Slashdot Mirror
Towards Molecular Computing
pq writes "The NY Times has a progress report on molecular computers: the results are finally rolling in. This July, HP and UCLA reported molecular logic gates; now Yale and Rice are reporting the ability to cycle those gates on/off and HP is announcing conducting wires less than a dozen atoms across. Interesting review - to quote, `this should scare the pants off anyone working in silicon.' " Mmmm...nano.
It seems to me that newer isn't always better
From everything I've heard, read about, and drempt of, this is far better. Silicon chip makers aren't going under today, chances are the existing companies will have a large stake in the new technologies. Nanotechnology is the next step, and I find it vey unlikely that it will just be some "useless, new" technology no one will care about. IMHO, we're in the beginning stages of a major paradigm shift, starting with the microcomputer revolution of the late 70s to the PC explosion of the mid/late nineties, and the shift ending (and new paradigm beginning) with the AI and nano revolution of the 2020s-2030s. That's judt my guesses, we have a few years yet =)
The Good Reverend
Ok, sure building a really tiny gate is cool and all, but the question is: of what use is this?
Sure it may be tiny, but if it can't make a state transition fast enough to keep up with today's silicon - forget it.
Yes, smaller may be better for Place and Route, but what about the REALLY important stuff like - timing. (ok, the guy does P&R for a living, begin your rant). I'll argue that there are more time critical hardware designs than size critical hardware designs, but correct me if I'm wrong. Faster is better than smaller.
You say you want a revolution?
If it does manage to remain the place to be, I'd suggest Assembler Alley, Polymer Pass (too chemical, maybe) or Nanogate Notch.
--
Time is Nature's way of keeping everything from happening at once... the bitch.
I'm just pondering over the problems inherent in Quantum Computing.
The role of a technician could change drastically. Instead of fixing problems, a tech would just sit and observe the system at every given moment so that it runs properly. If the tech stops observing for any length of time the system would go haywire.
Then again... maybe it's not so different after all. :)
LouZiffer
There won't be a difference between human and robot. Robots wil never happen, because nanotech allows for people to be modified to suit the purpose you were going to build the robot for. What would you use a robot for? Mining? War? Service jobs? We have millions of poorly educated and more-or-less self-maintaining people out there who will do the same job with a couple of nanotech tools and some quality propaganda.
I just saw Three Kings and thought it was the best movie I've seen in a very long time. Go watch it, and then ask yourself if the powers that be would or wouldn't chop-n-channel your sorry ass instead of building a robot from scratch.
SP
"Nothing was broken, and it's been fixed." -- Jon Carroll
I agree that feature size is not the sole determining factor driving processor performance, but it has been the major factor in the speed increases of the PC. Architecture may squeeze out an additional 10% here and there, and software another 10%, but that is about all. Some tasks are inherently unparallelizable, and those are the ones we need faster CPUs for the most!
What people should realize is that there are other chip technologies which can be used long after the feature size bottlenecks are reached. Bipolar logic has always had the potential to be faster, but it is far more expensive in terms of power consumption and real estate. Mixtures where non-FET-based logic is mixed with traditional logic can still yield impressive speedups. There is also the active cooling of silicon to consider. After that, there are other semiconductors such as GaAs an InP to work with.
The molecular computing research reminds me what I heard when the first optical-based computing methods were being investigated. Yes, there is a lot of promise, but we won't see it until it is actually better than the existing tech. None of the articles I've seen even mention the potential speed limitations or yield rates of these devices. It is too soon to tell. It may be that they switch too slowly, or have such a poor reliability that they will never come into major use. Don't count silicon out yet! It will be around for at least 40 years, although niche areas will be gradually taken over by the "exotics".
Actually lifespans could be much longer. I heard that, if it were possible to prevent aging, people might be able to live on the order of one hundred thousand years each on average.
Moderators, please travel at a -1 highlight threshhold so you notice interesting posts like this hidden down threads.
As linux is compiled with the C compiler, I would think that all the C programs would be the first thing to be ported to the new architectures. For that matter, aren't significant chunks of all the interpreted languages mentioned here written in C also?
