Slashdot Mirror
How Printable Computers Will Work
Gart points to this article on printable computers, an "Illustrated narrative [that] shows how users will simply download microchip designs from the Internet and print out a working ink-based, plastic processor on a desktop fabrication machine, similar to an ink jet printer." This is a nicely lucid account, and straightforward about the reasons that you probably won't start printing out a new motherboard this evening. Still, a glimpse of the future; this is one technology it will be cool to watch emerge from vapor.
This should have been one of the perils in the prior article on over-clocking.
:)
Not only do you risk ruining your CPU, but now if you overclock your video card, you could melt your CPU!
wonder what it'll do for recycling - hitech and lotech plastic recyclables.
I've been printing computers and then building them during class for the last few weeks. I get them from the Paper Mac page, and print them out with my laser printer. Then I take them to class and, while the teacher is babbling dumb stuff about what she has done with computers, I build macs. I've got a small army of them collecting here...
Posted from the wireless couch.
The original 80386 (with 275,000 transistors) fits in a 13.1 by 13.1 mm square.
It turns out that it's much worse than this. The minimum feature size gives you the minimum gate width you can use (if you're lucky). A transistor, contacts and all, takes up easily ten times this in both directions (after spacing rules and so forth have been taken into account).
Speed is also inversely proportional to the square of the feature size. Going from 0.25 micron to 25 micron slows you down by a factor of about 10,000 (give or take). You'll end up with microchip wallpaper that runs at 50 kHz.
There are uses for this (big active displays comes to mind), but computers aren't one of them.
I have a name to suggest for this: Breadboards!
Neat, huh?
--
--
"Outlook not so good." That magic 8-ball knows everything! I'll ask about Exchange Server next.
Picture people jumping out of windows of call centers. "What kind of graphics card do you have?" "Uh.. well, it's a Voodoo 5 base design, but my friend added another texture unit, and then we kinda played with the blending mode logic..."
Now there's a strange thought.. presumably, condoms would be a workable substrate for circuits.
"Um... no... that's not a webcam printed on my condom, or something.."
Just because you can doesn't mean you should.
There could be a licensing agreement, but how do you track pirated hardware?
There seems to be a new trend to have software contact a company's servers to see if it's been pirated or not. Hardware could do the same, and if it's pirated it could explode and leave an inky mess all over just like the capsules banks put in with stolen money.
Hush. Somebody will figure it out. :)
Sure, I'll pick it up from my parents house tonight and post the author. There is a small chance that I misremembered the name and it's 'A' for Anything, but I was pretty sure I remember it as being an odd title because it didn't match up.
Kintanon
Check out JoshJitsu.info for Brazilian Ji
Good question. I don't think the book ever addressed the point of Land ownership at all...
And it didn't really occure to me until you mentioned it that Land really should be the most valuable item in a society like that... I think I'll reread the book tommorow and see if they mentioned it while I wasn't paying attention.
Kintanon
Check out JoshJitsu.info for Brazilian Ji
Surely, if you're "printing" a chip, even of today's standards of size, you're still looking at the printer moving the substrate along minute amounts to get the design right. It would probably take several days...
The article says we can expect roughly 1/100 the speed of current silicon. I guess that means CPU's with 500khz - 1Mhz clock speed. The kind of machines that started the PC era.
At the same time, we're seeing a strong attack on the programmable, user-controlled computer by the Intellectual Property cartel. Maybe we'll end up using CP/M on 1979-vintage machines because the current crop of hardware is too locked-down and tamperproof. A machine like that could be open-sourced and downloaded from the internet.
On the flip side, I see many possibilities for abuse. Manufacturers could build flat computers into cereal boxes, shipping labels, even software packaging. Lots of places to monitor, advertise, and present license agreements. Lots of stuff that quietly 'phones home'.
Is it good or bad on the balance? Doesn't matter - nobody can roll back technology.
For some reason when you said Trojan hardware I imagined a condom with "chips" printed onto them. Oh the possibilities there.
Yeah, then we would have to put up with script kiddies writing VBScripts that add unwanted anchovies to those aforementioned pizzas.
What about the metal interconnect? Aluminum or copper... or not metal at all? What about metal layers and via's? What kind of package will this "die" go into?
