How Small Can Computers Get? Computing in a Molecule 143
ScienceDaily on what the future might bring for atomic-scale computing: "Joachim, the head of the CEMES Nanoscience and Picotechnology Group (GNS), is currently coordinating a team of researchers from 15 academic and industrial research institutes in Europe whose groundbreaking work on developing a molecular replacement for transistors has brought the vision of atomic-scale computing a step closer to reality. Their efforts, a continuation of work that began in the 1990s, are today being funded by the European Union in the Pico-Inside project. ... The team has managed to design a simple logic gate with 30 atoms that perform the same task as 14 transistors, while also exploring the architecture, technology, and chemistry needed to achieve computing inside a single molecule and to interconnect molecules."
well thats more just the processor... (Score:3, Insightful)
Re:well thats more just the processor... (Score:5, Insightful)
Smaller transistors means more efficient transistors. It's not just about size.
Re:well thats more just the processor... (Score:5, Insightful)
Re: (Score:2)
You seem to think computers operate using a combination of "processors" and "magic". You are mistaken.
Re: (Score:1)
Re: (Score:3, Funny)
You seem to think computers operate using a combination of "processors" and "magic". You are mistaken.
You just rocked my world view.
Re: (Score:1, Funny)
Re: (Score:3, Funny)
mostly, yeah.
Re: (Score:2)
You seem to think computers operate using a combination of "processors" and "magic". You are mistaken.
NOT LISTENING TO YOU!!! The magic smoke is REAL!!! It's real magic!!! And love, the secret ingredient is love, damnit!!!
Re: (Score:2)
And love, the secret ingredient is love, damnit!!!
Funny, my computers work just fine. <Kicks cat&$@#*% NO CARRIER
Re: (Score:2)
You seem to think computers operate using a combination of "processors" and "magic". You are mistaken.
No I think a video card goes in there somewhere.
Re: (Score:2)
>Smaller transistors means more efficient transistors.
Not always.
Those molecule sized transistors are much more efficient yes, but this rule isn't true anymore for the transistors we have in our current CPUs..
That's why now the frequency that our CPU doesn't increase anymore, it used to be to each generation of transistors were smaller and were more efficient so you could make the CPU run faster, but as this stopped the increase of frequency of our CPU has stopped also.
Re: (Score:2)
That's why now the frequency that our CPU doesn't increase anymore,
Efficiency != speed. While speed has remained stagnant over the past several years, power consumption and heat dissipation have gone down or remained the same while computations/second have still steadily gone up, which means higher efficiency as less energy is used to do the same amount of work. If you're still measuring the quality of a processor by its speed, then you have a lot of catching up to do since 2003 when the GHz race ceased to be relevant.
Re: (Score:1, Informative)
The problem is synchronization of the electronics. When you reach a certain speed the speed of electron flow comes into play. Likewise the shape of the ramp than denotes the hi / low states becomes an issue. When you 'scope what you believe to be a square wave you'll see that the edge is not the 90 degree edge one tends to think these waveforms are.
Re: (Score:2)
Smaller transistors means more efficient transistors. It's not just about size.
It also means noisy circuits.
Re: (Score:1, Funny)
It also means noisy circuits.
You mean like in Formula 1?
Re: (Score:2)
That's not what SHE said.
Re: (Score:1)
In the 1960s (Score:5, Interesting)
Combine this kind of idea with recent research on PNA (a more robust molecule than DNA which shares many of the properties) and the long term prospects could be very interesting - self-assembling memory, for instance.
Re: (Score:2)
But what would one do with a handheld that's smaller than today's? Seriously a lot of those devices have already hit the point where they're difficult for some of us to use due to the tiny buttons. I'm not sure that shrinking the size much more is going to make any sense for most people. Sure you can change the interface possibly to voice activation or direct neural interface, but tiny electronics are easy to lose.
I'm sure there's reasons why shrinking the electronics within the shell is desirable, but smal
Re: (Score:2)
You need to turn in your geek card.
