back to article Google says its quantum computer is 100 million times faster than PC

Two years ago Google and NASA bought a D-Wave 2X quantum computing system and the Chocolate Factory has now pronounced itself very pleased with the results. "We found that for problem instances involving nearly 1000 binary variables, quantum annealing significantly outperforms its classical counterpart, simulated annealing. It …

Silver badge

Re: Annealing???

Absolutely.

This technique's name is inspired by the metallurgic process of annealing to remove stress.

https://en.wikipedia.org/wiki/Simulated_annealing

6
0

Re: Annealing???

Simulated annealing is a technique for optimisation technique for computers, so named because it was inspired by exactly what you mentioned, annealing of metal to soften it.

SA is hill climbing, but with the addition of probabilistic moves in directions that lead to worse results. This allows SA to escape local minima - in a sense it is "softening" the function being optimised, smoothing and blurring it to allow the global optimum to be found.

6
0
Silver badge

Re: Annealing???

At last, I think I've followed something. That bit sounds like part of the Mechatronics course I did with the OU about 17 years ago.

1
0
Anonymous Coward

Maybe it works and doesn't at the same time or until you look at it you don't know?

0
0

"Maybe it works and doesn't at the same time or until you look at it you don't know?"

The problem is that if you look at it then you change it and so you have to start all over again...

0
0
Unhappy

Wahoo....

...Google now can calculate a billion different ads that "may be of interest to you" by tracking a billion different ways of using a billion bits of internetofshit

5
1
Anonymous Coward

I realiise

I have been spending too much time watching Vines, D-Wave made me laugh.

0
0

actual benchmarking says "meh"

http://arxiv.org/abs/1504.07991

Tail of the time-to-solution distribution is considerable fatter with quantum annealers for some reason versus classical simulated annealing. Its a promising approach, worth being aware of but not yet worth shelling out for on its own merits.

2
0
Anonymous Coward

I think the real test they should be running is comparing power consumption. The total power consumption of the D-Wave machine is far greater than a single processor chip. Additionally the D-Wave isn't a general purpose computer so perhaps the best comparison would be against the best ASIC for solving the given set of problems. The fight would therefore be: D-Wave vs Same total power consumption group of ASICs. If there's still a benefit to using D-Wave then they are on to something good. If this was BitCoin the current comparison would be Desktop PC vs ASIC miner.

1
0
Silver badge

OMG the oompa-loompas will set fire to the sky!

0
1

This post has been deleted by its author

A billion flies can't be wrong

How many recipes does it need to analyze before it can make a tasty meal?

0
0

Re: A billion flies can't be wrong

You know you'll end up with something that is almost -but not quite- entirely unlike tea.

1
0
Silver badge
Holmes

Re: A billion flies can't be wrong

Or make a nice warm cup of tea?

0
0
Anonymous Coward

Crytpo-pop

Bang goes the current cryptography keys.

Is everyone going to have to get one of these just to remain secure?

2
4

I do wonder how much speed increase one might achieve by simply designing a completely new architecture based on modern design techniques. As far as I know, most computers we use these days are highly refined versions of the Von Neumann architecture - Quantum computers are a different beast so they have exactly that opportunity leading me to wonder whether this is part of why they show some impressive jumps in pace.

1
1
noj

Now we know the REAL reason for global warming.

(that was a joke)

Guess we're all going to need longer passwords soon.

(that was not)

1
0
Silver badge

Blimey, look at all the Plutonium badges in the comments...

I feel like a noob...

1
0
Silver badge

Re: Blimey, look at all the Plutonium badges in the comments...

Don't feel too discouraged. Eadon had a gold badge!

0
0
Anonymous Coward

But can I use it to farm Bitcoins.

0
0

Terribly frightening

An army of Google war robots powered by quantum AI armed with the accumulated knowledge of the human race. Monkeys are in SERIOUS trouble.

1
0
Joke

Well...

Can it quantify 'up to'

1
0
JLV
Silver badge

Dumb ass question but are these types of quantum computational algorithms applicable to Bitcoin-style computational mining in theory? If so, what would be the currency implications of a breakthrough where someone would suddenly out speed everyone else for a while?

0
0
Anonymous Coward

If you can afford to buy and run one of these super-cooled monsters, I suspect you don't need to worry about mining fake plastic tokens ...

> If so, what would be the currency implications of a breakthrough where someone would suddenly out speed everyone else for a while?

Everyone would realize that they entire crypto current scam ^H approach is a ponzi scheme designed to make the guys who got in early rich, and go back to using normal money?

1
2
Silver badge

"

Everyone would realize that they entire crypto current scam ^H approach is a ponzi scheme designed to make the guys who got in early rich, and go back to using normal money?

"

On the contrary, it is traditional currency (most of us no longer use money, only currency) that is akin to a ponzi scheme, with the government controlling the scheme and reaping the rewards. Bitcoin OTOH, if it were to be widely adopted, fulfils all the criteria needed to be real money (inherently limited, durable, portable, divisible). The downfall of currencies are that they are not inherently limited - governments can create as much if it as they want very quickly and with no effort, and all have done so to some degree or another. Bitcoins, like gold bars, can only be created at a slow rate and with a large effort.

