Quantum computer to be presented the next week

santyhammer

Newcomer
I don't know if this is the correct forum for this ( pls move if not ) and could sound like Start Trek but I think could be the most important news in years....Well, here is :

It appears an enterprise called D-wave is going to announce this week the first quantum computer with a LIVE demonstration ( with two boring scifi apps ).

img_5568-2.jpg


( is that white thing TEFLON? used to seal well cables from the helium gas? I think that's the refrigerator unit )

lef-plate-tunnel.jpg


( mechanism to filter the input electrical noise? )

sample-holder-with-europa-chip_small.JPG


( basic socket, notice the sockets for refrigeration arm and electronic pins/wires )

img_0748.jpg


( a detail of the CPU die... core is made of niobium and quatz? Can work without an air-protection cover? )

20070110_d-wave_orion_processor.JPG


( functional schematic... what are those letters in the middle? We can see the 16qu-bits here running in parallel like any GPU stream processors. Why each unit is surrounded by 8 collectors )

All is about speculation and most scientists are quite sceptic, but the demonstration will take place the next week as is announced in

http://www.dwavesys.com/
http://dwave.wordpress.com/
http://dwave.wordpress.com/2007/01/19/quantum-computing-demo-announcement/
(diagrams, comments, etc )
http://www.wired.com/news/technology/0,70001-0.html
http://money.cnn.com/magazines/business2/business2_archive/2004/08/01/377387/index.htm

It appears it runs at absolute zero temperature, CPUS are made of "niobium/quartz" and an open-and-free Java/C SDK will be released too to program it.
With enough qu-bits ( you can add it VERY easy ) you could be able to perform gazillions of processing power ( yes, GI/raytrace at realtime. The author talks about "monte carlo ray simulations" ).

We are D-Wave Systems, Inc. The Quantum Computing Company. D-Wave is the world’s first — and only — provider of quantum computing systems designed to run commercial applications.
Please join us in February as we demonstrate a technological first: an end-to-end quantum computing system powered by a 16-qubit quantum processor, running two commercial applications — live.
For your convenience, we will host this event in two locations: Silicon Valley and, two days later, in Vancouver, B.C. near our home offices. We look forward to seeing you.
More than 300 international press pass have been given.
Im absolutely sceptic about this, but if this is true the world gonna change suddenly.......... you, know, Star Trek :p

More info at

http://www.eetimes.com/news/latest/showArticle.jhtml?articleID=197004661
http://www.hpcwire.com/hpc/508991.html
http://www.techworld.com/opsys/news/index.cfm?newsID=7972&pagtype=all
http://uk.theinquirer.net/?article=37536
http://www.freepatentsonline.com/6495854.html

Time ago somebody asked in a thread if we could perform GI illumination at realtime... well with this is apparently possible hehe...

Anybody could explain how quantum works? I think is something like "the bit is in state 0 and 1 at the same time" because simulation takes place in sepparate universes. Well, see:

http://en.wikipedia.org/wiki/Quantum_computer
http://www.quantiki.org/wiki/index.php/Category:Introductory_Tutorials

ps: I'm editing the post a lot to reflect more accurate info as I get it.
thx
 
Last edited by a moderator:
if that's true, how can they run all of that on a 16 qu-bit computer?
it sounds absurd, kind of life running quake on a ZX80.
 
No way are those claims real. We are a loooooooong way from having functional quantum computers. I'll believe the 16 qubit QC, but not the part where it runs anything meaningful.
 
Yes, the prototype works near -273C, but I read they expect to raise the working temperature ( a vapochill like http://www2.asetek.com/default.asp?showPage=startside.asp&contentSection=2 will be enough for this ) by 2008 and increase the qubits to 1000 which could lead in an unimaginable ( really ) horsepower. Actually they use 16 and is capable to search a molecule hash from a 1E24 database in a short time ( thats one of the apps they are showing it appears ).

The amazing thing IF THIS IS TRUE, is that to add more qubits to the system is really really easy. The difficult thing was to develop the base technology.

I added some photos btw, looks.... area51... I still can't believe it!.
The interconnection wires look to be made of an strage component haha, so amazing... Could we send a Beyond3D person to that event? It can be interesting, please!
 
Last edited by a moderator:
Bahh, it would be neat, but chances of this being anything meaningful are...how shall I put it...umm...slim;)
 
Yes, the prototype works at -273C, but I read they expect to raise the working temperature to -70C ( a vapochill like http://www2.asetek.com/default.asp?showPage=startside.asp&contentSection=2 will be enough for this ) by 2008 and increase the qubits to 1000 which could lead in an unimaginable ( really ) horsepower. Actually they use 16 and is capable to search a molecule hash from a 1E24 database ( thats one of the apps they are showing it appears ).

-273.15 or -273.0? There's a definite difference.

The amazing thing IF THIS IS TRUE, is that to add more qubits to the system is really really easy. The difficult thing was to develop the base technology.
They haven't explained how it is they can feed the amount of data needed to just 16 qbits, let alone something like 1000.
 