OK, I admit to lifting this from that one episode of X-Files (w/ the virtual life thing), but couldn't someone create a program that creates keys from reading the data on cds or mp3s? I'm already imagining encrypting my emails w/ one of Mozarts concertos.
Haynes. Snapper. McDonald's. With computers this cheap, the primary purpose they're put to will be "smart" materials and "smart" products. Your hamburger wrapper will flash and sparkle an ad for whatever product you didn't buy. Your socks will beep at you if you put on a mismatched pair.... All for an extra fifty cents (and an extra cost of a penny or two).
Now the rub on THAT whole idea is:
Will we have figured a way around population growth by that point in time? There IS something to be said for people dying by natural means. If things were to continue at that rate we would over poplate our galaxy. Of course I spose some people will still die in stupid ways...drunk driving/flying and the like. I don't know that I would WANT to live that long honestly. Of course I have severe moral issues with life extension in the first place. And futurist technology as a whole. That's for me to work out though. =)
"We hope you find fun and laughter in the new millenium" - Top half of fastfood gamepiece
"Fighting the underpants gnomes since 1998!" "Bruce Schneier knows the state of schroedinger's cat"
The question of interfacing is an extremely important one. Just because you have the means to make the connection does not necessarily mean you have the capability to understand the 'language' of signalling. I will admit that the visual system is one of the most studied of human physiology, but there are still many questions to be answered. Nanotechnology is not the ultimate solution to any one problem, particulary this one. . . .
STFU about slashdot bias.
The Canadian government will convert the new Level 4 Biohazard research facility into a nano lab... just you wait ;) well.. i can hope
"Manufacturing might involve assembling trillions of circuits and then identifying and mapping out the bad ones -- much as faulty sectors are declared off limits in today's disk drives.
"It's a very biological approach. Everyone's brain is the same, but the pathways are all unique."
Actually I interpret these two statements as a very strong connection between mulecular computing and AI (why not?) Imagine "pouring out" a motherboard and allowing it to form itself throughout its productive usage time.
Will bad circuits be isolated only in production, or on a continual basis throughout the circuitmass' service tenure? If the latter, then would this circuit not then gain "experience" in addition to data? I would at first compare this to the TCP/IP system, but on a vastly condensed scale. Does this not, at least, simulate learning process?
What about this statement:
"But researchers in molecular electronics are optimistic that they will be able to
In addition to an evolving system that maps out its own bad sectors, consider self-assembly (I know this is a stretch, but...) A circuit mass that can map out its own bad sectors, self-assemble (reproduce?)
How distant then, would be a self-assembly process that detects the cause of bad circuits, and implements processes to aviod these causal situations?
At which point will have we stopped witnessing product evolution vs. genetic Darwinism?
For those who really care, 10^-18 is 'atto'. But the prefix 'nano' seems to be already accepted for the purpose of molecular computing. I assume that the resulting processors would be 'nanoprocessors'.
But there will probably one day be better technology...
picocomputing - using electrons for computing (pretty much quantum computing)
femtocomputing - computing in base 6 using quarks
attocomputing - computing using whatever the hell quarks are made of (superstrings? energy patterns? green cheese?)
--
Win dain a lotica, en vai tu ri silota
And of course, once you have a nano robot, you can use it to assemble a smaller nano robot... until we start manipulating individual molecules.
Does anyone see potential in using nanobots to fix corrupted chips, boards, etc. once such an error is detected and diagnosed? Store a number of 'bots somewhere in the motherboard, in a little compartment, and send them out to fix stuff within the computer?
hey thanx!
The digital engeneering work will require exactly the same skill set (routing power raround the die will be different).
As for the people working in the fabs, they proably dont make a whole hell of a lot relitivly.
Here's a link to a project that already has some function at 2GHz, and is planning to step to 4GHz with some of the new technologies. Cool stuff, and a great group of guys.
This includes some chip shots, as well as descriptions of several of the new technologies that are being used for the project.
"It's tough to be bilingual when you get hit in the head."