I'm curious as to how they plan to address these issues
Or maybe the Gimp could have a plugin for printed circuits
...now Apple will be able to simply print out gay-ass color-coordinated Bondi Blue motherboards to stick in their iMacs. Wonderful.
Sean
I haven't read it in years, but I seem to remember in Clarke's "The City and the Stars" a passage regarding the village "foundry". Every village had some sort of self-contained "millworks" about the size of a refrigerator that produced "useful devices" (my quotes). Can my recollection be any more vague? Anyone have a copy on hand to refer to?
Come to think of it, this moves computer technolgy, such as creating machines to do "x, y, or z" into a freedom of the press realm.
The petty little potentates pop paranoia and profiteering must be trembling in hooror at this, once they figure it out and see it coming down the road.
This could be fun!
"It is a greater offense to steal men's labor, than their clothes"
Thinking out loud here... If a CPU or board design can be expressed in a printer markup language under this scheme, then I believe it would be possible to translate that description into plain english. Or a suitably different reverse-engineered piece of hardware could be described this way. It could be expressed out loud on street corners in songs or poetry or t-shirts. (Just like what happened with DeCSS.) Only free speech would be used to distribute hardware. Great for spreading not only the word but the means of useful hardware to the masses.
Now, granted, these would be pretty screwed up songs. And pretty long! And I guess haiku is out of the question. But in theory, this could lead to free-speech-like implications that the hardware industry may not have had to deal with yet.
Sing me a song, you're the VLSI man...
It's seems very unlikely that we would have a future where every single home user would download the new AMD Athlon K399 1 Billion MHz, although I must say that the idea of me going to my computer resaler, and having _them_ print a copy (no more "out of stock" replies") would be really cool.
Hey, it was just a dream. Not everything always makes perfect sense in a dream.
But good idea. Ooops... bad idea. We don't want this to actually happen.
I'll see your senator, and I'll raise you two judges.
I can't wait to have a 10x8 foot monitor! I'll scare the heck out of my neighbors with life-size Q3A!
std::disclaimer<std::legalese> sig=new std::disclaimer; sig->dump(); delete sig;
It'll raise the level of computer literacy by a notch. Not only will it be divided into the standard 2 groups:
Can program VCR
Cannot program VCR
The Can program VCR group will be divided into
Can read PCB
Cannot read PCB
Joking aside, it wouldn't be difficult to, like a script kiddy, just download blueprints and get yourself a homebrew mp3 player, but how can you 'trust' such kits, implicitly, any more than you can trust software?
Mostly it should be okay, but the odd virus here and there could wreak havoc on someone. So a new class of 'anti-trojan' software, as well as more literacy in computer skills in general, needs to taught.
I would like to have a private fab and rapid prototype lab in my garage. That would be muy nifto.
Louis
Geek dating!
GPL Deconstructed
Open source, man!
I spend time and effort to develop a car Ogg player and GPS receiver.
I am part of a group of people who have similar interests, so we all share our variations on the designs, LCD or display implementations, voice activated, IR, bluetooth, etc.
Over a period of months or years, we all have in car dash Ogg players, GPS recievers with voice activation and other random stuff.
Other people take this and adapt it to Visor handspring modules
They add power saving and cycling functions, or something. Or they make it smaller.
Feedback occurs, and the indash unit becomes smaller. We get enough space to add more features.
Back and forth.
Isn't this how the Open Source model works?
Geek dating!
GPL Deconstructed
--
A feeling of having made the same mistake before: Deja Foobar
"The fab jammed again..."
"Is there a jam?"
"...it says it's out of toner?"
Will the real Bruce Perens Please Stand Up
But if the printer could do a 3-d chip, and you could adequately cool such a chip, then a pentium class chip is possible.
Anyways, it might be possible to do a mini-cluster of cpus to give the same power. If "printing" hardware becomes cheap, I'm guessing the computer designs will adapt. Maybe, instead of a localized cpu/memory/etc, the components would be integrated. Imagine a "motherboard" with each square inch of it being its own cpu/memory. There would be no seperate vid/sound/modem cards, just the appropriate connects on the edge of the board. Rather spiffy.