Computer small enough to process audio inserted inside the ear. Reprogramble Bluetooth headset that sits in between your skull and skin. Medical diagonstic equipment that you swallow or is inserted into the blood stream by injection.Light weight glasses with mp3 playback(yes they existnow but only In bulky frames)
Of all of those a subdermal reprogramble(by light flashes like the old MSFT watch) would be awesome.
Re:In the 1960s (Score:5, Insightful)
Broaden your vision. This is about making smaller components.
What can you do with smaller components? Well, right away, you can put more stuff in the case. Your iphonanopalmtop thing can have a foldout screen and keyboard, or a bigger battery, or it can simply be lighter. I don't know about you, but I find an iPhone a bit hefty.
Now, if you look beyond next week, smaller components let you do entirely new things. You think technology is sufficient now to put a computer in a palmtop? Whatever, dude.
I want a computer in my eyeglasses. Optically corrected screens overlaying my vision. High resolution. And I want them to weigh the same as a normal pair of glasses. Don't forget to throw in a video camera for good measure.
Can we build something like that now? Or course not. That sort of thing today is either a huge bulky piece of headgear, or it's moderately bulky and has a terrible display. We need better components: much smaller, much lower power, faster.
Don't ever say we've reached the limits of useful computer technology. Until you're plugged in directly via your visual cortex and have a robot butler who brings you waffles in the morning, we haven't even reached the limits of uses we can already imagine.
Re: (Score:2)
I want a computer in my eyeglasses
Also streaming pr0n
you're too optimistic (Score:1)
i'll just settle for getting the flying cars they've been promising for the last 50 years. or just a direct brain interface to a standard PC/mac and a few applications that can take advantage of it.
imagination's nice and all, but frequently allows the marketing droids to lead us around by the nose.
Re: (Score:2)
I don't know about you, but I find an iPhone a bit hefty.
I think you need to start workin' out, mate.
Re:well thats more just the processor... (Score:5, Insightful)
OK but what if you want to put them inside nanobots designed to target and kill cancer cells or a zillion other applications that are made possible by smaller and less power hungry computation? Smaller also means more powerful computers at the 'classic' scale, for which we know there is demand for right now by way of the very existence of supercomputers.
Re: (Score:2)
Smaller also means more powerful computers at the 'classic' scale
That's a major understatement. Assuming we're working with carbon:
(6*10^23 atoms / mole) * (1 mole / 12g) * (14 transistors / 30 atoms) * (1 core2duo / 1.5*10^8 transistors) = 155 trillion Core 2 Duo processors per gram.
Even if there's a billion-to-one overhead for interconnects and so on, that's still 150,000 processors per gram. Yeah, that counts as more powerful.
Re: (Score:2)
True, and you're also one software/hardware bug away from creating the infamous grey goo - or at the very least killing everyone on the planet without creating a grey goo. I'm not sure if Kurzweil has ever addressed this specific issue, I guess he would say that biological life would be defunct at that point and we'd all hav either uploaded our minds to an artificial substrate or will have been wiped out and replaced by some machine intelligence. Happy new year!
Re: (Score:1)
Re: (Score:1, Redundant)
no, this is redundant (Score:1)
And 640K of memory is enough for anyone...
Re: (Score:2)
rather than the whole computer, i see no reason why consumer computers need ever get any smaller than a phone if you want it portable, or small enough to be fitted to the back of a screen for desktops
Why on earth would you want to interact with a computer via a screen? I want reality overlay, and some sort of neural interface (coupled with a gesture interface.) Ideally the computer that performed these functions would be no larger than a single chip to make implantation easier. Nasal cavity, mastoid, something (the latter is a little dangerous, in an impact it could become a serious health issue. But anyway.)
Re: (Score:3, Insightful)
Remember back in 1984 when the first Mac was released touted at a size of a stack of paper. Considered small and light enough to move anywhere. Then when the laptops (real Laptops, not the luggable) were released while today are considered huge and bulky but at the time they were small enough to carry with your books.