No, not unless they use organic molecules to create actual replicating cells. Virus' require a complete and functioning cell with replicating DNA in order to do their deed. On another note, however, if there were replicating-cell computers, you would then have to worry about your monitor giving your computer cancer.- -----------------
----------------------------------------
Well I'm 25 as well and that was my biggest question. Will I see it in my lifetime? Thank you for such a great response.
A bit of definition if i may:
by assembler do you mean mass production or just single unit fabrication?
"We hope you find fun and laughter in the new millenium" - Top half of fastfood gamepiece
"Fighting the underpants gnomes since 1998!" "Bruce Schneier knows the state of schroedinger's cat"
Well, yes, brute force could be faster. But then we'll just have to use those "micro-micro-micro" processors to find bigger primes. A processor can always encrypt better than it can decrypt. Currently encrypted stuff would suddenly be very ... vulnerable, thought. Makes you think differently about all those catch-phrases like, "using current technology..." or, "barring some incredible new advancment..." doesn't it?
I've lived through some changes, and all I can tell you is it's hard to see very far ahead. But if I were you I'd try to get some coursework in neural networks, genetic algorithms and fuzzy logic.
--
Time is Nature's way of keeping everything from happening at once... the bitch.
This is true, but could you imagine "growing" computers? And computers giving "birth"? I'd be all for that...it'd be weird, but talk about the end of WinTel....
Regardless of the hardware, they'll always need software jocks since advances in this area occur very slowly. So start out in hardware and migrate to software engineering as the technology changes.
Well, at the speed of light, would it not be possible to execute an infinite amount of instructions in no time what so ever?
Mikael Jacobson
I'm probably wrong about this though.
I will be the first to admit that I'm not clued in about either technology so I'm hoping the more clueful here can offer some insight. How does the performance benefit derived from such a device compare with what could be achieved from the research being done into optical devices (ie: those using light rather than electrons). It seems to me that if one of these two technologies gained favor, the research being done into the other would be lost. Is this the case? Is it possible to use the processes mentioned in this article to replicate an optical device? Besides eliminating heat emissions problems, I'm sure there are plenty of other benefits. Does anyone know?
Okay, so I lied, that stuff does interest me and I did look :).
:)
Yeah, some of that stuff looks like what my friends and I looked at and played with at RIT. Well okay, they played and I got to BS with them and see the wafers. Not really the same as doing it is it?
This space for sale
I could be wrong on this, but from what I understand the problems with replacing a retina are more concerned with the interface to our nervous system than any lack of ability to create an artificial means of input.
I'm not even sure that it would be possible to improve on the optic interface (by making more receptors with small components) since I'm pretty sure most of the way the brain processes optical information is pretty much "set". Maybe if you did it for a baby and they could grow up with it.
Unbreakable toys can be used to break other toys.
I think this has been around for a while; but for all of you who don't feel like signing up for the free registration:
Login: slashdoteffect
Password: slashdot
-- Imagine how much more advanced our technology would be if we had eight fingers per hand.
Current software will be trashed along with the Von Neumann machine architecture since the economics of molecules vs. silicon are significantly different. For instance, a fast pattern matcher could replace the ALU as the basic building block. We have no software to capitalize on this type of structure.
Bell's theorm is kinda' like this.
Basically if you take electrons spinning the same directions from the same atom and split them up and change the spin on one electron they spin on the other will change at the same time.
So you have a (theoritical) way to transmit binary from anywhere in the universe to anywhere the coorsponding electron is.
The first step to FTL communications?
Why the focus on molecular computing? (I guess to answer myself, it'd be an intermediate step).
I think that a more productive branch of research would be in quantum computing. (bye-bye binary. ON/OFF/BOTH will become far more interesting).
They have had logic gates for some time in quantum computing. (One of the left-coast universities, I believe).
A good starting link to find out about this stuff is http://www.quantumcomputing.com. You will also find information about quantum cryptography. I haven't looked at the crypto stuff yet, but by the fact that they have a section about it, it would seem that it would answer people's current questions about current encryption technology as embattled by molecular computing.
