D-Wave Launches in AWS Marketplace | Hacker News Reader

D-Wave Launches in AWS Marketplace

scientific-computing.com

70 pointsdonutloop26 comments

Seems like they're shrugging off the question of whether quantum computing has any advantages over classical, and going straight to "use it or you're missing out".

As far as I know, there's still no convincing evidence that today's quantum computers do anything better than classical computers. For example, Google made a very narrow claim of quantum superiority and even this was undermined by subsequent research [1].

[1] https://techcrunch.com/2022/08/05/googles-quantum-supremacy-...

Furthermore, I believe there's some question as to whether the kind of computer that DWave has made is even a true quantum computer.

It is worth noting that all known classical algorithms to solve this problem scale exponentially with the problem size, and fundamentally there is no reason why putting 60 or 70 qubits in the fridge is particularly more difficult than 53. The point is that while classical algorithm advances can take off orders of magnitude off the runtime, this has to be done _every_ time there is a small increase in the size of the quantum computing device in order to kill a supremacy claim.

(To clarify, I personally think optimization is not a remotely near-term application)

You're assuming that quantum algorithms scale better than classical ones. To the best of my knowledge (which is not much), the cases in which this is known to be true are extremely limited. We have a faster algorithm for factoring numbers, Shor's algorithm, and I'm not sure there's anything of note beyond that.

At this point in time there's no one that has enough logical qubits to do a useful quantum computation. It might be a few decades before someone figures out how to scale up the number of useful qubits. In the meantime a bunch of QC start ups will pop in and out of existence like particles. Very tough field, but that's exactly where I want to see more resources allocated: towards solving difficult problems.

Your comment is off topic.

D wave is not a quantum computer. It is a specialized hardware for solving certain classes of optimization problems using quantum annealing.

Is this different than what has been available through Braket? https://aws.amazon.com/braket/quantum-computers/dwave/

> Leap will be available in AWS Marketplace soon, and Amazon Braket customers will be directed to the AWS Marketplace to access the D-Wave system.

Are there a lot of entries in the AWS marketplace that you can't run on AWS, but instead access remotely? I thought it was mostly things like AMIs (Windows, etc) or managed services you run on top of EC2 VMs (databases systems, etc). But this is actually hardware operated by a third party...

Various SaaS vendors are available via the AWS marketplace, including ones that aren't running on AWS.

For example we pay for Datadog (using their EU region which happens to be on GCP) via our AWS bill. As we have 30 day invoicing with AWS this makes it all very convenient.

I'm not sure how this deal works behind the scenes, but AWS already rented access to D-Wave devices previously via AWS Braket. I wonder if they're continuing to operate the hardware for this. I agree it would be an odd choice to put this in the AWS Marketplace if they're not.

The actual quantum computers don't live inside AWS datacenters

https://aws.amazon.com/braket/faqs/ (under security)

Most of AWS Braqet is orchestrating emulators (ie fake quantum computers running on EC2).

What can I do with it?

A fine question!

> Our solvers include the binary quadratic model solver, the discrete quadratic model solver, and the constrained quadratic model solver.

Apparently this include graph colouring and travelling salesman problems, as well as things expressible in https://github.com/MQLib/MQLib .

What was the time complexity of the solution to salesman problem?

Or are quantum algorithms not measured in the same way?

The problem is np hard and d-wave runs quantum annealers, so the ultimate benchmark is likely wall time.

How far off is the tech to be able to run Shor’s algorithm?

It is not general purpose quantum computer, so it will not run Shor algorithm.

Isn’t it, anything but slower and more costly than anything you could do with a convention computer?

Oh wait, it’s just a very small subset of problems, nvm.

Constrained optimization

You can find the prime factors of the number 15.

I wonder how far they have come in speeding things up. A good while back I read an article that said the reviewer was not sure what was running on the regular chips and what was not but the performance gains (back then) was not worth the cost.

So is quantum computing a real thing?

D-Wave is not a universal quantum computer. It’s a quantum annealer [0].

[0] https://www.quora.com/Why-is-D-Waves-quantum-computer-not-co...

There is no known reason why it is theoretically impossible.

However, despite many people claiming otherwise, there are currently no situations (other than contrived benchmarks) where a quantum computer is better than a digital computer.

Of course it is! But, can it catch up to the 60-year head start that classical computing enjoys? That remains to be seen.

Reply on news.ycombinator.com