As computing shrinks our idea of size goes down too. The Stack of paper Mac, was considered really small because computing at the time the easiest job for moving anything with computing was th
Re: (Score:2)
Re: (Score:2)
Thomas J. Watson, the founder of IBM, claims he said he thought there would never be a need for more than 5 computers in the United States. We laugh at this today, because it was impossible for Tom to see what ubiquitous computing might make possible. Just as Tom was blind to the future, so you are blind. Ima
Uh (Score:2)
How about some actual data? This article is extremely watered-down ("1/100 of a nanometre (that is one hundred millionth of a millimetre!)") and essentially has nothing beside speculation about what these transistors can be used for. They don't even say what element the atoms are for fuck's sake. It's pretty amazing that they made the equivalent of 14 transistors with 30 atoms, but the article makes it sound like they just pushed some atoms together under a microscope.
Re: (Score:2)
I didn't actually RTFA (yet), but they probably did just push atoms together under a microscope (although doing so would require specialized equipment which is more than simply a "microscope"). The damn thing will probably fly apart once it gets above more than few tens of degrees absolute. It's great if they can put a P4 chip on the head of a pin and not need a huge heatsink for it - but it's useless if you need to carry around a couple of thousand pounds worth of cooling equipment to use it.
(Having read
Re: (Score:2)
Yes! Would it kill the online science magazines to post a link to the actual research web site or a technical report? Geez, I'm getting tired of having to search through multiple websites starting with the researcher and/or department name to find the actual research because the Google search is swamped by the press release that's being flogged. All of this could be avoided if the damn press release would contain a link to the research page and the science mags would include this link.
How small can computers get? (Score:4, Insightful)
Re: (Score:1)
Not a problem. You just need to interconnect such computers to create data-center inside of your phone.
Re: (Score:3, Insightful)
Re:How small can computers get? (Score:5, Funny)
Electronic viraga?
"Mind passing me my phone before we start? Just need to SSH into my penis."
Re: (Score:2)
Re: (Score:2)
Yeah, you're right, SSH is totally inappropriate. It should be a VPN so you can send other things there.
Re: (Score:2)
Re:How small can computers get? (Score:4, Funny)
Re: (Score:2)
Which emBEDded OS would it run!?
Not one that has anything to do with "micro" or "soft."
Re: (Score:2)
Electronic blowjobs! Who needs viagra when you can directly control the brain?
Re: (Score:1)
Re: (Score:2)
That will only be a problem until the invention of a telepathic interface.
How small can a computer get if it is implanted into your brain?
Re: (Score:2)
Re: (Score:2)
It would be great to have a computer the size of a halfpenny
Uh, you mean, like the CPU in the computer that you're using right now?
Re: (Score:2)
A CPU doesnt make a computer, a CPU is a computer in need of a power supply, memory, clock source and BIOS. Fit that all on a penny and call me back.
Re: (Score:3, Informative)
Re: (Score:2)
You can always do what we do all the time: make it the size of the current processor and say that it is a million times faster.
Re: (Score:1)
but how would you access it?
Just what I was thinking. Obviously other technologies would have to be developed alongside this processor/memory/whatever form it eventually possibly takes.
The day is coming though when we can take a piece of 'paper' from our pocket and the paper is the computer. The possibilities in regards to pervasive computing are interesting.
The true question isn't about when they can make a full usable computer using this technology. The real show-stopper is when they can make it affordable and market it in a world
Why is this insightful?!! (Score:3, Funny)
STOP
I'll stick with my telegraph, thank you very much.
STOP
Yabbut (Score:2)
Re: (Score:2)
STOP
I think those crazy radio-wave-transceiver early-adopters will find themselves only the object of more ridicule once this little fad burns out.
