 100+ gigabyte hard disks. Back then, having one whole K of ram was heaven - last week, I bought 512 meg for $20. Back then, the clock oscillator could be made from a simple L-C circuit, and it ran several hundred kilohertz. Now, it's a PLL-controlled internal oscillator, using an external crystal oscillator, all running at frequencies that make a microwave oven look slow.
All this, in thirty years. That *really* makes me feel old
Lemon curry?
Doesn't the modern Intel Pentium 4 run just about as fast as this one too?
:)
With XP
As I recall we had a Model-33 Teletype for software development. We punched the program into paper tape, called up a system using an acoustic modem and used their cross-assembler. Or maybe I'm just having an antacid flashback.
Wouldn't it be more appropriate to make a bigger deal when this thing is four thousand four years old? Or perhaps, when it turns 44?
I've been waiting for the year 2015 to be the first poster with the story. I was really looking forward to all the extra karma gained with the mod ups.
Dang, foiled again.
Here is a link that has a simple graph from the 4004 to the P7 (Merced Pentium II) that shows how Intel has obeyed Moore's law (at least until the P2.)
Jesse Wolfe Sr. Manager Systems Integration
Intel 4004 Turns 30
:P
from the middle-aged dept.
I'm 33 and that ain't middle-aged. I take offence!
/*drunk.. fix later*/
Intel summed up the "speed" [sic] difference between a 4004 and a Pentium 4 with an interesting assume-this-basketball-represents-the-sun-like analogy: "Intel's first microprocessor, the 4004, ran at 108 kilohertz (108,000 hertz), compared to the Pentium® 4 processor's initial speed of 1.5 gigahertz (1.5 billion hertz). If automobile speed had increased similarly over the same period, you could now drive from San Francisco to New York in about 13 seconds."
There's a Dutch auction over at eBay with 100 of these little babies at $10 a piece. Just search for "Intel 4004". The date code on the picture shows that they were made in 1975, so they're not the ceramic and gold ones. Auction ends Thursday evening. Don't outbid me, or I'll mod you down.
But imagine a beowulf cl...
.13 micron technology, making them a tiny fraction of their original size ? Throw down a hundred of them on a board and have it run a massively parallelized app of some sort at 25 cents per node.
Hey, seriously, wouldn't it be groovalicious to have a bunch of 4004's produced using today's
Why the hell not ?
-Billco, Fnarg.com
Of course, it is best balanced with Mr. Hoff's interview, as they seem to have different ideas on how much everyone contributed, the language and technical communication barriers were definately there.
Bleh!
Still, the Z80 now is nothing to what it might have been. I remember when Z80/CPM systems were the standard for serious desktop computing. Even Apple II people used a Z80 card to run business apps. (There was a MS version -- their first hardware product!) I think this Apple/Zilog combo was the most common desktop business computer at one time. If things had gone just a little differently...
I liked the 6502 instructions. There was an elegance to them, a symmetry that seemed missing in the 8080 and its ilk. When in doubt, Intel threw registers and special instructions at the problem (never mind the Z80's two complete register banks), whereas MOS seemed to favour soft solutions (don't need no stinking "multiply"), and of course memory-mapped I/O.
This next song is very sad. Please clap along. -- Robin Zander
But Hoff did accomplish something important -- a lot more important than inventing a particular gadget. He demonstrated that simple general purpose computers could be built that could replace a lot of the complex custom hardware that was then being built. In so doing, Hoff started us down the road to making computers ubiquitous.
Well, yes, of course you could. I didn't mean to suggest that you couldn't, just that those IN and OUT instructions represented special cases, and the 6502 philosophy seemed to be against special cases.
This next song is very sad. Please clap along. -- Robin Zander
I *LIKED* the Z80's alternate register set. It made interrupt handling much easier!
Fascism starts when the efficiency of the government becomes more important than the rights of the people.
If we take Moore's law and extrapolate. . . 33 years is 396 months, so technology must be 2^(396/18) = 4,194,304 times more advanced now than it was then. Did you guys even have fire yet, or were you still confined to nice warm Africa?
I think we need to call up Guiness (aside: isn't it strange that a record-tracking group also makes beer?) and update the records. This is a major archealogical find.
-- Imagine how much more advanced our technology would be if we had eight fingers per hand.
The 6502's initial excitement was the price -- it was the first real $20 microprocessor. The 8080 cost quite a bit more in 1976. The 6502 was, I would suggest, a far more elegant chip than the 8080. The 6502 was Federico Faggin's second microprocessor. He did the Motorola 6800, then jumped ship to MOS Technology. The 6800 was loosely inspired by the PDP-11, a very nice machine for the day. The 6502 rethought a few of the 6800's design decisions and was even nicer. Apple picked the right chip.
BTW the original 6501 was pin compatible with the 6800, but didn't ship because Moto made some threats. The 6502, I think, added an internal clock; there were a bunch of other 65xx chips that followed. Rockwell was the second source and kept the line going after MOS Tech, which had been bought by Commodore, tanked. I think it's still fairly popular in the embedded world.
The 8080's instruction set was more reminiscent of the PDP-8, and rather ugly. The 4004 was not really a microprocessor, just a controller, because it lacked interrupts and some other features needed to be a "computer". The 8008 was a major improvement. And only 18 pins -- a lot of jelly beans needed to run it! The 8085 was the easier one to glue to. I recently read an article explaining the history of those early Intels, but I don't recall where.
At the time I thought this was Big Stuff. Unix on a microcomputer! But in hindsight, you really need a VM to have a serious modern OS. That's why there's no Linux for the 80286! I think the first commercial Unix to do this was CTIX, which Convergent Technologies created (System III/V, with the BSD VM, running on a 68010 with proprietary memory management hardware) for its MegaFrame box. The MegaFrame was a disaster (tried to be too many things at once), but it paved the way for the first 680x0 Unix boxes -- Convergent's main claim to fame before they were absorbed by Unisys.
Anyway, a conveyor belt dropped bottles from a wheel going around (a horizontal disc) onto straight rows of pins, also moving. Required some trigonometry and timing, especially when starting the machines up. It was controlled by a 4004, the code lived in 7 256-byte uv eprom dip chips.
We had an assembler written in Fortran, it ran on either a Honeywell 1648 or a Dec PDP-10 (both notable machines in ARPANET/Internet history). When I got there, they used to type the hex assembler output into the prom burner by hand! Burning the 7 proms took 18 hours of person time, and was error-prone. I wrote some code to do the eprom download automatically, with a paper tape or something, cut the process down to an hour and a half, made some folks pretty happy.
My memory of it is REALLY fuzzy. I had a lot of projects on the burners, and it was just one hunk of hardware among many that were floating around. It was a curiosity because it was from IBM and a prototype, which is the reason I remember it at all.
That said, yes, it was around 1983/84. I do remember that it was rather bulky, perhaps like a cash register, but I don't recall any membrane keys, sensor inputs, etc. It's entirely possible that they took a unit like that and created the prototype from it? On the other hand, I might just not remember the membrane keys.
Sometimes it's best to just let stupid people be stupid.
That's a very good question. On the one hand, they were quite far apart on the Unix family tree. On the other hand, I don't recall anybody abandoning their V7 ports and starting over with System III, when AT&T decided that the latter was the official commercial Unix. Probably everybody just folded in System III features. I'm guessing that V7 backward compatibility was never an issue, though I'm hardly the expert.