There was also a show about it on Science Friday some months back. (http://www.sciencefriday.com), although I couldn't find it in the archive.
And by then we'll be able to port ourselves from carbon to something more long-lasting...perhaps to whatever's supplanted whatever's supplanted whatever's supplanted (and so on...) silicon by then. :-)
--
"HORSE."
"HORSE."
-Flaming Carrot
That has crossed my mind also. I have concluded that it will make my most favorite fantasys come true(except the one about the blond).....we will have to find other places to live...fast. I see huge ships full of New Worlders heading out to all corners of this uni/multiverse. This is what we are here for...to see the continuation of ourselves. The only way we can do that is to get off of this single rock before it disappears. Overpopulation will quickly push us there. But I am not syaing anything a hundred fantasy scifi writers have said better and before I have.
It's fun just to think about.
Pop over to http://www.foresight.org and check out Robert Freitas' work on nanomedicine in the "what's new". Having processing power like that in a watch will be childs play. You could have them replacing your red blood cells, and still doing better at oxygen and CO2 transport than the original blood by an order of magnitude. I have artwork on my high-bandwidth homepage of nanotech bloodcells at work.
:v)
Vik
and why would it be unchristian? and should unchristian things be forbidden? Buddhism is unchristian. Atheism is unchristian.. planning to ban those too? well.. go ahead. I don't live in the US, but found your note funny
//rdj
No one can understand the truth until he drinks of coffee's frothy goodness.
--Sheikh Abd-Al-Kadir, 1587
I think the New York Times is really dumb for requireing a free l/p just to read stories. Anyway, heres and easy one to remember...
login: wheredoyou
password: wanttogotoday
Mark "Erus" Duell
Get a login and only select linux news. You happy: you only see linux news. We happy: we don't get more posts like this.
//rdj
No one can understand the truth until he drinks of coffee's frothy goodness.
--Sheikh Abd-Al-Kadir, 1587
" way the brain processes optical information is pretty much "set"." :) the brain could be taught to accept the new nerve input. Quite how you'd do that for retinal implants I don't know, but at least it shows that the brain can be dynamic in the way it operates.
I don't think this is actually true. I seem to recall people who'd had their hands / feet stitched back on following a mechanical injury recovering utility of the limb. Immediately following the injury & operation, the hand would be incorrectly "wired", so that trying to move your little finger would result in your index finger moving. But over time (with plenty of physioterorism
Strong data typing is for those with weak minds.
There's two differences.
(1) You're talking about repairing previously existing functions, whereas he was talking about actually adding new ones. The former is simple enough, given time. The latter, however, is not. This is the reason that, e.g., 'feral' children never learn to speak, or people born blind can't suddenly 'regain' their sight.
The brain is extremely flexible at birth, and grows progrssively less so as you age. This is a necessary: performance is improved through specialization, which inevitably results in a more rigid system. An adult brain is still tremendously flexible, but seldom sufficiently so to make any drastic changes. Adding new features is very definiately a drastic change. Repairing damage, OTOH, is a matter of reversing drastic changes, so natural it works quite a bit better. (For a related example, phantom limb (partially) the result of the brain not being able to adapt to it's new state, sans limb. I'm not sure if/how you can recover from PL, but it takes a very long time to do so.)
(2) It's not a matter of new input in the case of super-retina's. There no input whatsoever, because there's no optic nerves connecting the added features to the brain. You's have to totally rebuild the visual system, at least up to the LGN, to make it work.
To me, this was the most interesting line of the story:
The Clinton administration is now considering the possibility of a National Nanotechnology Initiative as early as next January to set up financing and help organize diverse research activities in nanotechnology.
Not that government involvement is always a good thing, but the field can only benefit if a government with the resources of the United States decides to provide official backing. Also, it will help to lend more credibility to the field, in the eyes of Universities and other governments that might not be as convinced of the possibilities as we are.
Of course, if this happens, in a few years we'll have to listen to Al Gore tell us how he invented Nanotechnology, just like he invented the Internet.
darren
(darren)
Okay, so...if they're using organic molecules does that mean they could be self-sufficant and considered alive? Hmmmm.....what happens when your computer can catch a cold?