-273.15 or -273.0? There's a definite difference.
Yep hehe! I think a temp near absolute zero to produce some kind of superconductive effect. I doubt any technolody in this planet could stay at pure-absolute-zero heh! The main arm on the photo #1 uses hellium to cold the CPU ( it looks like the vapochill tube ).
Here http://stochastix.wordpress.com/2006/09/29/d-wave-the-quantum-computing-company/ we can see is -269C really, sorry for the errata.

They haven't explained how it is they can feed the amount of data needed to just 16 qbits, let alone something like 1000.
I agree. In the comments of their blog somebody mentioned is easy to add more qubits to the CPU and cheap too ( in fact they planned 1000 qubits by 2008, have 16qubits working now )

Other thing they should explain is how to fetch the results from the parallel universes. As you can see here http://en.wikipedia.org/wiki/Multiverse_%28science%29 is that the universes are independent, you can't modify or access a data from an universe different than yours... but well, I'm not a physicst and is only a theory... It appears they use something called "quantum tunneling" described here http://arxiv.org/abs/cond-mat/0412604

I hope they clarify all this in the Feb 13 conference, but if they are providing a live demonstration gonna be interesting. Personally I'm gonna download their C SDK when is available and see how monter-carlo raytracing can be done. If these chips reaches the consumer market ( it appears this is what d-wave wants ) gonna start a CUDA/CTM vs this fight haha! Well... all is unclear atm !

Btw, if this progresses and is not a fake, like somebody mentioned in the links, this computer could put at risk the current security algortihms, which is preocupant.
 
Last edited by a moderator:
Wow, cool :) I remember reading about the first Qbit a few years back.
If its workable, thats totally sweet.

Should be interesting to see what sorts of apps they can actually run.
 
http://dwave.wordpress.com/2007/01/

"The base temperature at which we operate (5mK, or 0.005 degrees above absolute zero) is about 500 times colder than interstellar space. In other words the difference between interstellar space and the base temperature of our fridges is about 5 times greater than the difference between room temperature (about 300K) and interstellar space (about 2.7K)."
 
Wow, cool :) I remember reading about the first Qbit a few years back.
If its workable, thats totally sweet.

Should be interesting to see what sorts of apps they can actually run.
I think they are showing two applications in the show:

The first application is a pattern matching application applied to searching databases of molecules. This is an app that we developed internally at D-Wave. This is an example of how to apply Orion to problems arising from association (or conflict) graphs.


The second is a third-party planning/scheduling application for assigning people to seats subject to constraints. Anyone who has tried to plan seating arrangements for a wedding should be familiar with this one. This is an example of applying Orion to constraint satisfaction problems.


The first apps is made using their own Java/C-compiler specially optimized for this. It appears you need a special compiler to make programs. They said this SDK will be available and free in some days.
It's logical, but I wonder how they can convert the qubits to 64, 32, 16 and 8 bits C current approach and how to fit their SDK inside an existing application.

http://dwave.wordpress.com/2007/01/
The base temperature at which we operate (5mK, or 0.005 degrees above absolute zero) is about 500 times colder than interstellar space. In other words the difference between interstellar space and the base temperature of our fridges is about 5 times greater than the difference between room temperature (about 300K) and interstellar space (about 2.7K).
All the information is confusing atm. In some places I've heard about -269C liquid hellium. Other indicate -273C. Other that the 2008 prototype will run at a higher temperature ( cascade vapochill ftw )... We need to wait for the official description. They are also changing their web, perhaps that refers to an old prototype but that http://www.leidencryogenics.com/index.asp sounds scary...

The event registration is now closed :( We can sign for the video content, schematics, etc in their old web it appears
 
Last edited by a moderator:
I wonder how they can convert the qubits to 64, 32, 16 and 8 bits C current approach
Well, 16 qubits can contain 65,536 simultaneous states so that's equivalent to a 64bit Float isn't it?
 
even if all the claims were true, probably still not enough power to do full-blown raytracing in realtime @ 60fps with movie-grade graphics. ;)
 
Well, 16 qubits can contain 65,536 simultaneous states so that's equivalent to a 64bit Float isn't it?
No it is equivalent to a 16bit float.

The real question is what is the speed like because I recon this has a lot slower clock and IPC then most processors. Really 16 qubits is fairly limited that should allow them a substance with 16 electrons really well but that isn't many electons. Most classical problems can if they can be done in parrallel you'd prolly be better going out and just buying a bunch of FPGAs. Obliviously if they can double the qbits it things are getting a lot more interesting.
 
Last edited by a moderator:
No it is equivalent to a 16bit float.
Oopsie, been a while since I paid much attention to that sort of thing :oops:

Anyway, clockspeed is largely irrelevant.
A quantum computer is all about doing stuff that supercomputers take hours or days to do but does them in seconds.
For stuff that takes less processing time then indeed, you might as well stick with a normal CPU.
It could be measured in minutes per cycle & still come out ahead.
 
Back
Top