next »

[perkiset]

Man this stuff turns me on, and makes my head hurt all at the same time. PopSci shows us an example of the first r'reals quantum computer chip. I think it does only a single bit (hmm, they have a ways to go...) but the physics of even this little advancement and the fact that they have it in hardware is DAMN amazing.

Check it: http://www.popsci.com/technology/article/2011-09/two-key-computing-advances-bring-quantum-computers-closer-reality

[kurdt]

Man this stuff turns me on, and makes my head hurt all at the same time. PopSci shows us an example of the first r'reals quantum computer chip. I think it does only a single bit (hmm, they have a ways to go...) but the physics of even this little advancement and the fact that they have it in hardware is DAMN amazing.

Check it: http://www.popsci.com/technology/article/2011-09/two-key-computing-advances-bring-quantum-computers-closer-reality

You mean qubit? Actually quantum computing brings in some serious problems for current technologies. Like for example it can be used to crack encryption way, way, way more faster than normal computer. Of course another thing is that what exactly are you going to use qucomp for? 3D rendering? Games? Encoding? But this also means that everything needs to be rewritten unless they actually manage to do some sort of mediator that takes existing code, converts it for parallel processing and then computes. There hasn't been a proper solution for this type of stuff even when there's clearly a need because of multi-core CPUs. And how does qucpu work with memory? Memory is still running on single bus doing one single thing at the time. Are we going to witness similar situation than what we now have with harddrives? Is there even theoretical technologies known that could keep up with qucpu? Holographic memory maybe? What about lower level data synchronization that's preventing lockups? Are qubits going to create loads of race conditions just because they run parallel? I'm not being all doom and gloom but more like realistic that qucomputing won't be here any day soon. It won't be just plugging in the new quCPU or buying a new quantum computer. The capabilities of qucomputing are so enormous that it will take a total redesign/rewrite of almost every aspect in computing. Most likely the first consumer implementation will probably be more of a time sharing type of stuff over the network which means that consumers will only enjoy faster processing of the data but not have qucomputers themselves. And before we have technology to get it to customer's hand, we probably have such fast networks all around the world that there's no need for that anymore.

[perkiset]

I don't think your being doom and gloom, but just perhaps a bit old school.

The notion of "a complete rewrite" used to be extraordinarily onerous and prohibitive. That is no longer the case. Consider that quBit computing will help us figure out quBit interpretation. The very model you cite (decrypt in record time) can be used to accelerate programmatic implementation as well.

I agree, it won't be here any time soon. But "soon" today is the 3-5 year timeline, which is a millennia compared to even 2 years ago. I for one am pretty excited about the developments they're working through.

[kurdt]

I don't think your being doom and gloom, but just perhaps a bit old school.

The notion of "a complete rewrite" used to be extraordinarily onerous and prohibitive. That is no longer the case. Consider that quBit computing will help us figure out quBit interpretation. The very model you cite (decrypt in record time) can be used to accelerate programmatic implementation as well.

Not sure I follow. You mean the age-old notation of programs programming programs?

I agree, it won't be here any time soon. But "soon" today is the 3-5 year timeline, which is a millennia compared to even 2 years ago. I for one am pretty excited about the developments they're working through.

I just hope this isn't the flying car of our time - "In the year 2000 we will have flying cars"

[perkiset]

Not sure I follow. You mean the age-old notation of programs programming programs?

Well, kind of. I mean that modeling has come a long way, and that figuring out how to reinterpret has come a long way. Considering even something as simple as Rosetta for Apple we can see how bridges are built. The way this works is that there'll be a small cadre of people that understand, which will spread to in-the-know geniuses, which will spread to the general technoratti and Poof! We'll have critical mass. It's the depth and ease of communication that will make factoring for qputing easier than any transition we've seen so far.

I just hope this isn't the flying car of our time - "In the year 2000 we will have flying cars"

LOL no lie. It's been nice to see Terrafugia and Moller make headway, but we've a LONG way to go for that stuff. In fact, I see that as more challenging because it won't be limited to the smart people, it'll be used by average boneheads. Much more challenging.

[kurdt]

Well, kind of. I mean that modeling has come a long way, and that figuring out how to reinterpret has come a long way. Considering even something as simple as Rosetta for Apple we can see how bridges are built. The way this works is that there'll be a small cadre of people that understand, which will spread to in-the-know geniuses, which will spread to the general technoratti and Poof! We'll have critical mass. It's the depth and ease of communication that will make factoring for qputing easier than any transition we've seen so far.

I do hope it will be that simple