Abort Retry ACHOO?
Also on a side note, anyone know the current status on organic nurel(sp?) nets? IE artifical organic brains?
Before the revolution takes place on the nano-technological scale, they need sufficently advanced tools to automate the processes inherient to nano-technology. If the building of machines is going to rely on a team of scientists using scanning-tunneling microscopes to put the individual atoms in their place, we'll never get anywhere. and machines on that scale will cost a buttload. the first nano assembler though.. whew. who ever creates the first decent one has got a fortune for themselves.
--
Gonzo Granzeau
Gonzo Granzeau
"Nothing the god of biomechanics wouldn't let you into heaven for.." -Roy Batty
Ok. We have hard drive size getting smaller and cheaper. Chip/etc size getting smaller. So just when am I gonna get my own 500mhz watch?
Oh yah. How long will it be until somebody writes the patch to cluster them?
I can see it now. All the geeks are wearing watches up and down both arms. All of them are running Seti@home. Team slashdot could kick some serious butt.
Click here to read too much about my personal life
I realize that CPU speed are supposed to double after a given amount of time.
Correct me if I'm wrong, but chip speed have a limit to them. For example, right now, speeds are held back by the material they're made from, copper and silicon. Wouldn't it be true that the fastest a chip could ever go would be a chip made from a couple of thosand atoms, with the electrical impulses in the chip running at the speed of light? Which brings up a question as to how close we are to the limit.
And once we hit that limit, would the only way to boost performance be to create paralell processing units?
So, I guess, really, what I'm trying to say is...
...WHEN CAN WE USE BEOWULF ON IT!?!?!
( OK, maybe that's not what I wanted to say. Maybe something along the lines of "When will we see any of this cool nanotech which seems to exist...but not really." There, that's better. )
aÍÍ©ÍÌÍ£Ì'̽ͩÌÍzÍYÌÍÌY
It seems to me that newer isn't always better, in the job market that is. Look at what a Cobol or RPG programmer makes today and what a C++ programmer makes today. The new stuff is in the news - the old stuff is what gets things done. Plus there are thousands of kids right out of school willing to work all day and all night on the new stuff for nothing while the old stuff is in great demand. I should have stayed with RPG on AS/400...
This is definately a step foreward in the nanotech department. I, as well as others I'm sure, am eagerly waiting for all that is nano to arrive
"I swear I won't break you if you let me take you where the willows never weep" -- Switchblade Symphony
Could someone with a bit of information in this area provide what they see to be a reasonable timeline for nanotech? I have been doing alot of pondering on technology growth and futurism lately and wondering how much I'll see in my lifetime, especially considering how much I've ALREADY seen in my lifetime.
"We hope you find fun and laughter in the new millenium" - Top half of fastfood gamepiece
"Fighting the underpants gnomes since 1998!" "Bruce Schneier knows the state of schroedinger's cat"
I wish I could be around 200 years from now to hear someone go "They were using silicon wafers for what?"
:>)
Speaking of silicon, I guess the valley will have to come up with a new name. Free beer for the best guest
Hates people who have stupid little sigs
Currently pundits are predicting the end of Moore's law with 9 micron interconnects. In my opinion, we'll see this limit as soon as 2002. If it takes til 2014 to ramp up manufacturing processes for molecular CPUs, this means that the CPU you buy in 2003 might actually remain state-of-the-art for an unprecedented 10 years!
The wait for tech support doubles every 18 months... Any likelihood they can solve your problem halves. Foosters
login: dammitbitch
pwd: dammit
The people working in silicon can just buy
the new technology.
I'm a bit familiar with the research at Rice, and basically they are building gates by deforming nanotubes in certain ways to alter their electrical characteristics. They've been playing with it and playing with it until they can build structure which act like logic gates. The implications boggle the mind.
I say its high time for nanotech logic to start ramping up, so this is very exciting. Silicon is only going to carry us so much further. Once you get down to a semiconductor gate that is 20 molecules across, the Physics get much more interesting, and electromigration starts eating your lunch.
A very interesting field to follow...
