Slashdot Mirror
50 Year Old Computer Still Going
The Angry Mick writes "Geek.com is running a blurb on a 50 year old CSIRAC computer that is apparently still functional, if lurking in an Australian museum. Sporting a whopping 2K of RAM and screaming along at a blistering 300 khz(!) it proves the adage that they really don't make 'em like they used to . . ." Yes, because if they did, they'd be really, really slow.
.. how do they cool that thing down?
I fuse with Mercer every single day...
...30 more years and my Apple //e will have been running for 50 years! Woohoo!
// Agent Green (Ian / IU7 / KB1JQO)
// IEEE 802.3: All 10base Are Belong To Us
I find the idea of a massive computer lurking rather funny. Of course, it could be the 4 Guinesses I just polished off. Oh well, time for bed. I hope I don't have dreams of ENIAC or some other thing now!
_damnit_
It's my job to freeze you. -- Logan's Run
A Beowuld cluster of- oh never mind, where would you fit it?
Information doesn't want to be anthropomorphized anymore.
and linking that to yesterday's discussion about the lack of quality these days, i bet we won't have any/many of today's computers around in another 50 years time... or 50 days for some of them...
- Welcome the coming of the New World Odour
By reading the horde of nested articles, I got the impression that the machine hasn't run in decades, and probably would not if powered.
Correct me if I'm wrong. But please quote a piece that says it is actually running now.
Bot Assisted Blogging
Although I have the advantage of having a whopping 64k of ROM. I only have to use the RAM for my data. I would expect that computer also has to store the program binary in the 2k. Overlays, anyone?
Lately I've been finding it worth my time to spend a few hours recoding some functions in order to shave just a few bytes off their stack usage.
Kids these days, assuming everyone's got 128 megabytes for their application. They just don't code 'em like they used to.
Request your free CD of my piano music.
My first computer, the C64 runs at a massive 1Mhz, only about 3 times faster than this machine.
:)
Commodore released the 64 in 1982, this puts it at 20 years of age. That's 30 years between these two machines. When did Moore make that law again?
Yikes, imagine what the computer world will be like in 30 years time! Assuming MS haven't screwed it up for everyone.
Not bad for a living dinosaur. Listen to it yourself :)
It does annoy me that people , even though its in good humour , snigger at these old machines with their "paltry" 2K memory and slow speed. Yeah , sure they're not exactly a Cray. But look at what was done with this one. Skyscraper design , cloud droplet simulation, antenna design! Lets see even the best programmers used to point and drool GUI interfaces and hand holding wizards try and do that in 2K now using little more than paper tape! The people who designed, built and programmed these machines REALLY knew what they were doing and probably forgot more about efficient programming and code compression than todays "top" coders ever knew in the first place.
Is hard to imagine that this Computer was used for weather prediction. ...
Those tasks usually require large amounts of data to be processed
--
Stefan
DevCounter
An open, free & independent developer pool
created to help developers find other developers, help,
testers and new project members.
- Don't older computers than this run air traffic systems in airports?
- Is it Y2K compliant?
- And last and least, imagine a Beowolf cluster of these.
And BTW, at the link mentioned, they are questioning whether the computer is even running: "From what I read in the links the computer would definitely not work if powered, instead it would probably catch fire".Sex - Find It
god I feel old...
Years ago, when I worked at the CSIRO I worked on this machine for a while. I'm amazed it didn't die long ago. It used RPN for calculations, which takes getting used to, but is far better then algerbraic.
It's processor (not CPU - it consisted of multiple chips) is a hardware FORTH type. The jokes about FORTH programmers are true!
Just imagine how big a computational problem could be solved in 50 years with contemporary P4 hardware, if only Intel would build its hardware to last 50 years. ...Or anything over 5 years - for that matter...
Now this may not be a problem for home users that buy a complete new system every two-three years (regardless of environmental effects), but I'm sure happy they don't sent out space probes which rely on today's state-of-art.
--
The most likely way for the world to be destroyed, most experts agree, is by accident. That's where we come in: we're computer professionals, we cause accidents -- Nathaniel Borenstein
"A lot of its components would not stand having voltages applied to them again," says Thorne. "I think it would probably catch fire."
Reminds me of some of the old linux kernel code, and thinking its good to have a sense of humor.
Trying to get a printer working and getting a kernel message saying Lpt:1 on fire!
It's easier to fight for one's principles than to live up to them.
