Slashdot Mirror
HP's Crossbar Latch... Next-Gen Transistor?
moojin writes "CNN.com reports that "in a paper published in Tuesday's Journal of Applied Physics, HP said three members of its Quantum Science Research group propose and demonstrate a "crossbar latch," which provides the signal restoration and inversion required for general computing without the need for transistors.""
Additionally, the crossbar latch can be locked across the steering wheel to prevent car theft.
Unknown host pong.
Some funding for the experiment came from the US Defense Advanced Research Projects Agency (DARPA)
Yeah. I'm betting that "some" of the tech used came from the same source as well. I mean, if it's been proven [tt] that transistors couldn't have been invented the all of a sudden way they were in 1947 (or even using today's technology), then how are they expecting us to believe that this new tech isn't reverse-engineered UFO tech? We're currently still miles away from acheiving anything in quantum computing, and now we're suddenly expected to believe that HP has this kind of working tech? Give me a break.
Man is a slave because freedom is difficult, whereas slavery is easy.
Does mean that I can finally replace the vacuum tubes in my computer? I am hoping for something that can fit in my bedroom.
if not then im not interested.
This post brought to you by Captain Obvious.
Take that, all you people who said that this year would be the year the moore's law beats us!
I think I speak for the sarcastic majority when I say this has little, if anything to do with Quantum Computing as defined thusfar.
..the Original statement by HP and even more important HP's paper in the Journal of Applied Physics.
This Sunday at the Engine Room
QUANTUM TRANSISTOR AND THE CROSSBAR LATCH
with special guests
SIGNAL
$10 cover, must be 21 to attend
I thought HP only did printers :)
A million monkeys and this is the best sig they could come up with...
I will be the first to admit that eventually there will be some limit to how small we can make a transistor (or transistor replacement) it seems that we still have a ways to go. I remember recently a spate of doomsayers going on about how circuits couldn't get much smaller than they are now, and how this would be the end of easy processor speedups. Well I guess they were wrong again. I don't think that will stop them from telling us that circuits can't be built smaller than this however.
Philosophy.
it is going to transform the way we computer simply out of the sheer computing power we'll be able to throw at things.
Yep. After that no more thinking will be required. Just brute force everything. Chess, unsolved equations, protein folding, you name it. We won't need science any more.
I work for a team in research at HP. This latch has potential, but it hasn't been fully tested. The PR dept just simply went off and decided to get everyone excited.
Just pray it ends up passing all testing and becoming everything they expect. Otherwise we might end up with an Intel-like pentium division problem on our hands...
Yeah, I'm a Republican AND a geek. It is possible.
How long before this goes the way of Alpha or HPUX or Bluestone or PA-Risc. Nice to see HP's still doing cool stuff; but I can already see Carly thinking "Hey, Dell doesn't have that kind of cost center, let's cut it".
Smaller, faster, and easier to build.
I must say, sexy, darn sexy.
In Soviet Russia, asses suck this joke.
Next thing you are going to tell me is that this will run Duke Nukem Forever.
Sigs? We don't need no stinking sigs!
We are dealing with quantum applications here. Maybe you should look a little harder. It's small after all...
If it works, would actual cpu-producing factories be able to implement it, or would it require a new process and new fabs to produce?
And they claim impressive (potential) performance gains... do the average computer user really need more than 4 Ghz? Or will the market for this new technology be supercomputer-class computers only?
Eureka Science News - automatically updated
EETimes story
It's Patent #6586965
An Indian-American Hindu committed to non-violent thought/speech/action alarmed by the global explosion of radical Islam
I got this in my computer world email subscrib, a bit more info then then CNN article. Computer World
Hold up, wait a minute, let me put some pimpin in it
See full text of article here
If this pans out to be viable, it will be interesting to see if it is promoted as a scientific (i.e. open) discovery, or a patentable (i.e. closed) invention.
Comment removed based on user account deletion
Except that computers already make heavy use of quantum mechanics - anywhere that things are very small, quantum effects come in to play, and it is the understanding of these principles (or at least, the spooky coherence of reality to our model) which lead to the development of the transistor.
Is this a mechanical gate or electrical?
You're wasting processing power by posting here on Slashdot!
Okay, I'm/was a SETI user also. However, I still think that there is plenty of evidence to support that we have yet to find intelligent life on this planet, or another.
Didn't Chevy invent those back in the 50s? What is this crossbar latch and what does it have to do with computery?
(Cackling: Eye of GNUt and hair of GNOME, give me root and get me home!)
-"...bad old ideas look confusingly fresh when they are packaged as technology" - Jaron Lanier (Digital Maoism on Edge.o
It'll be region coded. All the real power and functionality you want will be available in another region.
Life sucks and then you upgrade.
A feeling of having made the same mistake before: Deja Foobar
HP develops product with strange name and amazing powers!