--Lenny
I like this idea... I think this will be the next revolution (since the Silicon) We might be looking at a new era in computing history... Kinda neat, what what?
---
---
Killroy Woz Here
I don't think it has anything to do with linux. slashdot.org -News for nerds (it doesn't say anything about "news for linuxusers")
Mikael Jacobson
I dont see anything on slashdot specifically stating "Linux News for Nerds." Just go back to bed and sleep thru this next century if you're not interested in it.
One day there will be something, and it will almost certainly be at the nanotech level where all the existing stuff breaks down .... but it's going to be a long ugly transition .... Si will hold on as long as it possibly can (kind of like how modem tech kept coming back after people predicted it' demise) I bet longer than most people predict .... and molecular level stuff's going to be initially unreliable (and probably get a bad rep as a result).
Me I'm still hoping for the non-electrical nanotech stuff and assemblers so we can get away from this 2d chip paradigm we're currently stuck in
I was reading in a science magazine a number of months ago (Discovery I think, before the new editor) about the so called "spooky nature of photons." I am not exactly sure how the expirement worked or how the scientists got one photon to be tied to another photon, but here is the jist of what happened.
They sent two photons along fiber optic cables at 90 degrees from each other for approx 20 kilometers (is this right?) Then they ran expirements with these photons such as stopping one photon in travel, etc. The spooky thing that happened was that when one of the photons was stopped, the other photon also stopped 100's of kilometers away at the same instant (in the mathematical definition).
What this has to do with computer speeds (if anything) is that if we could harness this power then we would have near _instantaneous_ data transfer and processing speed. Of course there would still be a little bit of light speed overhead if we made the chips transfer any light, but the possibilities are amazing. Imagine zero latency downloading! It would probably change the entire way that architectures are built.
If anyone has anymore info on the subject please post it. My memory may be flawed and so may be my data but this is what I remember. I'm not a physicist, just a programmer.
"I can hoist a Jack. I can lay a track. I can pick and shovel too. I'll do anything you hire me to." - John Cash "Legen
To quote Conan O'Brian's "In the Year 2000":
"Computers will become so small that millions of them will fit in a tablespoon. They will be used to enhance the flavor or soup."
Hee hee hee hee
P.S. Don't kill my Karma, I marked it off-topic in the subject!
"Software is like sex- the best is for free"
-Linus Torvalds
Load up on the math and physics. In the end everything is math and physics. If you have a good undergrad background in these you can go either way latter on. Figure your career is about 50-60 years, so stay broad Good luck
Some of the interesting stuff happening at Rice involves getting the buckytubes to behave as either metals or semiconductors, depending on the tiling pattern of hexagons. To build circuits with this, you'd need to be able to join tubes of different tiling patterns. This was an active area of research two or three years ago, I haven't heard more since. At the same conference, I heard that it was pretty easy (from a chemist's pov) to "functionalize" buckytubes, i.e. stick little molecules on there that did useful things. Probably that would be useful in joining tubes with different tiling patterns.
WWJD for a Klondike Bar?
Well, I'm a long-term optimist, so to me, being born is like buying computers... It'll always be better to do it later, but you have to do it some time. :P
Technology is neither Christian nor un-Christian. The moral and ethical implications of what people do with technology, on the other hand, is a different story.
You know, ten years is a long time to work at McDonald's -- just take a job that will pay for your Physics PhD...
Sounds like this tech will be more likely to be able to provide that. Also, quantum computing may well have limited applications (good at searching and sorting, lousy at running a Window Manager). We still need quicker, smaller, cheaper processors that work with understood logic to solve these macro-scale problems.
Still, I can see quantum processors combined with Turing-style memory readers and semi-mechanical procedural processors to make kick-butt systems that store and process their terabytes in the space of a shirt-button.
How many buttons can you sew on a shirt, anyway?
Well, increasing evidence points to the brain being more plastic than previously thought as far as forming new synapses. I believe you are correct that some training could be accomplished, but probably not as much as would be required to make full use of such an interface.