" Sadly, it's not an option to make CSIRAC operational again today. Time has taken a toll on this fragile dinosaur.
So what exactly would happen if anyone tried to relive the magic by switching it on?
"A lot of its components would not stand having voltages applied to them again," says Thorne. "I think it would probably catch fire."
300kHz may not sound like much, but with overclocking and a decent watercooling setup you could crank it as high as 334kHz!
-- Ed Avis ed@membled.com
When I'm at my wittiest, she just sort of rolls her eyes, groans, and goes back to whatever she was doing.
Request your free CD of my piano music.
The Geek.com article says:
" A half-century old computer, called CSIRAC, is still operating in Australia. The computer, which was Australia's first, ran at a blistering 300 kilohertz, had 2 KB RAM, and 2.5 KB storage."
But the Inquirer article linked by the above Geek.com article says:
"The machine was the fourth computer to be built anywhere in the world, ran at 0.001MHz, and had a massive 2000 bytes of memory and a behemothic 2500 bytes of storage."
Which, by my calcuations, would be 1000 hertz or 1 kilohertz. I tend to believe the Inquirer, since they're running the source article. And besides, the 1977 Apple ][ was only 1 MHz, Don't you think there was a bit more progress than less than doubling in processor speed from 1949 to 1977?
Information doesn't want to be anthropomorphized anymore.
My first computer (10yrs or more) was a 386 with a 5quater drive and a 100mb hdd, 640k ram and no mouse. Windows 3.1 was going to happen. Yah, it all sounds funny now, but with the recent remark that Moore's law may soon become obsolete, we may not get MUCH faster computers on current technology. I guess we'll have different architectured.... quantum computing and DNA computing seem to be hot areas. :p
A question though, was it just built 50 yrs back, or has it had an up-time of 50yrs ?????
|/________
|\A|ALYS|
2K of RAM would be very useful (I got 640 bytes). Although direct access ROM is very nice (better than talking through a serial port anyway), having extra RAM would be lovely, letting you compress stuff better, and mix data by category instead of by whether or not it changes.
It's not just that the simpler chips are more reliable, but they use less power, generate less heat, cost less and take up less space and don't weigh as much.
I have heard that the ARM chip is the most popular for embedded applications these days, and many of the ARM chips in use are quite tiny, have no cache and run in the 40 Mhz range, like the ARM7TDMI.
68000-based chips from Motorola are also very popular.
And check out uCLinux, a linux port to several microprocessors that run without a memory management unit.
Why bother with an MMU when there's no disk to swap to, and the failure of a user program would mean the failure of the whole system?
Request your free CD of my piano music.
Check out this page which tells us the history of the said computer. In the end, it says the following:
Following the University of Melbourne's purchase in 1964 of a Control Data 3200 from the USA, CSIRAC was donated to the Museum of Victoria. At this time it was realised that CSIRAC was the oldest computer still in operation, and worthy of preservation so it was carefully dismantled and stored.
CSIRAC is now the centre-piece of the IT display at the Museum in Melbourne.
As I understand it, the music was recorded by building a replica of the sound hardware and connecting it to the emulator. People who heard the music have confirmed it sounds pretty much like the original in 1955 (IIRC, it was around that time).
Perhaps the coolest thing that they did with CSIRAC was build a HLL and compiler for it, which they called Autocoder IIRC. It looked like a cross between FORTRAN and BASIC and avoided some of the thinkos of FORTRAN, as far as I could tell.
CSIRAC is now permanently on display at the museum in Melbourne, Australia. It's the only complete, original machine of its generation in existence, and well worth a look if you come down our way. There is also a book on CSIRAC called "The Last of the first", which is a fascinating read if you can get your hands on a copy.
One of my university lecturers, Peter Thorne, got his start in computers as an operator for the machine. He met his wife there - she was a fellow computer operator!
Any sufficiently advanced technology is indistinguishable from a rigged demo
--Andy Finkel (J. Klass?)
Moore's Law includes price. Did you take into account, that you might have payed less when purchasing your 1982 C64 than was spent on CSIRAC, 20 years earlier?
Btw, C64's feature 64kB which is 32 times 2kB, so at least memory size doubled five times in 20 years, that is: it doubled every four years.
--
In theory there is no difference between practice and theory. But in practice there is -- Jan L.A. van Snepscheut
The story states that it is still operational. If you follow the links, at the end of the the big write-up they ask a what would happen if someone tried to power it up. The reply was "probably catch fire".