The biggest use of these is likely to be in units that have to travel to places where getting replacement hardware is difficult. Such as extraterrestrial expeditions, or armies at war on earth. Combine this technology with a fabber and you can build whatever replacements you like when things break.
The far ahead possibilities include virally spreading across the universe using fabbers that build more fabbers...
I wonder if E-Paper could be adapted to form circuit boards.. Download the chip, plug it into your e-paper, watch it assemble itself. I bet it could be done.
Douglas Adams
1952-2001 :(
The headlines:
2010 eGates hardware over throws governments
2012 eGates hardwar controls 90% of the internet
2013 TuxBoard Manufacturing, ESR CEO started in conjunction with the GNU Project has begun. The EG Supercomputerweb laughs.
2013 People sick of having their homes burn down try TuxBoard Manufacturing out of curiosity and hope
2014 EG SCW now scared. FUD attacks: TuxBoards don't burn down, they must not be capable of the high-level processing.
2015 Cox on a Chip capable of making eGates hardware work without burn out and without the need for Windows 9000 XPIJFLSFN
2016 Helsinki almost nuked, but bombs fail to go off, and land 300mi out of range...hardware failure.
2018 Redmond uprising attempts to overthrow EG SCW
2020 After the 100 GHZ war, Linus Torvalds reigns as occaissionally benevolent emperor over 98% of the world.
2021 New Zealand finally ready to cater to AmigaSketch users.
2024 Massive solar flare damages most of the computer world.
2025 Etchasketch becomes popular
2026 Printable punch cards arrive on the outskirts of the remaining ecivilization.
2027 Last Beowulf cluster dies in captivity.
2028 Torvalds frozen and displayed for all to look at.
3028 Cockroaches and lawyers wonder why there is a frozen person on the European continent. Lawyers wonder if he has money (which looks a lot like punch cards)
3030 The Troll Niscenus and a band of AnonCows re-establish world order with a central hub of infromation they call DotSlash.
So, who's up for volleyball?
"Yeah...it was the numbers that were irrational, not the murderous cult of vegetarians...." -- Hippasus of Metapontum
I can only assume we'll need Cox on a Chip to write the drivers for the new e-paper.
I don't think we'll see printers sketching 8-atom wide paths on unborn microchips, ever.
A) The "printer" would cost way to much for any common use.
B) Have you seen the size of the machines that currently do this? They're ENIAC, hyperbolicly speaking.
C) By the time anyone gets this going, computer will be communicating on their boards and processors with 8bit light streams rather than electonic anything...err, not that I know anything. If Century Tel didn't buy GTE from Verizon, you wouldn't not know anything either.
I usually get turned into a troll everytime I say this, but, "Who's up for volleyball!?"
"Yeah...it was the numbers that were irrational, not the murderous cult of vegetarians...." -- Hippasus of Metapontum
I hate to spoil all this dreaming with realism, but you'd really be pushing this technology to make even an 8-bit CPU. (See the third page of the article.) The smallest elements you can print are about 25 micrometers (.001 inch), which is over 100 times the linear dimensions in a modern Pentium chip. And that means the transistors will be 10,000 times the area, and 10,000 times the capacitance. Add the inferior performance of the plastic transistors to the complexity limitation implied by the dimensions, and what you wind up with is something like a 6502, but running at a few KHz...
There are actually many control applications where a small 8-bitter running at KHz speeds is quite sufficient. But you can buy good 8-bit silicon CPU's for $0.50 to $15 (depending mainly on how many pins you need), and I don't see any chance of these plastic circuits beating those prices.
What this might be good for is the custom interface circuits that are virtually always needed between the CPU and the world. These usually wind up either as a large number of generic components, soldered onto a fair-sized circuit board, costing perhaps $20 to $100 to utilize a $5 CPU. Or you can use a few programmable logic device chips -- but these cost more than the CPU. So if they can get reasonable price/performance, you might eventually see printed plastic circuits containing the "glue logic" as well as the resistors, capacitors, and ESD-suppressing diodes -- so you just solder on the CPU chip and its ready to go. (But not for a Pentium motherboard -- think about the speed.)
