How To Work A Quantum Computer
Rather than moving Orion, D-Wave is developing remote access software. Writing quantum annealing algorithms for solving binary quadratic programs on Orion is very different from classical programming. You can work directly in the system’s machine language, directly choosing the current flowing on the input lines on the chip, but Rose expects that will only appeal to scientists studying the way quantum computing itself behave. Demos like the image matching system are written as problems in what he calls ‘industry standard ways to state combinatorial optimization problems’. A conventional computer converts that for Orion so that a solution corresponds to a pattern of current in the qubits that takes up the minimum amount of energy – the annealing.
“It’s like trying to find the lowest point in a valley when you have a ball and you let the ball go; it can find the lowest point by finding its natural state,” Rose explains. “It’s really easy to learn to use the system at this level, but figuring out how to recast the problem you really care about - say image matching - to use this new capability is very hard. Typically the folks who "get it" at this level have PhD-level discrete math backgrounds applied in an industry setting.”
Rose wants more people than that using Orion; “I’m a big fan of opening things up as much as possible to anyone wants to use them and making them easy to use even if people don’t understand quantum computation.” For the rest of us, D-Wave has produced a compiler that means programmers can state problems in SQL using a new FIND command. “This level of access allows anyone who is an expert database programmer to begin using the system within about 30 minutes,” claims Rose, “most of which is in learning the syntax of the FIND command, which is very similar to SELECT.” Developers who want to prototype applications or get familiar with the interface can try the programming model out with D-Wave’s Web service solver, although this currently runs a software emulation rather than sending commands to real hardware.
D-Wave’s demonstrations have generated plenty of controversy, partly because the company hasn’t published the kind of information that comes out of less commercial concerns in peer-reviewed journals. Critics suggest that what D-wave has is an analog computer that won’t achieve quantum performance; Geordie Rose believes time will show it’s a true quantum computer. “It’s known that there are several universal models of quantum computation, just like there are several universal classical models. The one we picked, adiabatic quantum computing, has significant advantage in that it’s easy to implement with large qubits.”
Using semiconductor manufacturing techniques means D-Wave can create a new version of the processor every month and keep tweaking it to fix any problems. And he’s confident there’s enough demand that we’ll see a usable quantum computer within years, not the decades some predict. “There’s a huge push from business and technology and that push is not going to go away any time. This is not like super high temperature conductors or fusion; this is something that’s going to be pushed until it works.”
- 1 / 3
- Next
-
Latest Miscellaneous News
- 05/28 – Samsung's First PIN Pop-up Stores Hit London Tomorrow
- 05/28 – UK's New Cookie Law Is in Effect (But Lots Aren't...
- 05/28 – Intel Picks UK for Research Institute on Sustainable Cities
- 05/27 – U.S. Hacks Al-Qaeda Affiliate Website in Yemen
- 05/25 – Specs for Dell's Windows 8 Clover Trail Tablet Leaked


LIKE A GLOVE!
I WANT TO SEE THAT THING PWN SOME GAMES >< I KNOW A LITTEL OVER BOARD BUT I CANT DREAM CANT I?
dam you beat me to 1st post ?> o wait that was me i just was to fast about posting i did not sigh in o well i cant prove it now but i know it was me >< LIKE A GLOVE
oops i said a bad word dont report me >< any ways the 1st post was me i was to quick to sigh in and yes i have made 3 post so sue me. LIKE A GLOVE
I still wish someone would explain exactly how quantum mechanics and quantum computers work. Do they actually defy the laws of physics or what? I didn't think that was possible.....
I meant to say "in layman's terms"
Very cool, i'd also like to see that pwn some games. However judging by what they say, it'll be very difficult to do if it's even possible.
BTW what does it matter who got first post anyway? lol
This is some wild funky alien technology sh#t... Yeeeehaaa!
