Not sure how much of the story is true, based on information I noticed machine is designed for solving one specific optimization problem, and can solve it 100 million times faster than single core PC would.
True quantum computers are apparently still decades away.
The technological advances are impressive, I wish we could advance collectively spiritually a little too. Quantum computers, whenever materialize, can become pillars of good or evil. I hope we will be mature enough to use them wisely.
Yeah as you rightly point out, quantum computers currently have to be purpose-built to perform a specific task, kind of like where computers were back in the 40ās and 50ās before the first programmable computer was invented.
Some bright spark will eventually create the first programmable quantum computer though, then we have some interesting possibilities. How about a space MMO like Star Citizen with an entire galaxy fully simulated in real time, with precise emulation of newtonian and einsteinian mechanics in play? An entire planetās population procedurally generated with their own history, relationships, personalities and backstories (a bit like Dwarf Fortress but with less eccentricity).
Keep in mind though, the D-wave systems arenāt āfullyā quantum computers, and therefore lack some/a lot of the truly weird characteristics that can be used.
Also, the speedup is from a comparison between a 25 KW quantum computer with an ordinary 100ās W computers. And the ordinary cpu is running a simulation of what the quantum computer would do. So itās by no means a fair comparison.
Itās still an impressive machine for all I know - but itās not by any means that impressive. Itās maybe a few hundred to a thousand times quicker on very specific tasks, and even that speedup is arguable. āFullyā quantum computers only have qubits in the single digits atm, and barely work with that.
That is all of course if quantum entanglement and the wave collapsing can ever be solved. Its entirely possible that a quantum computer will be physically limited to a small number of bits.
Quantum computers are simply not that impressive to me. One hundred million times faster than a single core computer may have been impressive back when we only had a single core and graphics cards could barely handle 2D let alone 3D.
By the time quantum computers have finally been realized will there actually be that big of a gap over conventional?
@Ryiah - Remember, quantum computing is sort of like time travel, anti-gravity and advanced, lifelike artificial intelligence that is indistinguishable from human intelligence. Itās a kind of magic people believe in. Once we get itā¦ they think everything will change forever, and of courseā¦ we are closer than anyone knows.
When in reality, if/when we do figure this stuff out it will be used in very non sexy ways to solve boring problems.
@Master-Frog Itās sort of like those as in āweird hard to understand things possibly possible in the futureā but time travel (to the past) and anti-gravity donāt have any scientific foundation that I know about, where quantum computing does.
And the issue of lifelike artificial intelligenceā¦ itās interesting. Computing power definitely seems to be growing steadily exponentially up to today, so eventually youād reach some levels resembling human computing capability. That capability can than further improve the chips used, and then growth would become double exponential and shit hits the fan. And as you say - this could be closer than anyone knows if you consider in the massive billions some governments/companies throw at supercomputers. Note here; Googleās search result is a mix of a hundred or so factors; A primary one has become trained AI the past year or so.
Judging by the info I gathered, thereās class of problems that can be solved at very high speed with quantum machine, and conventional machine wonāt be ever getting close to that. For example, it is supposed to be able to crack most of the modern cryptoalgorithms, see Shorās Algorithm ( https://en.wikipedia.org/wiki/Shorās_algorithm ).
Iām not sure about practical applications of that hardware in normal programming.
The math: http://www.ams.org/samplings/feature-column/fcarc-quantum-one http://www.ams.org/samplings/feature-column/fcarc-quantum-two
is quite different from traditional computing, so it looks like the program will need to be designed from grounds up around qubit behavior.
As far as I can tell, it is like having a machine that can very quickly solve your problem, but only when you sing the problem to it in ancient sumerian. Or something.
Speaking of which, IIRC there were couple of people warning about AI trouble within next 5 years.
One of theories is emergent intelligence, where intelligent behaiovr in sufficiently complex system just emerges due to glitch or set of external factor. Right now thereās blue brain project, and IIRC google also had couple of ai-releated projects + they bought boston dynamics for some reason. If thereās a company that can accidentally create skynet, thatās probably google.
I have great hope for the future. Itās clear most problems are of greed and wanting want we canāt have or wanting to be what we cannot be rather than enjoying who we are.
Using quantum computing, I think weather-proof carbon-fibre (or other appropriate building materials) homes, cars, and using other more relevant materials to the problems at hand, health and safety can be engineered for everyone but quantum computing still canāt control peopleās lust for what belongs to others. Certainly quantum computing can be used to pull people out of the past and into the future. There really comes a point when you have to question the relevance of picking at the scabs of the past and reject those that use the unpunishable past to try and leverage power over others today. There are still seal and whale hunters based solely on a phony claimed desire to keep alive oneās cultural traditions and yet those same people couldnāt modernize quick enough for their own comfort and longevity despite all the cultural traditions those technologies required them to abandon.
I think it will be useful for rendering and can āsettleā into the correct solution much more rapidly in unbiased rendering. Game AIā¦not so much. Transform and vectors, rotations and suchā¦I do not see the benefit. Perhaps a mini quantum computer on graphics chips to speed up realistic unbiased renders would be a highly useful pro/consumer everyday use for itā¦ Anybody want to start the next gazillion dollar startup with me:)
It is not yet known how anything is going to be changed. It is going to be a step by step process most likely. Maybe comparable to what we see with multi core since a few years. Just because the possibility exists doesnāt mean that everything has to be reimplemented completely. Many things donāt need to be multi threaded at all and for everything else, there are usually step by step improvements possible.
I didnāt have enough time to thoroughly read through that qubit math article I read earlier. AFAIK, qubit technically represents constrained vector and operations on it are probabilistic, plus thereās whole superposition/entanglement thing. Basically, when qubit represents 0/1, before you read data from it, it exists in both states at once, and when you read data from it it snaps to one of them, with certain (known?) probability. Traditional binary can be visualized as on/off switch and operations on it are 100% sure. Meaning I have reasonable reason to doubt that, for example, C language would work well for qubit hardware due to completely different lowest level architecture. It might turn into something like quantum coprocessor, though - sorta like we handle GPU/CUDA programming and shaders these days.