People, check out their site. They do this kind of stuff all the time. It's research - not an actual product. Why aren't there stories like this every time they have a press release?
Check out this announcement that declares an extension of Moore's law for 50 years!
I hope you posted from home :)
Go here for teh [sic] funny.
So I can finally play Doom 3 at >10fps :)
Seriously though, forget protein folding and chess and the DNA computing crap, think about the 'quantum' enabled games we could get !! And ofcourse 9 new incompatible quantum 3d apis from MS, QuantGL and godKnowsWho.
Greetings,
I read the Press Release and this "has the potential to"... My guess is that HP are suffering at the moment (AIX machines are cheaper and more powerful than HP-UX ones, guess which we are buying less of) and this Press Release was published as a way of boosting the stock price.
Given that HP are dropping PA-Risc in favour of Itanium and that Intel appear to be dropping Itanium, HP seem to be dropping out of the large Unix market. I am sure that the PC Server market is good to them but surely diversification is the better way to stay competitive? Before anyone suggests it, there are some things that you just can't do as efficiently on lots of little servers that you can do on one larger server. For example distributed databases have locking issues that monolithic ones don't, and some of our legacy applications are still single threaded in parts.
Z.
-- Under/Overrated is meta-moderation, and therefore is Redundant.
Ah...that explains why my XP system does spooky things sometimes...
The NSA: The only part of the US government that actually listens.
As near as I can tell, what they've done here is implement levels of titanium and platinum nano-wires which pass each at right angle. However, to prevent leakage, at the crossover points they are held apart by Rotaxan molecules.
Rotaxan molecules are organic, and have this nifty little molecular ring which enables them to be conductive or not based on its position. Thus, you get your binary switch. This little animal is the "crossbar latch," apparently. And it can be done in something like 40 nanometers, making it scads smaller than current conductive strips.
Unfortunately, I'm having a great deal of trouble tracking down technical details. HP wants to keep its secrets, obviously, but Berkely and Stanford should be a little more forthcoming, think I. Anyone have links to more technical information? It would be greatly appreciated...
What he wants is more important that what I want. What he wants is also more important that what you want.
We will come up with harder problems.
Karma: -2147483648 (Mostly affected by integer overflow)
From nanoinvestornews:
A molecular crossbar latch is provided, comprising two control wires and a signal wire that crosses the two control wires at a non-zero angle to thereby form a junction with each control wire. Each junction forms a switch and the junction has a functional dimension in nanometers. The signal wire selectively has at least two different voltage states, ranging from a 0 state to a 1 state, wherein there is an asymmetry with respect to the direction of current flow from the signal wire through one junction compared to another junction such that current flowing through one junction into (out of) the signal wire can open (close) while current flowing through the other junction out of (into) the signal wire can close (open) the switch, and wherein there is a voltage threshold for switching between an open switch and a closed switch. Further, methods are provided for latching logic values onto nanowires in a logic array, for inverting a logic value, and for restoring a voltage value of a signal in a nano-scale wire."
From USTPO
"A novel switching device is provided with an active region arranged between first and second electrodes and including a molecular system and ionic complexes distributed in the system. A control electrode is provided for controlling an electric field applied to the active region, which switches between a high-impedance state and a low-impedance state when the electrical field having a predetermined polarity and intensity is applied for a predetermined time."
According to the article switch time is approximately a tenth of a second, or 10Hz.
Don't get me wrong, this is great and all, (see a better article at EETimes) but to implement microprocessor-complexity devices with single nanometer technology, we need single nanometer scale wires and the technology with which to 'draw' them onto silicon.
We already have enough trouble at 90nm with wiring, and it's only getting worse at 65nm.
This looks like a great leap in device technology, but we need similar advances in lithography to really use it.
No, no, no! Looking harder will simply collapse the function down to a single outcome and prevent it from working. The trick is NOT to look. :)
Compression too
But what would happen to poor old encryption? would it be possible to have keys long enough to prevent brute force and short enough to encrypt quickly?
I agree with your point, but I just think in case people don't know, this isn't true quantum computing, per se. Though the technology does rely on quantum mechanics, and the science will quite possibly lead to quantum computing, all this is is a better transistor. Think transistor is to vacuum tube as nanoscale latch is to transistor. A true quantum computer is actually an entirely different type of computer than we see today, even moreso than a ternary or analog computer is different than a binary computer.
i ntro.html>good article</a> on this:
Cal Tech has a <a href=http://www.cs.caltech.edu/~westside/quantum-
-Amalcon
Yeah -- but watch those papers fly out at the speed of light!!
Does anyone have a published white paper from HP anywhere to read some technical writing about this? I'm interested, but news sites just don't tell me what I want to know.
I do security
Now with keys of 2^google plex digit length!