To take another approach, perhaps it would be possible to examine the existing interface and come up with something that mimics it well enough that the brain doesn't have to do a lot of rewiring to figure it out. For a damaged eye (using your example) if some of it was still working then perhaps it will someday be possible to observe the way the undamaged portion works and then improve upon that.
Creating a completely new and improved interface might be difficult. To do something like Jordy's visor in STNG would probably not be possible without training from birth.
Unbreakable toys can be used to break other toys.
It's not healthy for humans to have this kind of technology. It's very un-Christian.
ABC Nightly news (Nov 3 or 4) had video with interviews of Mark Reed and James Tour. They showed what looked like an animated video of the HP picture of a molecular wire shown in the Progress report article. They had the quote about making molecular computers becoming as simple as making photographic film. Transcripts may become available http://abcnews.go.com/onair/worldnewstonight/trans cripts/wnt_transcripts_index.html ABC news could also be contacted. Here is a link to a good paper by others working in the field of molecular computing. http://www.mitre.org/technology/nanotech/Arch_for_ MolecElec_Comp_1.html
CE's don't work with silicon directly anyway, you use HDL and other high level languages and tools to design circuits, which are then simulated in software and layed out automatically. Why would you think smaller general purpos computers wouldn't use logic gates anymore, regardless of how they are implemented?
Why is it that the anaylists keep saying that we can't continue to shrink transisters fast enough anymore to keep up with Moore's law? The relentless quest to make Quake run faster is enough for most engineers to make the impossible happen. This is a big step in microprocessor design, and when combined with the insulating properties of aerogells, will allow us to pack more and more transisters into tighter areas. Before too long, we'll be manipulating the state of a Quark to turn switches on and off. It's a wonderful time. :) However, one day.... In the mean time, sit tight and wait for the 10 cent processor.
Well, the quantum computer won't be available this quarter, unfortunatly.
Mike
Hmm... in the space of a single Pentium [23] we could have thousands of micro-micro-micro (that's 10^-18, for those who care) processors... what does this mean for cryptography? Brute forced keys in fractions of seconds.... *shudder*
b underscore geiger at hotmail dot com
o/~ All God's children shall be free in Pirates of the Caribbean, when we reach that Magic Kingdom in the sky... o/~
I'm interested in the possibilities of these minuscule gates to run on the tiny bits of power from a glucose/oxygen fuel cell. With some molecular photodetectors, gates like these could be used to make an artificial retina and restore sight lost due to age, injury or disease; with the tiny size of the gates, they could be made smaller than the cells they replaced.
Biomolecules don't seem to like heat very much, so really high-speed (and high-power) operation might not be the best use for these. But making up for it with massive parallelism, and taking hints (or outright copying) natural systems could lead us to a whole new world of technology that we might have trouble recognizing as computing. Still, I'm game for it!
--
Time is Nature's way of keeping everything from happening at once... the bitch.
Hemos, can you sit down with Roblimo and show him how to submit interesting articles rather than creating the dubious fluff he usually thinks up?
Thanks!
What really has me curious is the chemical/electrical interfacing. From the Times article I'm assuming that what they have right now is inorganic, so are the gates still transfering electrons? Does anyone know of work in this area?
Come now, you're being way to hard on fluff by comparing it to Roblimo's drivel!
W S B Knocking on heaven's backdoor...WSB
Naw. What you get out of a computer-related degree is the ability to think analytically about problems. This is something which will always will be useful to you.
They say that this technology will be dirt cheap to produce, but, honestly, what company will sell a computer that is that fast at a low price? If anything, they'll price the computers the same as higher end silicon machines and try to gouge us for as much money as possible, while they're raking in the dough. This especially applies for the first company that provides the technology, while there's no other competition.
Dr. Eldarion
- Running Windows on an Athlon is like driving a Lamborghini with the parking brake on. -
I work @ jpl (a sub-contractor for NASA) as a part-time student. We recently had a lecture here on a very related topic, of using atomic structures like "bucky-balls" (essentialy a chain of carbon atoms formed in the shaps of a soccer ball's black and white patches) to create near sci-fi materials. They could have conductive or insulating properties depending on the orientation of the atomic bonds.