Every time the total amount of currency in existence is increased, it transfers a proportion of the value of the currency in your bank account into the pockets of the government - which is the ultimate stealth tax because you do not see any loss in the amount of currency you have, only a loss in the value of the goods it is able to buy (which amounts to the same thing but most people don't see it for what it is).

1
1
Anonymous Coward

> Every time the total amount of currency in existence is increased, it transfers a proportion of the value of the currency in your bank account into the pockets of the government

And this isn't true for a bitcoin-like mining scheme because?

0
1

@ Cynic_999

... which is exactly why I have always considered BitCoin to be a resource rather than a currency, and one which should be regulated (for trading purposes) as such, like goldetc.

0
0
Anonymous Coward

> And this isn't true for a bitcoin-like mining scheme because?

It is a good idea to become at least minimally familiar with a subject before passing an opinion. As the critic that you are, I would have expected you to know the answer to that question.

The total amount of Bitcoins that can be generated is limited by design. You could of course change the algorithm, but then it would no longer be Bitcoin but some other thing. There are indeed a number of "digital currencies" about, I believe.

For the record, I have no opinion or interest in so-called digital currencies. For that matter, I have no opinion and very little interest in so-called traditional currencies.

0
0
Silver badge

Dumb ass question but are these types of quantum computational algorithms applicable to Bitcoin-style computational mining in theory?

Quantum annealing isn't, as far as I can tell, so DWave (or D-Wave or whatever they call themselves) are out. And I don't see any obvious application of any of the algorithms I know of that are in BQP to cryptocurrency mining.

In any case, as other people suggested, the economics don't work. The DWave machine is expensive enough; as far as we know, no one's built a real QC of any size, for any price.

1
0

This post has been deleted by its author

Silver badge
Terminator

Einstein would be horrified

Imagine if MSFT had this quantum technology, they could slurp data from the Other Side of the Universe

0
1
rav

WOW Quantum Computing from NASA......

next they'll be saying that Andrea Rossi has reinvented COLD FUSION and the EM Drive just needs Di-Lithium Crystals.

0
0
Silver badge

Re: WOW Quantum Computing from NASA......

In today's Empire of Wishful Thinking, that can't be far off.

0
0
FAIL

A cynical load of hogwash...

Yes I read the paper, and it is interesting, but not that exciting. If it were quantum, I would be the first to congratulate them.

However the 108 claim is absolute rubbish. We have machines with at least 106 computing elements, so why not compare their megafridge against THE BEST traditional machine? Not some weedy PC!

This is the reason why the TOP500 persists. It is too easy to fudge a benchmark your codes does well on that nobody has heard of.

I am not impressed (yet) and I even sat through their sales pitch....!

Is it possible Google and NASA are just trying to justify buying a very expensive fridge....?

P.

3
0
Silver badge
Holmes

Re: A cynical load of hogwash...

This.

I would also like to know whether it is actually using Quantum Effects in there or whether it is, in the end, just a special-purpose ANALOG COMPUTER (i.e. a physical system that models the problem you want to solve). Because I remember papers that were disputing the presence of Quantum Effects in a D-Wave in particular.

The abstract at least nicely elides the question.

Analog computers being millions of times faster than symbolic processing machines at a given specialized task is not a particularly surprising finding.

2
0

Some interesting comments but ...

... the real question is - how will affect my Facebook experience?

Please tell me it can't get any worse!

0
0

we already know....

the answer is 42....

0
0

Carefully chosen benchmark comparison

Classical simulated annealing is not a particularly good optimization technique which was acknowledged in the paper:

"It is often quipped that Simulated Annealing is only for the 'ignorant or desperate'."

There are much better classical optimization technique which would compare much more favourably but they used it because it was the closest analogue to quantum annealing.

Simulated annealing also doesn't do well with potential energy surface which contain deep, narrow wells where it can get trapped. Quantum annealing can tunnel very efficiently through them and so they chose a problem which suited it. From the paper:

"carefully crafted proof-of-principle problems with rugged energy landscapes that are dominated by large and tall barriers"

1
0

This post has been deleted by its author

Anonymous Coward

Is it too much to ask that the heading says "... faster than a PC"?

2
0

A good way of thinking about this machine is trying to open a regular pin padlock.

The traditional approach is to sequentially try a key with every different depth of cut in every position.

The quantum computing approach is a "bump key", where you simultaneously sweep across all cut depths on all positions.

The sequential approach can take a very time. It's trivially parallelized, although you are just one-for-one trading hardware for time.

The quantum computing approach is almost instantaneous, but you are never certain if you have found a correct solution. You always have to test the result to see if it works, and even then repeat a few times to see if the result is the same. If you consistently come up with a bad result, you need to reformulate the problem to specifically exclude that result and try again -- much like a lock pin always getting suck on a nick.

2
0
Windows

I realize it may be a challenge, but...

I'm sure Microsoft will rise to that challenge as it as done in the past and make Windows and M$ Office run 100 million times slower...

1
0

Page:

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Forums

Biting the hand that feeds IT © 1998–2018