InThane
It's not that we wouldn't need science any more, it would just change science radically. The problem space would change. We might solve all the old problems but then there would be a whole new set to work on.
It would be quite interesting for sure. I'm just imagining the computer games we could have.
The ratio of people to cake is too big
- Mean operations til failure: ~100
- Switching speed: ~100/sec
So they just need to improve its reliability by a factor of 10^16 or so, the switching speed only by a factor of 10^7 or so.so as Caltech. BTW htmlgoodies.com has good html href help too.
No Sig for you.!
Does this mean that in a few years you will be able to charge $2000 for a transistor audio amplifier to a dorky audiophile who claims that "crossbar amps does not sound as good as old technology" ?
Quantum cryptography of course!
yes >
I thought it was interesting that this device was patented (U.S. Patent 6,586,965) in 2003 on an application filed in 2001, so that the core technology isnt really new as the article implies. The patent has a better discussion of the technology actually used than the cited articles.
r ?Sect1=PT O1&Sect2=HITOFF&d=PALL&p=1&u=/netahtml/srchnum.htm &r=1&f=G&l=50&s1=6,586,965.WKU.&OS=PN/6,586,965&RS =PN/6,586,965
Here is a link to the patent:
http://patft.uspto.gov/netacgi/nph-Parse
Can we get a "-1 Wrong" moderation option?
For some reason this kindled the mental image of a giant gorilla throwing poop at a brick wall which suddenly transforms into a cure for cancer. I don't know who is worse off the person who thinks that sheer computing power will solve everything or me :-/
SIGFAULT
Since were talking about quantum computers here, why wouldn't they be using Quantum Encryption (which non-quantum computers can use too), which would be protecting data with the laws of physics (using quantum entanglement and other things on the actual photons/whatevers sent down the fibre/phone/whatever line).
Seems pretty hard to crack to me, well in fact.. it seems impossible.
Google for more info, I've been pretty breif since I'm not an expert on how it works.
I will be the first to admit that eventually there will be some limit to how small we can make a transistor (or transistor replacement) it seems that we still have a ways to go.
:-)
... which we're also nowhere near.
I knew all that research I did for my novel might come in handy one day.
The theoretical limits of information and computational density (based on quantum density limitations and reletavistic constraints on signalling, i.e. speed-of-light limits) are Bremermann's Limit and the Bekenstein Bounds, and we're one hell of a long way away from that. Practical limitations may be an order of magnitude or two less
The Future of Human Evolution: Autonomy
Most people don't know what quantum computing is and/or what it's good for. There are some things that quantum computing won't do very well at all. It isn't the end-all/be-all of computing like some folks think.
Funny that HP would be flaunting this. I read about Quantum based, nano gates in Nasa Tech Briefs nearly two years ago. Can't submit such things to here from there, as they don't keep their articles on-line.
Kinetic stupidity has a new brand leader: Allen Zadr.
Yeah, I was typing that kinda quick, and forgot to preview it. oops.
-Amalcon
http://www.cs.caltech.edu/cbsss/pdf/SniderG/NanoAr chI.ppt
"Can there be a Klein bottle that is an efficient and effective beer pitcher?"
Maybe I'm being troll like, but everytime somebody "announces" some dramatic breakthrough lately (medical, computing, so on...) We are always at least 10 years from actual stuff I might be able to buy.
It's unfair to tease me and then never come out with the stuff!
Once we figure out the basics, it is going to transform the way we computer simply out of the sheer computing power we'll be able to throw at things.
No, it won't.
Quantum computing (which has very little to do with the parent article) will change the way we think about computationally "hard" problems. Things like prime factorization, things like NP-completeness, things like cryptography.
But quantum computing will not replace the general-purpose Turing-complete model of computation we currently use. We will more likely see the idea of a quantum-coprocessor, something that you can interact with through a conventional CPU.
The problem with quantum computing involves the complexity of doing simple tasks... Yeah, it can factor absolutely mind-boggling numbers in one unit of time. It also takes that same one unit of time to figure out 1 + 1 = 2. The problem there involves the length of that unit of time - Between loading a state onto a set of qubits, them almost instantanously solving the problem, then reading the state off of them, you could have done potentially billions of cycles of normal CPU ops (no, I don't have a time-scale to quote for this, but I would consider it exceedingly optimistic to hope we eventually get it down to the millisecond level).
This development has so much potential because it points to a very, very major leap in the size of what we would currently consider a transistor... From 90nm, used by Intel and AMD's absolute latest mass-production facilities, down to a few nanometers. This means lower power requirements, faster CPU clocks, and much better areal density of functional units (getting down into the range of a few dozen atoms per switch, rather than hundreds of thousands at 90nm). The linked article also vaguely alludes to easier manufacturing techniques, but skimps on that one.