STOP
Re: (Score:1)
You wouldn't have a computer the size of a halfpenny. You'd have a computer the size of a cell-phone, with the processing power of a 32-node Beowulf cluster.
If you have one of those, you have exactly zero use for the cloud: so long as you have a monitor, keyboard, and mouse, your computer can be exactly where you are.
Obviously, you want some backup, but that is easily solved by have two or three of these things, one of which sits on your desk, and you sync to it when you're at home, so if you lose your port
iPod shuffle size of its controls (Score:2)
Re: (Score:1)
If Microsoft got involved, it would need at least 3 controls...
(Ctrl-Alt-Del)
Re: (Score:2)
Angels dancing on the head of a pin time (Score:2, Insightful)
Really the question is how small becomes impractical? I remember the calculator cold wars. It hit the limit when everyone realized how silly a pen one was when you couldn't read the display or use it without a tiny stylus. Eventually the cost of reducing the size will be astronomical so even if you can what's the point? We can make antimatter fuel it's just so insanely expensive that without a major technical leap you aren't going to be powering a car much less a starship with it. There may be uses justifyi
Re: (Score:2, Insightful)
Re: (Score:1)
Create cells? We've got to understand them first! We'll all be having nano-ipods implanted in our skulls before we even come close to being able to engineer one human cell.
Re: (Score:2)
replace failing cells
Create cells?
I took it to mean the dead cell is discarded and the nanobot moves in to take it's place. Rereading his post, it seems he did mean creating cells, but that might not be necessary. We just need a nanobot that can serve as a proper replacement.
Re: (Score:2)
We can? Link to article?
Re: (Score:2)
Re: (Score:1)
halfway there? (Score:3, Insightful)
The real key to all of these and all non-trivial efforts at Nano technology is for these devices to be self assembling. By non-trivial I mean other than "simple" things like nanotubes or quantum dots. These simple compounds can now be produced in industrial quantities through basically chemical/physical means.
While it is very very impressive that they can do this, in order for this to become practical, they will have to make millions, no billions, no trillions, no quadrillions... of these things at once or they have to be able to duplicate/reproduce themselves. The (self) "assembler" is, of course, the holy grail of nano-tech.
Hope I see it before I die and that it doesn't cause my (and all of our) deaths! :P
Re: (Score:1)
Assuming such a self-assembler was created, I would expect it to require some very controlled situation for reproduction - extremely high temperatures, or extremely cold ones, or possibly a flow moving from 400 Kelvin, followed by a rapid cooling to 150 K, then a move to room temperature, at which point the process could be repeated, given enough materials to produce more.
Such materials would probably just be raw Carbon, hydrogen, and maybe a few other obvious elements, but I wouldn't be surprised if you ne
How small can they get? Do they run Vista? (Score:2, Interesting)
Meanwhile, atomic-scale computing is created, and at this stage its hard to say whether this is a step or a giant leap in the right direction.
Re: (Score:1, Flamebait)
Re: (Score:2, Funny)
No, but someone ported Linux to it...:D
Soo... (Score:4, Interesting)
No (or rather, probably not) (Score:2, Informative)
> is now more than exponentially larger?
It's impossible to tell if it's scaling linearly or exponentially or whatever from just one data point; however, unless the atoms are working in a totally different computing paradigm (like quantum computing), it's unlikely to be more than just a linear factor of improvement.
Re: (Score:2)
might it be the plateau that Moore's law predicts?
Yes:
In terms of size [of transistor] you can see that we're approaching the size of atoms which is a fundamental barrier, but it'll be two or three generations before we get that farâ"but that's as far out as we've ever been able to see. We have another 10 to 20 years before we reach a fundamental limit. By then they'll be able to make bigger chips and have transistor budgets in the billions.
-Gordon Moore, 2005 http://en.wikipedia.org/wiki/Moore%27s_law#Ultimate_limits_of_the_law [wikipedia.org]
Nanotransistors - Nanocomputers? (Score:2)
But about complete computers, well, still dont know if all components could be stick together in a single molecule, or that it retains all the components functionality in that way.