More realistically, there are many applications where large, low-performance circuits would be ideal. Displays, for instance -- you can etch a wonderful display into a silicon chip, but you need a microscope or a good projection system to read it. Maybe they will soon be able to print the same circuit in plastic at readable size.
This wouldn't be suitable for most computing applications, where designing software for a generalized processor would be easier and more effective most of the time.
But for dedicated and experimental electronics, it'd be golden. Just think how much this would accelerate prototyping and design for BEAM robots, since you don't have to mess around with soldering bunches of components!
cryptochrome---If you can't trust a nerd, who can you trust?
Technology Review had an article about the very same topic months ago. Surprised it didn't hit slashdot then. Well anyway, it is an interesting read as well for anyone who want to hear more.
This reminds me of a book called "'A' For Everything" where an inventor created a duplicating device. You put the object you want duplicated on one side, and press a button, and a copy comes out on the other side. Within hours almost everyone had one. Within days the economy had reverted to Slaves being the only valuable form of property. From there people figured out how to clone other people using the device and a bizarre slave/owner hierarchy was born. A very interesting book, I recommend it to everyone.
Kintanon
Check out JoshJitsu.info for Brazilian Ji
However, I doubt we'll see any sort of IC fabs capable of producing anything as complicated as say, a 555, on the desktop within the next 50 years. Call me skeptical.
Electronics buffs, we build stuff that we download, design, read in books, etc. Automation would be getting access to a plotter or a UV rig for making PCB construction much, much easier.
-bugg
Check it out here!
Its true the density isn't as great, but in this case the medium is just a piece of plastic, not a VERY expensive piece of silicon, so real estate is not a commodity... however it sounds like their switching times are much slower then silicon transistors ... given that and the distance between the resistors growing (because they're less dense) it sounds like the problem will infact be speed.
Free Techno/Jazz/DNB/MI Music by guys obsessed with monkeys!
I had a dream some months ago.
Printable PDA's were developed. You could have them printed on the surface of your skin. No more need to wear a wristwatch, you would just print a PDA on your wrist.
Problem was (in the dream) that they would wear off, and you would have to have them reprinted (Or have a newer better model printed) onto your wrist.
They became so common that everyone had to have on in order to just carry on daily life, business transactions, etc.
Poor people couldn't afford $40/mo to print these things on their wrists.
So good ol' corporate greed stepped in to help. You could get a sponsored low-end PDA printed onto your right hand by agreeing to also have a color animated advertising banner printed across your forehead. Next month, come back, pay homage to corporate greed. Repeat.
Don't laugh. I really did dream this. (I said, no laughing.)
Those who can, do. Those who can't, use Windows.
I'll see your senator, and I'll raise you two judges.
Now that's the real trick, isn't it? In modern integrated circuit design the interconnect uses up more area than the transistors. Even if you could do all the wiring in other layers (by the way, only VERY recently have ICs come out with lots of layers. One or two wiring layers was the standard for YEARS) you would still need lots of vias to move the signals between layers and down to the transistors.
And as for cost, I just checked MOSIS and if you needed a 6.5mm by 6.5mm square silicon chip fabbed, MOSIS would charge you about $70000 for a lot of 25! Kind of pricey for a 68k processor, don't you think?
There are a lot of reasons silicon is useful, and I'd be VERY suprised if people started printing chips out on their desks.
downloadable and printable hardware, if it develops to fruition, will destroy the need for a computer hardware insdustry in certain sectors.
the gamers, the kids, everyone who uses their computers as a hot rod and doesn't have mission critical stuff running on them, will pirate hardware.
There could be a licensing agreement, but how do you track pirated hardware? use it on the Net and it sends out a call signal? I don't think so. Even if it could be traced, people would just set up parallel intranet networks with the stuff instead of using it on the Internet, where cops, et al could track it.
All in all, opening the doors to terrorists, foreign intelligence agencies, and anybody else who wants to reliably gather information without much expenditure and without being traced.