My first intent was to agree with you an mock the statement,which is a bit hyperbolic.
but is that not what has already happened? The raw power we have gained over the last 50 years has changed how we approach problems. New levels of tech change the way we approach science. No reason not to expect that it wont happen again. As others have said,"the space will change".
What do you say to the man that has nothing? Cast it away!!
On the other hand, imagine inlaid crossbar latches on a printing matrix. Higher accuracy... Imagine 4 billion DPI - full color?
Kinetic stupidity has a new brand leader: Allen Zadr.
FWIW, the entire Inca civilization rose, thrived and fell - without the wheel.
Can we get a "-1 Wrong" moderation option?
HP probably won't release mac drivers for this...
Care about electronic freedom? Consider donating to the EFF!
[...] how are they expecting us to believe that this new tech isn't reverse-engineered UFO tech?
Obviously (from the names) we're talking about the "transtator" of Star Trek fame.
Bones (or Kirk, or some other crewperson or person from another of the fleet) must have lost another communicator on one of the trips back in time to Earth, as he did on that Mob-run planet in _A Piece of the Action_.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Would the average computer user like to transcode video in less than realtime, or turn their home movies into DVDs without launching a compression run before they go to bed, or emerge -uDp world in a reasonable amount of time? Yeah, I can think of a quite a few uses for faster processors off the top of my head.
The great part is that once we actually have them, some smartass from Argentina will come out with a hot new technology that we'll all want but that won't run very well on our puny 4GHz systems. Then, some pundit on Slashdot will say that no one really needs a 10GHz quad-core processor, and someone else will prove that KDE sucks by pointing out that Mandlebrot translucency vectors can't be computed in realtime on their 3.6GHz Opteron and that power users should check out XFCE (with built-in Composite support).
Dewey, what part of this looks like authorities should be involved?
That's for militaries and governments, not for us. The way in which it works to prevent interception also works to prevent relaying... so unless you have a fiber line or laser link directly to whoever it is you are communicating with, it's useless as I understand it.
is not necessarily the invention, but the fact that HP has any research going on at all. I thought that they were still too busy dumping past inventions.
Hopefully, with all that power, I'll finally be able to run Half-Life 2, Doom 3 and Far Cry at full speeds...
Bah, I'll never use a nanoscale latch computer. My transistor computer sounds warmer and less sterile! Plus it distorts prettily...
JOIN US FOR PONG!
They'll keep working through an EMP that would fry semiconductor electronics. You can reduce the size and MTBF of your glass envelope system by large scale integration of circuitry on the anode. Fewer glass envelopes means fewer failures. Ideally, you could get the whole system in one envelope. Then just keep some spares to replace as needed.
Given that HP are dropping PA-Risc in favour of Itanium and that Intel appear to be dropping Itanium
Where in the world did you get that Intel was dropping the Itanium?
And yeah, they may be dropping UNIX (HPUX) proper, but check out their linux support.
PA-Risc/HPUX boxes had their place in time for things like CAD workstations and things. But a regular PC with a good graphics card is just as good if not better today. A company down the hall is switching from dual PA-Risc boxes to regular dual PCs and they could afford to rebuy new equipment every year to 18months with the equivalent money that they were spending on maintenance contracts alone.
I spoke with some reps from HP last week and there was no indication whatsoever that Itanium support was going away, nor Linux support.
This is all nice and all, but they only gave us a pointer to an article in money.cnn.com. Does anybody subscribe to the Journal of Applied Physics and has seen that article? How does it work, what kind of performance can we expect, etc?
A Google search turned up this paper on nanoscale circuits that appears to be related. It mentions HP's crossbar latch patent in particular. Interesting stuff.
Patrick Doyle
I mod down every jackass who puts his moderation policy in his sig. Oh, wait a sec....
And banks.
not for us... unless you have a fiber line
In decent sized cities, more and more businesses are using fiber lines. And oh by the way, it's becoming more popular in homes too.
Still sound far-fetched?
She is going to pretty steamed when she finds out there are a few people left not devoted to figuring out ways to get customers to buy more ink.
You were mistaken. Which is odd, since memory shouldn't be a problem for you
"Crossbar Latch" sounds so anachronistic! How about something cool and futuristic sounding like psychofraculator or something?
I might know what I'm talkin' about, but then again, this is Slashdot...
HP got a patent for this in July of 2003 (in patent #6,586,965), filed Oct 2001. What is news is that they've successfully demonstrated it working.
> Quantum crypto algorithms which is a subject I know very little about.
:P
Here you go:
msg XOR key
"...this thing is not a transistor... hence, end of moore's law."
If you look at the orignial paper, Moore is talking about "components," not specifically "transistors." There's no semantic reason why this couldn't continue to apply to the new technology.
I'm not sure "quantum mechanics" is exactly the right terminology for the principles behind transistor physics. At least not at the time of their invention. Although the principles behind energy quanta and states are utilized, there's no entanglement, tunneling, or other such "spooky" quantum effects going on as far as I remember.