Of course, by 1962 CSIRAC was years behind the state of the art.
Any sufficiently advanced technology is indistinguishable from a rigged demo
--Andy Finkel (J. Klass?)
running the search engine on Sourceforge, right?
It's Christmas everyday with BitTorrent.
The actual hardware is dead, but as I've said elsewhere an emulator does exist and many of the old programs have been recorded.
Any sufficiently advanced technology is indistinguishable from a rigged demo
--Andy Finkel (J. Klass?)
Who's there?
*60 second pause....*
CSIRAC!
On one occasion, they gave a demo to an organisation called the Institute of Radio Engineers (IRE), but apparently a memory error occurred and the thing printed "CSIRAC welcomes the members of the IRA) :)
Any sufficiently advanced technology is indistinguishable from a rigged demo
--Andy Finkel (J. Klass?)
For those of you who didnt read the article pointed by the parent poster before listening, that is NOT music being played by the CSIRAC itself but rather a software recreation of the original hardware, and a modern recreation of the original speaker. :)
It basically sounds like my old spectrum, only a bit worse
I wonder what all the background noise on it is, though, it certainly sounded like they had a massive computer in the room while playing it.
No, they had a Z80A running at 3.54MHz.
Stick Men
That's nonsense. They all ran at the same speed, 3.54MHz. The ZX81, the Spectrum's predecessor was slower, at 3.25MHz. I had (still have) a ZX81 and a Spectrum 128.
Stick Men
One of my neighbours helped to build CSIRAC. My guess is the computer looks better than he does.
Great old guy. His wife does a great pumpkin scone.
we have a working IBM 1130 here (and the IBM engineer that it was assigned to...) hehe
www.aconit.org
The CSIRAC was a vacuum tube based machine. From http://www.cs.mu.oz.au/csirac/design.html:
And on top of that, ICs weren't invented until 1958.
Information doesn't want to be anthropomorphized anymore.
> they really don't make [computers] like they used to
If automobiles had evolved at the same rate as computers we would all be driving Jaguars that went 250 miles an hour, got 500 miles per gallon, cost $1000, and self-destructed once a year, killing all of the occupants.
That's true. According to the book I've got, the CPU of CSIRAC was synchronized to the mercury delay lines, which completed a cycle in about 1 millisecond, so I suppose you could call the clock rate 1 kHz. Each instruction took either 2, 3, or 4 memory cycles to execute (the initial design had every instruction take 4 cycles, but an improved control unit design took advantage of cases where that wasn't necessary). Hence, the machine ran at about 500 instructions per second.
Any sufficiently advanced technology is indistinguishable from a rigged demo
--Andy Finkel (J. Klass?)
It says "typical" not "state of the art". Most desktop PCs *ARE* around 500Mhz and *DO* have only 64mb of RAM.
The article is clearly dated "Dec 10 2002" so it's not "from around a year ago" at all - no idea where you got that from.
Nick...
The C64 was a CONSUMER item. When the CSIRAC was built there was no such thing as a computer for consumers. It would be more appropriate to compare the CSIRAC to the so-called supercomputers that were availiable in 1982. Machines like the Cray X-MP and Cyber 205 were availiable in 1982. The costs to own and operate them are comparable to what it took to operate the CSIRAC in it's day.
i ca l/computers/history.html
The UK's weather bureau give specs on the Cyber 205 they were using in '82:
http://www.met-office.gov.uk/research/nwp/numer
CDC Cyber 205
200Mhz Clock
1 MegaWord of memory
The Cyber had a 64 bit word size so that amounted to 8 MB of ram. So clockspeed has increased over 600 times and memory has increased over 4000 times in that time frame. This is just confining myself to the 205. I didn't look for the specs on other large machines like the Crays that were availiable then.
Computers as something just anyone could play with were pretty much nonexistant prior to 77 (true you could build something ENIAC-like anytime in the seventies if you were REALLY good with electronics). It's more instructive to see what the kind of money they had to spend on the CSIRAC will get you as time moves forward. Power comparable to the C64 was availiabe in the early sixties for that kind of money.
