Quantum Computing Advance Begins New Era, IBM Says (nytimes.com) 28
Because of their intrinsic ability to consider many possibilities at once, quantum computers do not have to be very large to tackle certain prickly problems of computation, and on Wednesday, IBM researchers announced that they had devised a method to manage the unreliability in a way that would lead to reliable, useful answers. From a report: "What IBM showed here is really an amazingly important step in that direction of making progress towards serious quantum algorithmic design," said Dorit Aharonov, a professor of computer science at the Hebrew University of Jerusalem who was not involved with the research. While researchers at Google in 2019 claimed that they had achieved "quantum supremacy" -- a task performed much more quickly on a quantum computer than a conventional one -- IBM's researchers say they have achieved something new and more useful, albeit more modestly named. "We're entering this phase of quantum computing that I call utility," said Jay Gambetta, a vice president of IBM Quantum. "The era of utility." A team of IBM scientists who work for Dr. Gambetta described their results in a paper published on Wednesday in the journal Nature.
[...] The IBM researchers in the new study performed a different task, one that interests physicists. They used a quantum processor with 127 qubits to simulate the behavior of 127 atom-scale bar magnets -- tiny enough to be governed by the spooky rules of quantum mechanics -- in a magnetic field. That is a simple system known as the Ising model, which is often used to study magnetism. This problem is too complex for a precise answer to be calculated even on the largest, fastest supercomputers. On the quantum computer, the calculation took less than a thousandth of a second to complete. Each quantum calculation was unreliable -- fluctuations of quantum noise inevitably intrude and induce errors -- but each calculation was quick, so it could be performed repeatedly.
[...] The IBM researchers in the new study performed a different task, one that interests physicists. They used a quantum processor with 127 qubits to simulate the behavior of 127 atom-scale bar magnets -- tiny enough to be governed by the spooky rules of quantum mechanics -- in a magnetic field. That is a simple system known as the Ising model, which is often used to study magnetism. This problem is too complex for a precise answer to be calculated even on the largest, fastest supercomputers. On the quantum computer, the calculation took less than a thousandth of a second to complete. Each quantum calculation was unreliable -- fluctuations of quantum noise inevitably intrude and induce errors -- but each calculation was quick, so it could be performed repeatedly.
There are problems like that, but how many? (Score:3)
AFAIK there are only a very few problem types for which a quantum computer is known to be theoretically better. There may be others, but how many?
It's better at factoring prime numbers.
It's better at simulation quantum interactions among entangled states.
What else? I'm sure that there ARE other applications, but I don't know of them.
Re: There are problems like that, but how many? (Score:5, Funny)
Well if you tell me in advance the number is prime, I can also factor a prime number pretty quickly.
Re: (Score:3)
Re: (Score:1)
Combinatoric problems may be more amenable to quantum computing. That entanglement allows a problem with many possible inputs to be computed in a single quantum clock cycle which is where the supposed quantum supremacy comes from.
You've probably heard that quantum electrodynamics is the the crown jewel of science, the "most successful theory" for accurately predicting the anomalous magnetic moment of the electron to something like 10 or 11 decimal places. (Aside: the true nature of this calculation and meas
Re: There are problems like that, but how many? (Score:2)
"That entanglement allows a problem with many possible inputs to be computed in a single quantum clock cycle which is where the supposed quantum supremacy comes from."
Does this remind anyone of a large language model that computes weights of all input words in a sentence in one step?
Re: (Score:2)
Grover's algorithm provides theoretical speedups for search. A lot of problems can be formulated as a search.
This is an enormous number of problems, the most interesting problems, and the most likely to be practically benefited by a quantum computer.
Re: (Score:2)
Re: (Score:3)
Quantum computers are good at any problem where the size of the problem space scales exponentially with the size of the inputs.
More long-winded: when people talk about quantum computers, they always start with how each qubit can be in a superposition of 0 and 1 and then go into the Bloch sphere and woo! quantum! And it's all irrelevant. You can represent any number of isolated qubits states to arbitrary accuracy with a linear scaling to classical bits (e.g. to represent a single qubit state with double-prec
Re: (Score:2)
Not a lot. No "normal" computing benefits in any way. And with the current tricks they are claiming will finally make QCs more powerful than my 30 year old pocket calculator, the number of things they can do shrinks down to a very small set indeed.
It is pretty fair to say that it is still completely open whether QCs will ever scale to practical usefulness.
Re: (Score:1)
"It's better at simulation quantum interactions among entangled states." That's IMO the real use of quantum computing, and here it's exponentially (in the d.f.) than conventional computers. It includes finding energy levels of molecules in battery design and how proteins fold.
Anyone who thinks IBM's computer is small has never seen it in person. The support stuff is enormous. It's also not clear that IBM's approach scales up or that it will be the best tech.
The current theme is hybrid classical-quantum
Important, but *lol* (Score:1)
step in that direction of making progress towards serious quantum algorithmic design
A step. In the direction of something serious.
This a rare moment; savor it! A Quantum Computing announcement used honest literal language to describe where they are and how what they did fits in. Bookended with hyperbole, but you can't have everything.
Error correction or avoidance is where the rubber meets the road. Probably, they'll still be slower than a regular computer at everything, because conservation and information theory. But a try requires this, and most people are still ignoring it.
All of that s
Re: (Score:2)
Two things, first, they'll have to do this kind of research to figure out how to make things less noisy and therefore requiring less error correction. Second, they are designing error correction in software that can later be created in hardware. It's just research still, no one is going to trust these things with anything important yet. It is very interesting to know the actual constraints, though.
I'm going to have to do secondary research to figure out my question - are they using Bayesian methods und
Re: (Score:2)
Well in probabilistic computing if you can get a problem that will give you the answer 51% of the time (guaranteed) then by running it repeatedly you get the answer with 99.99999% probability. You can get a highly probable answer as long as the solution gives you more than 50% probability.
Presumably the QC in question can solve this problem so quickly than running it several times is not a problem and you can converge on the correct solution quickly.
So that is a viable solution, as long as the problem runs
Hmm.. (Score:2)
So far virtually all the quantum supremacy claims I've looked at [essentially] simulate quantum systems. You see the problem with that? It's like claiming you have an extremely promising and really fast molecular interaction simulation computation system that, so far, has proven it can very quickly determine whether a liquid has acidic properties. It works by dipping a litmus paper in the liquid.
Re:Hmm.. (Score:5, Insightful)
"I have a truly marvelous demonstration of this proposition that this margin is too narrow to contain."
Re: (Score:2)
Sure, just like having a great fluid dynamics simulation is just like putting a plane in a wind tunnel.
Re: (Score:2)
But the difference is that in fluid dynamics simulation a computer isn't actually building a physical model of the plane in a wind tunnel.
Re: (Score:2)
Neither is it doing so in a quantum computing simulation of a quantum system.
Most conventional computers are electrical systems. A typical one uses tiny patterns etched on a silicon die.
Most quantum computers are also electrical systems. IBM's, for example, uses superconducting transmon qubits. These are tiny patterns made of aluminum.
When you a plane in a wind tunnel with a conventional computer you map your problem onto the behaviour of those patterns on silicon using fluid dynamics equations, a programmi
Re: (Score:2)
Indeed. The claim is complete nonsense.
IBM i / AS400 (Score:1)
Re: (Score:2)
IBM desperation (Score:2)