Quantum computers process information with the "electron pairing" effect they naturally operate with, meaning it operates in such a way when one of the 28 bits is "doing something", it affects other parts of the chip in such a way they "do the same thing" naturally...it has very deep implications because it would basically SCALE LIKE CRAZY....NATURALLY....we are still in its infancy w/ 28 QBits, i dont even immagine what is will be like with 512...1024....a Giga QBit.....a Tera Q Bit.... they would all scale naturally... whatever we an do now whit it will be done faster and better when wider versions of the chip come out, withought the need to re-coding i immagine (Totally different from the way we program now with the current multi-core chips needing special programming to take advantage of more resources).- my 2 centz.
Quantum computers process information with the "electron pairing" effect they naturally operate with, meaning it operates in such a way when one of the 28 bits is "doing something", it affects other parts of the chip in such a way they "do the same thing" naturally...it has very deep implications because it would basically SCALE LIKE CRAZY....NATURALLY....we are still in its infancy w/ 28 QBits, i dont even immagine what is will be like with 512...1024....a Giga QBit.....a Tera Q Bit.... they would all scale naturally... whatever we an do now whit it will be done faster and better when wider versions of the chip come out, withought the need to re-coding i immagine (Totally different from the way we program now with the current multi-core chips needing special programming to take advantage of more resources).- my 2 centz.
basically, a quantum computer does not have to do "work" to get an answer. The "work" it does involves electrons and atoms. The problem is, is that when the "computer" reads the atoms/electrons, they are destroyed. So, the data can only be read once. But as mentioned in the article, code cracking would be very easy, as the computer would only have to work with one atom. Basically, the future quantum computer will be able to crack any code, instantly.
"I still wish someone would explain exactly how quantum mechanics and quantum computers work. Do they actually defy the laws of physics or what? I didn't think that was possible....."
Quantum Physics are laws that are far more complex than regular everyday Physics. They do not defy but rather redefine and enhance our understanding of physics at a much smaller scale (Smaller than atom size)
Games aren't really an area that would benefit directly from Quantum computing. The benefit is for other fields, like mathematics, communication, cracking encryption, physics. Quantum does not mean faster, it means they operate in a fundamentally different way and have different applications. Just like the stack of IBM QS22 Cell Blades next to me (Which don't even have video cards), they are for a totally different purpose and in some cases even if they could run a normal game might even be slower. High-Performance-Computing is a very different field, and it has to be specifically programmed for, which is why the article mentions they are advancing the new SDKs to allow things to run.
My computers don't even have a real "OS", they have micro kernels which only have one program installed, and that program is specifically crafted to fit inside the 256KB of L2 Cache on the IBM Cell Processor to prevent the system having to page regular, slow, system RAM.
The stuff is written in C, and then hand-tuned in assembler.
That does not mean the development is not exciting, its just that like most major science changes, the action is far from the end user and it will take products a while to trickle down to something we use everyday. Ironically, by the time it gets to us, most people will have no idea which technologies brought them to fruition in the first place.
I wonder how fast an E8500 could run at -270ºC.
I wonder how fast an E8500 could run at -270ºC.
You might have a good idea when the OC competition ensues. I'm sure you'll have some guys throwing -200C nitrogen at 8500's.
One question: Does it overclock? That cooler should at least allow for a 30% overclock...
I want one!
this is all good and fine. Will it play Crysis
Intel will buy them in after they make all that profit from Nehalem and release quantum computers for commercial use as soon as they figure out how to make a "safe", commercially viable, 3 Kelvin refrigeration units. Coolermaster will then release a 2 Kelvin unit for overclockers 3 months later...
Unfortunately it will NOT be able to play Crysis at 60FPS.
Reaching temperature as mentioned above would lead to super conductivity . Which people are yet to incorporate in daily usage . This is no area for child's play (games)
maybe the crysis problem wont be solved but i bet that i will finaly see
duke nukem 3d run at 954,555 fps!! that is, my friends, is what they call... THE S#$%!