Arthur C. Clarke wrote a book (whose name escapes me right now), about a crystaline structure that was strong enough to support massive amounts of weight, but was so narrow, and light, it was nearly invisible. This material would have allowed for suspending things in geo-synchronous orbit with the earth, almost like an anchor, and one could move up and down these things just like an elevator.
Cool thing is, this isn't nearly as impossible as it sounds, and with the development of nano-technology like that in this story, and that of the lecture, it really makes you realize that it's a great time to be alive.
Q: What do you think about American Culture?
A: I think it's a good idea.
(adapted from Gandhi)
I'm a first year student studying Computer Engineering. Every time I see a story like this, I start to wonder about my choice of major. I love the work I've done so far and the prospect of hardware design, but what use is there in continuing to study this if I'll be graduating at a time when silicon has about a decade left to go?
I've always entertained the hope of pursuing grad studies in either bio- or nanotech, and now, both of those are looking more promising than ever. What can I do now to better prepare mysyelf to specialize in one of those fields in graduate school? Do I need to look seriously at a new major or should I just concentrate on taking some specific courses?
Damn. If I end up getting that Computer Engineering/English double major I'm shooting for, that English degree could conceivably be more useful. Shudder.
-jay
If we do someday get a radically new hardware paradigm that is significantly advanced enough that there is no question that people will switch to it ASAP, might not that spell the death of architecture specific technologies?
For instance, take interpreted computing vs native compiled. For all those interpreted programs running out there on current hardware, to "port" them over to the new hardware, one only needs port the interpreter. To port a natively compiled program, the compiler and the program needs to be ported. All those programs written in interpreted languages will beat the others to this new market - by far! I think that may be the second most interesting outcome of a radical shift in the underlying hardware of everyday computers.
The most interesting outcome, might be a sudden dominance of the Open Source software model, for much the same reasons. If we're all forced essentially, to "start anew", OS might win in some key areas.
For instance, if a brand new operating system had to be developed from scratch to accomodate the new architecture, would an Open Source OS win right off the bat because of it's greater speed to market, so to speak? And it's speedier fixing of bugs and implementation of new features?
To ask it another way, if suddenly all PC's were replaced by Amiga's, who would be first with a stable, high-quality operating system: Microsoft, or the Linux community?
And what kind of software would be ported to it first? Java, Perl, Python programs, or C, assembly, Pascal programs?
First, make it work, then make it right, then make it fast, then, make it bloated!
If I understand the topic correctly, a mollecular computer should be make todays computer look like a turtle racing a ferrari. Current encryption methods are based on the theory that there isn't enough CPU power to use the brute force method of breaking the key. These would change that and could make e-commerce much less secure. I guess 56 bit encryption in the face of this would be laughable. Yes, I know we're talking about years before this is ready, but it's still exciting stuff.
There are plenty of architectural improvements yet to be made in processor design. Once feature shrinking becomes even more difficult, clock speeds may begin to change more slowly, but chips will still improve in performance. More effor will be thrown into developing advanced architectural features around existing gate technology.
And don't forget the software side of technology! As more and more software is written towards a multi-threaded architecture, the speed advantages presented by multi-processing and multi-threading architectures will become even greater.
One development I am watching with rapt attention is the transition to Simultaneous Multi Threading Processors (SMT). This is still in the works, but processors such as the Alpha 21464 will be built around this design in the near future.
SMT procs move some of the process table down on to the processor itself such that the processor can fill time while waiting for a cache miss to be serviced by task switching to a seperate thread. Further, SMT allows simultaneous dispatch of instructions from *multiple* instruction streams. This sort of architecture makes much more efficient use of parallelism in hardware than current superscalar processors. Further, executing on the same chip, the different threads can synchronize *in cache* which is far more efficient than hitting memory like current SMP systems are forced to do. Very exciting...
There's no way a processor company would sell the same part for 10 years. If they can't shrink their gates any further, they'll just find new ways to exploit parallelism in hardware with more advanced architectures.
--Lenny
What does this have to do with linux?
NEXT.