Okay, some quick math:
50 years * 366 days/year (rounding up) * 24 hours/day * 60 minutes/hour * 60 seconds/minute * 300000 cycles/second = 4.74336e14 cycles
Now, my Athlon XP 1600:
4.47336e14 cycles / 1400000000 Hz / 60 sec / 60 min = Roughly 89 hours
So even if this machine were still running (which, incidentally, it's not. RTFA), in terms of pure cycles of functionality pulled out of the machine, my Athlon beat it in the first four days. It's a lot easier to maintain a pair of shoes than it is an airplane. And of course, this machine ISN'T still running, and would likely execute an HCF instruction (Halt and Catch Fire) if powered on, so you really can't call it reliable.
(Of course, my Athlon's running Windows (needed a games machine), so it's debatable whether or not these cycles have actually been functional...)
--AC
I would have thought many of them would no longer be manufactured. (Computers went solid state- discrete transistors- in the late 1950s and integrated circuits in the early 1970s.)
"Which, by my calcuations, would be 1000 hertz or 1 kilohertz. "
With a clockspeed of 1000 hertz you'd actually be able to hear the thing go "OOOOOOOUUUUUOOOOUOOUUUUOOOOOUUUOUOUOOOOOO".
Man that !has! to have sucked completely to be a developer back then: "WATCHA SAYIN'?? I GOTTA WHAT??? CHANGE THE POINTER?? I !CAN'T! !HEAR! YA!!"
naah sig schmig
Computer evolution and Windows de-evolution are separate issues.
--
If you moderate this, then your children will be next.
Here's a brief page about some ibm tube logic modules, schematic for an 'inverter', etc. Anybody with an old 650 laying around I'll gladly cart it off for you.
try { do() || do_not(); } catch (JediException err) { yoda(err); }
Uh ... since semiconductor cost is directly proportional to size, increasing the density reduces the cost for a die with the same circuits. Likewise, increasing the density also allows you to put more circuits on the same die and to run them at higher speeds, increasing performance. Moore was most defnitely talking about price and performance in his "law", he was just not spelling out every little detail for you.
> Computer evolution and Windows de-evolution
> are separate issues.
Oh, you're right, I forgot that using Linux keeps your hard drives from failing, forces everybody to perform daily backups, keeps worms and viruses from affecting your system, makes CPU fans last forever...
Actually, just looking at the monitor and reading the words off it probably takes far more than 300,000 calculations per second. Consider that OCR still doesn't work well with computers 1000 times faster than that.
Hey kids, there's only 5 days left 'til Yak Shaving Day!
A few things that should be considered along with the lifetime of current PC's
a) Heat and dissipation: They run hot as hell. Yeah, this was filled with Vacuum tubes and probably got fairly warm as wellone probably got fairly warm as well, but in modern PC's the heat tends to be focussed over particular components, leading to detioration over time.
b) Moving parts: Fast-spinning hard drives, fans (see heat, above), etc. The more moving parts you have the greater chance of failure. It also takes more power impulses to start a motor spinning up (hard drive, CD-ROM).
c) Expected time of usage: We're going through PC's a lot faster than we used to. How long was CSIRAC in use? For most home users, you can usually expect an upgrade at least every 5 years. Perhaps not a new PC, but at least a component. Why build a PC that's going to last forever if it's going to be obsolete very soon - except for consideration to servers, etc.
(You can look at it online if you want)
Sadly enough, yes.
It wasn't major, really... just a Turing machine project for a homework assignment. It calculated the function y = 2x + 1. In unary, of course.
Strangely enough, writing Turing machines didn't greatly increase my appreciation of 0s. My appreciation for having an instruction decoder, however, went through the roof.
bytesmythe
Hypocrisy is the resin that holds the plywood of society together.
-- Scott Meyer
I'm sure it would power up in the morning and clear its throat for about five hours, then go in the bathroom for two, then have salt on its oatmeal.
And all the while, there's a Sun machine thinking "Why can't you just short?! Short and be done with it!"
I'm having an episode!
Keep your packets off my GNU/Girlfriend!
Has somebody told the NetBSD folks about this one? (Read 2nd line from the top center of the link if you don't get it.)
There will be a celebration to jointly celebrate it's 50th anniversary and it's completion of calculating pi to the 4th digit.