But then again, maybe I need to go back and take a second read-through of the "Semiconductor Physics" book we used in our "Physics of Electronics" course in college.
you could simultaneously simulate a giant gorilla throwing poop at a brick wall, a stucco wall, a steel wall, and a glass wall, although you would have to stipulate the condition which would select the simulation results you actually get - such as which wall yields the largest splatter. That would obviously be the last calculation we would ever need.
(A lot of people don't realize that only one of the 2^n calculations is returned, but one can conditionalize which calculation is returned, so that ultimately only the "useful" calculation is returned. All of the other calculations are sent to our unlucky quantum brothers in other universes. In fact, I wouldn't be surprised if the cosmic backround radiation wasn't found to actually be such a residue from our lucky quantum brothers who have already developed quantum computing!)
((Yes, I'm just kidding.))
Ben Hocking
Need a professional organizer?
Intel has a good overview on what leakage is all about. Leakage has nothing to do with jumping wire channels, although the electric fields generated between one wire and another in small process geometries cause signal integrity problems such as noise and delay.
so unless you have a fiber line or laser link directly to whoever it is you are communicating with [Last clause ommitted by WaterBreath] Whooosh, it went straight over your head. The fiber line you use needs to go _directly_ to your recipient -- quantum encrypted data can't go through a relay or a router (read: go across the internet or any other packet switching network).
I want a new world. I think this one is broken.
But without a similar leap in software design methodology, we'll still be writing crap software. The more powerful the software, the more complex it will be. The more complex, the more bug ridden. History already demonstrates what happens when you give more computer power to programmers: eye candy that crashes a lot. All quantum computing will do will be to make your software crash that much faster.
I'm not being cynical here, I'm being realistic. I work on a large project, and software complexity is our achilles heal. Computing power is increasing faster than the ability to design software for it. The tired mantra of "garbage collection prevents all bugs" is wearing thin. The only software technology that manages complexity is "simplicity". Smaller is better. Standardize and decouple all components. Go the Unix route of tiny utilities doing only one thing but doing well. Eschew the Microsoft model of jamming more and more features that no one will ever use into a twenty year old code base.
Don't blame me, I didn't vote for either of them!
Quantum computing is supposed to remove the concept of transistors and use the spin of various subatomic particles to reflect numbers and number theory.
Or am I way off.
Unfortunately, you need a DIRECT fiber line between the endpoints of the message.
That is why quantum cryptography is generally only used to distribute the keys for more mundane cryptography. Essentially, they're using a OTP and distributing the key w/ quantum cryptography. That way they KNOW the key got to the destination unmolested, before they start sending the encrypted data. OTPs are theoretically unbreakable, if the keys are generated from a true random source(Hard to find.). The problem is that the same key is used to encrypt and decrypt, so you somehow need to get that key to everyone who wants to receive the message, and if any copies of the key fall into the wrong hands, it's officially cracked wide open.
Quantum computing is overhyped in its computing potential. It has been shown that quantum computers do not gain a computational advantage in these problems without an exponetial number of gates.
Interesting as well is that the factoring problem that made quantum computing famous has never actually been proven to be hard on a classical computer. This is to say, it is perfectly possible that factoring can be done on a regular computer just as fast as a quantum computer.
The issue isn't available fiber, the issue is you need a point to point connection for anyone you want to use Quantum Key Distribution with.
Fiber to the home = point home to point Phone CO. You could communicate securely with the Phone CO, and that's it (if even that, it probably goes to some intermediate relay before the CO or equivalent).
QKD (at least w/r/t BB84 style protocols) only works with an uninterrupted, unrelayed quantum channel (a single piece of fiber).
-Greg
But, he added: "This could someday replace transistors in computers, just as transistors replaced vacuum tubes and vacuum tubes replaced electromagnetic relays before them." Top of page
Someday, probably by the time their patents run out. Btw, whats the deal with saying that "HP" did it, rather then the researches who actualy worked on the thing. I'm sure there were plenty, but I'd try to at least get their names in (if I ran a big company).
ReadThe ReflectionEngine, a cyberpunk style n
I must say, I feel quite sheepish now. =(
I'm sure IBM wouldn't mind selling PowerPC to HP if they are short of a decent engine family ..
Yeah, or else the cat might die too.
Slashdot: come for the pedantry, stay for the condescension.
Notice I said "appear"... Itanium as a workstation processor has been dropped by HP (Intel claimed it was never meant to be a workstation processor if I remember correctly) and I have heard rumours (from Intel Engineers no less) that they are cutting back significantly on Itanium work (leaving just enough resources to fulfill their contract with HP)... I don't know if this is truly the case but that was why I said "appear"...