Damn (Score:2, Funny)
Sometimes I misplace my laptop. How will if find my tiny computer
in the future? Will I wash it by mistake? Can it take the dry cycle?
Grrrrr.
Wow (Score:5, Funny)
Well isn't that nice (Score:2)
But what will you do for I/O, then?
What about cosmic rays ? (Score:5, Interesting)
So are we going to have to shield tiny computers with an inch of lead ?
Re: (Score:2, Insightful)
Doesn't normal RAM get hit by cosmic rays and radiation? AFAIK it also suffers from bits being flipped incorrectly. Even Flash memory suffers from individual cells dying without much problem.
I am sure there are ways to offer redundancy and failover between molecules, eg. you could create self assembling groups which all do the same calculation, a controller could then decide which ones are right based on probability, dead molecules could be marked in the FAT... err I mean the MAT.
Most of the technology we
Re: (Score:3, Funny)
Re: (Score:1, Insightful)
A bit of radiation whizzing by would not just 'flip a bit' and make the computer/program crash (or even worse - produce an erronious result) but could dislodge a few atoms and physically damage the computer.
So are we going to have to shield tiny computers with an inch of lead ?
Instead they will be like modern computers with fault detection, error correction and automatic rerouting.
Re: (Score:2)
I wonder how much complexity would be added by the fault detection and repair logic to offset the advantage of this kind of die shrink. I take it atom chips would quickly deteriorate in speed as more and more cores fail?
I have a bunch of devices back from the 80's that are barely warm to the touch and will work forever. I can't say the same about even a conservatively designed modern computer, let alone all the hand-held devices that are engineered to die from exhausted polymer batteries after a year.
Re: (Score:1, Insightful)
Won't work near Tchernobyl (Score:2, Interesting)
Heat and memory biggest issues... (Score:2)
... that are facing computing. CPU speed is far out-stripping storage and memory bandwidth. More efficient transistors = nice, but LESS robust to defects = bad. I have to wonder how fragile these atom transistors will be. I'm wondering if we're approaching a point where having too few atoms leads to much higher failure rate.
I can't be the only one thinking about how expensive this is going to be.
Re: (Score:2)
Reliability decreases with pow(number of atoms per transistor, 2/3), now when you apply some kind of error correcting, it increases with exp(number of transistors). If you maintain the chip size (that is, the number of atoms is constant), and the functionality, reliability will just increase with smaler transistors.
You get reduced reliability only if you want those smaler transistors to do more than the big ones, and even then, when you want them to do a LOT more.
30 atoms (Score:1, Funny)
... ought to be enough for anybody...
A Beowulf cluster? (Score:2)
Imagine a beowulf cluster of these! And a beowulf cluster of those clusters! Why, I could fit them on a matchstick head, call it W. Bush's brain!
Content-free article (Score:4, Informative)
I much prefer to read Eric Drexler's PhD thesis, Molecular Machinery and Manufacturing with Applications to Computing [mit.edu]. Chapter 11 (nanomechanical computational systems) is particularly interesting.
what even happened to "smart dust"? (Score:2)
One thing to say... (Score:1)
Re: (Score:2)
You drop a computer, you're screwed.
If the computer has a hard disk with moving parts and a giant, heavy heatsink that will chip the processor, then yes, you're correct. Drop a mini-ITX PC with a SSD and no heatsink and you'll find that there's really not much you can do to it aside from cracking the motherboard to bring it down. I'm sure you've dropped your cell phone numerous times (If not then congrats, you're abnormally un-clumsy) and unless it's pretty severe, it'll be just fine. One of the primary benefits of solid-state electronics is
Re: (Score:2)
I know! Let's write the operating system in Perl!
Re: (Score:2)
Something makes me feel that we won't have a problem burning all those cycles with several layers of VMs and emulation.
Re: (Score:2)