Goat sex free since 2001
It seems like once they have perfected this tech, a natural extension would be to repeatedly overprint with layers of conductive, nonconductive, and semiconductor ink to create a 3-D circuit. This could go a long way toward offsetting the registration size and natural slowness of the junctions, since you could stack them into a cube. If the layer size is comparable to the junction size, you could end up with breathtaking densities. With good massively parallel architecture such blocks -- they wouldn't be flexible and would probably be more like regular IC's once manufabbed -- might dwarf the performance of single-layer silicon.
Brackets contain world's first nanosig, highly magnified:[.]
Don't underestimate the importance of this.
I predict one of the earliest practical application will be a real credit-card calculator -- one the same size and thickness as your credit cards. Feature size will probably go down at about the same rate it has for silicon. Remember, back in the 70's feature size on silicon wasn't much better than what these guys are aiming for. And neither was the speed -- remember 4000 series CMOS? Yet these chips were the backbone of all technology when they were expensive.
This was a good article, well balanced and without a lot of the usual hype about how the tech will result in self-assembling skyscrapers. These people have modest, achievable goals. It will be interesting to see what they do.
Brackets contain world's first nanosig, highly magnified:[.]
This will bring a whole new definition to the word "warez"
---
---
---
"Against stupidity the very god themselves contend in vain" -Johann Schiller
Cripes, I'm old.
It must be 20 years ago now that Carver Mead was talking about having a machine on your desk that would fabricate silicon integrated circuits for you.
Anyone who's seen a fab line knows it's not that simple. The closest anyone came was e-beam lithography, but those machines are still the size of your bathroom, and still only do some of the processes.
It's pretty interesting to think about printing chip layers like a multi-pass color laser printer.
But can you imagine the toner-cartridge spam you'll get when there are ten kinds of toner material needed, and some run ten bucks a pass?
--Blair
"Dammit! I left my TiVo folded up in my pocket and it went through the wash again..."
One of the critical problems for long distance space endevours is what to do when/if you need to replace hardware in your systems when you're far far away from any fab plant. This kind of technology will go a long way to making the problem moot.
Demonstrant's Open Source Tools
This article has a few more facts about this technology, and its references are from journals like "Science" and "Applied Physics Letters" and "Chemical Review," so you can do some in-depth research if you wish. It's not so complicated that the layman could not read it, and it has some information not covered in the howstuffworks.com article. I did like some of the pictures in the howstuffworks.com article.
The electron mobility in polymers is MUCH lower than Si (a slow semiconductor), a fact that is mentioned in the article, but glossed over on this page. Overclocking these guys still won't get you very far.
One thing not mentioned is the short shelf life of these things. They tend to degrade in days to weeks, depending on the material.
I could go on, but I won't. I'm just glad to see this finally out in the popular media.
They are talking about achieving a 25 micron feature size. The current generation of processors is being done with an 0.13 micron feature size, meaning that the number of gates you can fit on your plastic chip is about 40000 (200 times 200) times lower.
Still, if they can get one transistor in 25 microns square, and handle all the wiring in other layers:
Trying to get much bigger than this (do the P5 in two inches square) is likely to be a loser because getting the signals across these large chips is going to be slow unless you use enough power to melt the plastic.
Memory: if you can do one bit in 25 by 25 microns, a square inch (2.54 cm on a side) gives just over one megabit (bits, not bytes). You're probably not going to be running Gnome or KDE on this.
... until they figure out a way to download and print a Pizza.
... hmmm
"Hello, Domino's? Can you e-mail me a large pizza with mushrooms and extra cheese?"
But then, that might bring a whole new meaning to "Spam Mail"
I wonder if they realize they are only going to sell a handful of these things... The first guy to get one of these printers is just going to print up more printers for his friends...
Trojan hardware.
In the future... motherboards may be nothing more than layers of nonconducting materials with thin conductors connecting various chips and sockets (into which other boards can be connected). Imagine the possibilities!
The same way that literacy, pencils, pens, ink, paper, and books destroyed the iron grip of the autocracy and nobility of hundreds of years ago, printable computers can break the grip of monolithic oligarchies dictating hardware and standards to people who don't need them.
Don't you have any sense of decency, to post such utter garbage in the first place, perdida?
(go ahead mark me as troll or whatever, it really doesn't matter)
Geek dating!
GPL Deconstructed
--
A feeling of having made the same mistake before: Deja Foobar