HP still have HP-UX on both PA-RISC and Itanium but the price/performance just does not compare with AIX on Power5.
You seem to have assumed that I was talkiing Dual processor workstations, I am thinking of the four to sixteen processor servers that we use a lot of.
Dual PCs (Running Linux or whatever) just aren't designed to be capable of the same levels of performance and reliability that the larger Unix servers are. The price for a fully redundant dual PC (server class hardware) is not much lower than that for an equivalent, non-Intel server.
Yes, maintenance contracts are expensive, but that's part of the package and you have to expect that. Older machines cost more for maintenance too, so upgrading to a newer machine can save you significant amounts of money.
HP never seem to give any indication that support is going to be dropped. But did you ever hear them talk about how they were going to incorporate Best Of Breed from Tru-64 and HP-UX into the newer versions of HP-UX? Have you heard them gradually backpedalling on those claims?
Z.
-- Under/Overrated is meta-moderation, and therefore is Redundant.
Sorry, I disagree...
Opteron is frequently described as being a better architecture than Itanium.
Itanium sales are improving but still rather lacklustre.
Intel have repositioned the Itanium in the market place. It's now a "high end only chip" whereas before it was supposed to be more general.
Intel have added a 64-bit X86 variant, if Itanium was all that was promised why would they have needed to?
Microsoft and Sun have both dropped proposed Itanium OS versions.
HP seem to be the only people still touting Itanium heavily. Actually, I found mention of a start-up, founded by the guy who originally designed the Itanium, planning to write software for it... But other than them...
Neither of us know for sure, but I see no reason for Intel to continue Itanium beyond contractual obligations with HP.
IBM Power5 appears to me to be a signifcantly better designed architecture than Itanium.
Z.
-- Under/Overrated is meta-moderation, and therefore is Redundant.
Not at all. Diode noise will do just fine.
Tom Swiss | the infamous tms | my blog
You cannot wash away blood with blood
Would the average computer user like to transcode video in less than realtime, or turn their home movies into DVDs without launching a compression run before they go to bed, or emerge -uDp world in a reasonable amount of time? Yeah, I can think of a quite a few uses for faster processors off the top of my head.
Sometimes I wonder how much of this "need" for faster CPUs could be eliminated by using more dedicated hardware, such as video encoders. For instance, if you're building a MythTV box, and you have three video capture cards, you probably can't find a processor capable of encoding all three outputs into MPEG2 (let alone MPEG4) in realtime. Maybe a dual Opteron could do it. However, if you pay a little more and get video capture cards with hardware MPEG2 encoders on board, such as the Hauppauge PVR-250, you could stick all three of those cards in a 1 GHz Celeron system and it should work fine, since all the system has to do is transfer all the data to disk using DMA, and the CPU is barely used at all. These MPEG2 encoder chips also don't require nearly the power that Opterons running at full-tilt require.
It is more difficult and more expensive to make dedicated hardware instead of just writing software, but for operations which many people have need of (MP3 decoding, MPEG2 encoding/decoding, etc.), it's probably worth it. So I wonder if there aren't other tasks that could be pushed off onto specialized hardware instead of consuming CPU cycles.
It was THEM. Oh, no, they're here!! *ZAP* ARGHHHHHH
How about "Heisenberg and the Quantums".
Now all we need is some singer called Heisenberg. *grabs white pages*
Aa...
So assuming the crossbar latch makes its way into general purpose CPU in 10-20 years, what're the odds we'll still be running x86-64(128(256?)?) on it?
Will this likely just end up as a 1 to 1 replacement for transistor designs (just denser) or would it potentially allow for different/"better" instruction sets? Hey, maybe they can make a chip fast enough that they could emulate x86 for all the legacy stuff.
Or will this just lead the way to miniturization so I can get all the current processing power of my computer into something the size of an iPod?
Here is the USPTO listing for this thing. HP (and a few other groups) have been working on this kind of stuff for a long time.
WWJD for a Klondike Bar?
Can i quote you in the next 30 years while this trend will be true?
It takes a man to suffer ignorance and smile
Be yourself no matter what they say
How many rural people actually farm? Almost none. Most of our food come from huge corporate-owned farms.
I grew up in the southeast. There's tons of rural people out there, and some of my relatives live very rural lifestyles. Do any of them farm? No. (Except for tobacco, which is about as useful for life as cocaine and heroin.) What do most of them do? Sit on their asses and complain about how there's no jobs.
Your farming excuse is downright pathetic. Most people live in rural areas because they grew up in rural areas, and refuse to change or move out even though they're dirt poor, have no job, and live in a shack.
Rural and semi-rural lifestyles are fine if you're tired of living in the city, and (most importantly) can afford it. Everything costs more in rural areas, except possibly for housing. Transportation especially is more expensive, because you have to drive so much more. Worse, it costs a lot of money to provide services to these areas, because the population density is so low.