There is something I find annoying with Slashdot, it's the bad habit of posters to leech news from other sites that already refer to a previous coverage on another site. This is absurd: I click on Slashdot's link to go to geek.com's link, which sends me to The Inquirer, from which I can finally have the real thing. Is this only me that is irritated or what? Hey, when I read the same news first on OSNews (who at least have the decency to redirect to original sources more often) and that some hours after I see that same story on Slashdot, but with the link pointing to OSNews, I find that a bit ridiculous. Not that I think it wrong to acknowledge that news posted on Slashdot came from another news aggregator (that's how one learns about the other ones), but the point is that you end up with a neverending arab telephone, and the guy down the line says black when you're posting white. Or else it's a new way to counter the slashdot effect, and I'm not just getting it.
the article mentions IBM's digital computer in america,
but doesn't mention that the first digital computer (the 'ZI') was designed in germany by: KONRAD ZUSE:
Konrad Zuse - Mark I
It's a tube computer, so it HAS to sound better then a solid state one.
Actually, just looking at the monitor and reading the words off it probably takes far more than 300,000 calculations per second. Consider that OCR still doesn't work well with computers 1000 times faster than that.
I doubt it. Digital computers and neural nets (brain) are just good at different things. Digital computers are good with things such as math, exact things. Brains are excellent at fuzzy logic. That's why it's so easy for you to read this comment. Your brain would say "That looks pretty close to a "T" so it's a T." A digital computer actually needs to run a neural net simulation for OCR, which takes quite a bit of power.
On the other side of things, try to compete with this 50 year old computer on algebraic formulas. You'd lose. Computers are just better at that.
If you don't understand any of my sayings, come to me in private and I shall take you in my German mouth.
Where can I go to buy one of these. I need to heat by back alley. I'll strech the power from the warehouse, nice and toasty cardboard box.
...is to combine the best ideas from old and new technology alike, and blend them into an entirely new result. That, to my eyes, is what real "innovation" or R&D is all about.
Some examples: DEC (Digital Equipment), in their heyday, came up with some great techniques for memory management at the hardware level. I'd be curious to know how many of those ideas got rolled over into more current stuff.
Another one; Where would we all be if Xerox's PARC facility had never come up with what has morphed into today's electronic rodent? Heck, IBM was using light pens years before that.
In short; You don't want to just ignore something because it's "old" or "obsolete" (Essence, I loathe that word!). You need to take the good ideas from the old stuff and build on them.
Somehow, I doubt that we would have so many tons of electronic junk choking landfills today if computer and electronics hardware was (a), really built to last, like the old stuff was; And (b), built to be easily upgradeable.
Bruce Lane, KC7GR,
Blue Feather Technologies
According to my calculations, if you were to port Unreal Tournament to this machine, you would be able to get 1 frame every ten days!
Lawrence Person (lawrencepersonh@gmailh.com (remove all "h"s to mail)
http://www.lawrenceperson.com/
Here's a photo
Tiller's Rule: Never use a word in written form that you've only heard and never read. You will end up looking foolish.
Think about it. That 300 KHz computer may still be running after 50 years, but those 50 years of CPU time add up to about 43 hours of CPU time on a 3.06 GHz Pentium 4. And that's just clock cycles; the Pentium 4 probably gets far more instructions per clock cycle. And, of course, the on-chip cache on the Pentium 4 far exceeds the 2 KB of RAM on that 50-year-old machine.
All in all, today's fastest Pentium could easily exceed the lifetime processing power of the CSIRAC in just a few hours, at a tiny fraction of the cost. Sure, it's cool that the computer still runs after 50 years, but let's put it into perspective here -- we get far more computing power out of modern chips, even if they fail within a couple years! Longevity isn't everything...
Deven
"Simple things should be simple, and complex things should be possible." - Alan Kay
The Linux kernel has support for a variety of hardware watchdog devices you can use on PC's.
Request your free CD of my piano music.
2kb is enough for anybody
13 year old white supremacists are shitty web designers.
I took some photos of CSIRAC while visiting the Melbourne Museum where it's now on display. They have a 2KB Casio electronic diary placed in the exhibit an example of a recent computer of the same power.
What is the inverse of the Matrix?
You vastly underestimate how many Pentium 100's there are still running Windows 95 or even 3.11.
Most people have no need to upgrade to gigahertz machines. I'm a software developer for ASP.NET applications and my Celeron 450 does me fine (albeit with 256mb ram rather than 64). Other developers in our office use Celeron 300s and some of the sales guys who only use e-mail get away with a P200. It doesn't matter how cheap an upgrade is, if you don't need it, most people won't buy it.
Only games players bother upgrading state of the art machines every couple of months. Office users just stick with what works and gets the job done.
If you told most office people that their computer would be taken away and upgraded to a new one every 3 months they'd have a heart attack.
Nick...