One of the last areas in the US was just recently provided with landline telephone service, in Louisiana. The cost was tens of thousands of dollars per person, but this was paid by all the other people in the state with a subsidy. Why?
Why should everyone else subsidize poor, lazy rural people just because they don't want to move?
Far too much money in this country (USA) is wasted subsidizing services for people who choose to live in rural areas.
Perhaps because the export $$$ those rural areas generate go some way to pay for the huge import bills that cities generate.
Like all pain, suffering is a signal that something isn't right
export $$$? What export $$$? People in rural areas don't do anything! There are no industries there! There's very little farming, if any. There's very little manufacturing: it all went offshore years ago. Some of my distant relatives were recently laid off because the jeans plant they worked in closed up so it could move out of the country.
What are these export $$$ you talk of?
The amount of computing that can be called "brute force" varies enourmosly between different problems.
For instance take the problem that Gauss is presumed to have solved in elementary school: add all numbers from 1 to 100. Gauss realized that the sequence can be broken into 50 pairs, 1 and 100, 2 and 99, etc, every which one adds to 101, so the answer is 101*50=5050. That's one example where we know one simple logical solution that takes much less computing effort than the more obvious addition of every number in the series.
OTOH, there are many other problems for which the only logical solution known requires a lot of computation. One well-known example is the four-color map theorem. Do you really believe that this theorem has been proved with "just" brute force, without the use of "science"?
I'm not an expert in the field, so I might be talking bullshit here, but as I understand it, there are problems in group theory where one can demonstrate formally that the only solution requires an amount of computation that's suspiciously close to "brute" force. The four-color map problem I mentioned required 1200 hours of computation when it was first solved in 1975.
and no dating, even
I doubt that we will ever figure out - and I suspect that even if we did figure out we couldn't do much about it
Of course, we won't be around to enjoy it. Somebody will complete the Towers of Hanoi and the world will end.
Haven't you ever seen two dogs meeting in the street? The first thing they do is to smell each other's ass. That's because a dog's ass has an absolutely wonderful smell! It must be true, how could a hundred million dogs be wrong?
Though if you're a neutron, it's free of charge.
Escher was the first MC and Giger invented the HR department.
Crossbar latches just don't have the natural warmth of sound that transistors do.
Unless, of course, you take pains so you can trust each relay point.
... but it's been a long while since I had stats class, and I could be wrong.)
But even if you maintain higher security at the relay point than you do in your own organization, the overall effectiveness of your security will drop. I don't care how secure each point in your network is, statistically, each added node will reduce your security. If you have 11 nodes at 99% secure each, your whole network is only 89.5% secure. (100*0.99^11
tasks(723) drafts(105) languages(484) examples(29106)
My brother, who is a PhD student at Caltech, told me about this article and indicated that his group has been partnering with HP for quite some time on this project. He personally is working with these latches on a daily basis and the group is currently investigating ways to produce them reliably on a mass scale. However, apparently, HP is claiming most of the glory in the press releases.
What are these export $$$ you talk of?
Primary Industries of all types... usualy happen in rural areas. I'm not sure what primary exports are worth to the US... probably a lot less than other exports, however, if you divide the exports by the population base, I think you'll find that the rural areas contribute a lot to the wealth of the cities.
Like all pain, suffering is a signal that something isn't right
See... one of the alien guys liked to collect old radios and technological antiques from his own cultures golden age and stuff. So first off we reverse engineered an old walkie-talkie and got the transistor. Then there was the microwave oven (it was strapped to the cubicle by period-accurate velcro, but of course that was generic hook-and-loop tape) and got the led display and microwave popcorn and the automatic lazy-susan from that. Then we moved on to the cute little sterio and got the CD, DVD and CSS/DRM (because the aliens would have never made it to the stars without Falmatrazze Spears music kept properly safe).
Now that they finally figured out how to get the elevator to work on the ship, we've gotten down to engineering and got a good solid tricorder opened and stole VLSI chip manufacture.
With the tricorder mostly reassembled we got the cover off the coffie machine and (and we all know how much aliens like their caffiene) and that super advanced machine is where we are getting all this hot cross-bar stuff.
Once we get some good alien caffiene in our systems (if we can find a frrozbat quarter to put in the machine) we will be wired up enough to give the shop manual a good looking over (it's worse than any VCR manual _ever_ conceived of by man) and we'll finally learn how to get the quantum computer (which was actually stolen form the elevator controls) to return an answer other than "basement".
See, you just have to know how to walk the tech tree. Once we have finished the earth-orbit section, we will send advanced teams to mars where they will find another alien ship, and they should have the transistor [the _martian_ transistor] for themselves to use in that round as soon as they can get 100 units of Vespene...
[And remember: watch out for the plutocrat-rush and the humanists will always send a hacked indistructable peon over into your base to try to get you all involved in a debate about abortion...]
Innocent people shouldn't be forced to pay for inferior software development.
--"Code Complete" Microsoft Press
You mean computer games like, uhm, let's build a real-time simulation of a solar system complete with an inhabitable planet and people and ...
generating the personalities of 60 billion people would be something of a computation problem. It doesn't need to do it fast, six thousand years will do, plus or minus a few hundred million. Simulated time, of course.
Or, here's an even better game -- tune your big bang.
It'll be region coded. All the real power and functionality you want will be available in another region.
Worse yet, there'll be no way to tell both how fast your system is running and what region you're in, at the same time.
All employees must wash hands before seeking equitable relief.
Computers do not make use of quantum mechanics. I repeat - COMPUTERS DO NOT MAKE USE OF QUANTUM MECHANICS. What they do make use of is the "whoa, wouldn't it be hot if it were smaller and faster and OH MY GOD WHY ARE THEY DOING THAT" princible. Quantum mechanics are why we can not (currently) make chips infinitely small. Transistors pipe many many electrons at once, and they hope that none of them are entangled or doing any other wierd stuff.
www.olin.edu
No. It's not about trusting the relay point. It's about the fact that it can't be relayed. Pick up a copy of Scientific American Jan 2005 for more details. Or just read an article on quantum crypto.
I want a new world. I think this one is broken.
the alien _is_ "god".
BBN has developed a quantum router. (www.bbn.com). These are the same folk who invented the internet. It does unfortunately require trusting the router. Therefore, you are partially correct - quantum encrypted information is only as secure as its routing points.
www.olin.edu
I think you misunderstood me.
Imagine a q-cryto link between you and your relay point, and another q-crypto your relay point and your destination.
I don't know much about fiber optic networks, but I imagine two 50-mile fiber-optic links is going to be a hell of a lot cheaper than one 100-mile link. Some businesses may find the lower short-term investment worth it, even with the long-term security cost.
tasks(723) drafts(105) languages(484) examples(29106)
Does anyone know any details about this new design? Is it similar to the FinFet design ( http://www.spectrum.ieee.org/WEBONLY/wonews/mar03/ multigate.html ) or is it something totally new that has an interesting design? Does anyone have a picture of this crossy latchey thingy? how easily can it be intergrated into common present industry models?
I want answers, not more questions...
-=gabe2=- macbook dual 2.0
Uh... perhaps you should do some research. Quantum mechanics is not a dump box for everything in physics which is 'weird' or 'cool', it's the formalised theory combining the principles of wave mechanics and matrix mechanics, the two equivalent and independently discovered ways of describing a physical system. Nothing USES quantum mechanics - it's not real. It's a theoretical model of the world. Most electronics, especially semicondutor physics uses the principles from QM to achieve things that we otherwise would not have believed to happen. I'd be very interested to hear why it's QM that means we can't make chips infinitely small and not something more like thermodynamics - infinitely small implies infinite information density...? Why would it matter if electrons became entangled in a transistor? How do you think a transistor works?
Make up your mind. Either he's logic or isn't consistent.
Sincerely,
Pan Tarhei Hosé, PhD.
"Homo sum et cogito ergo odi profanum vulgus et libido."
They are in fact not smelling their arses but their genitals which indeed have an absolutely wonderful smell for dogs just like human genitals have an equally wonderful smell for humans. (Or so I've read.) The anus is being smelled as a side effect and its smell doesn't differ between both sexes so it is not very helpful as a guide to copulation. (Dogs don't usually make anal sex.)
Sincerely,
Pan Tarhei Hosé, PhD.
"Homo sum et cogito ergo odi profanum vulgus et libido."
Then we will finally have some free time to get on with learning grammer.
sudo ergo sum
I'm not sure you could really say that opteron is a better architecture than itanium, but what you can certainly say is that the computing landscape is much more ready for AMD64. Itanium 2 is clearly a very nice bit of technology, it's just not what people wanted or needed, and it suffers from the lack of software support.
You can't win Darth. If you mod me down, I shall become more powerful than you could possibly imagine
Finite problem, finite solution.
Know your pads. One time pad: good for cryptography. Two timing pad: where to take your mistress.
I don't give a damn who this person is... if they have something valuable to say, they have a right to say it, and ad hominem attacks, regardless of their veracity, do NOT qualify as informative!
The sticking points for the technology which have to be worked out are the lifetime of the devices, and their switching speed. Currently, the devices only work for 100s of computing cycles, and switching speed is many thousands of times slower than silicon technology, comments Professor Moriarty.
Quotyed from a somewhat better news source. CNN sucks, dump it if you want to keep your neurons.
Be faithful to your obsessions. Identify them and be faithful to them, let them guide you like a sleepwalker. JG Ballard