IBM makes quantum computing available to the public

[HEXUS]

Via a five-qubit IBM quantum processor connected to the cloud.

Read more.

[CK_1985]

Hands up, I did not understand one word of that. Still, yay science!

[lumireleon]

A top scientist benchmarked the current popular D-Wave quantum computer and the speed is comparable to an i3 laptop.

[realstock]

With Quantum computing think of the advances, in targeted advertising google can exploit.... yay!

[HW90]

A top scientist benchmarked the current popular D-Wave quantum computer and the speed is comparable to an i3 laptop.

Google and NASA disagree, they published a report which found for applications where quantum methods can be applied they saw it was 10^8 times faster than a traditional processor.

[watercooled]

A top scientist benchmarked the current popular D-Wave quantum computer and the speed is comparable to an i3 laptop.

Google and NASA disagree, they published a report which found for applications where quantum methods can be applied they saw it was 10^8 times faster than a traditional processor.

Those two statements don't necessarily contradict - quantum computing is not well suited to some applications so there will be no speedup, and in fact even a big regression in some cases - don't forget quamtum computing is still in its infancy whereas traditionally processors have had time to get extremely fast at what they do. So while e.g. a given algorithm might see a 10^8 speedup in quantum vs traditional computation, you also have to look at the actual speed of the processor - the concept of fast and slow still exist in quantum computing.

To use an analogy, a GPU can be many times faster than a CPU at specific tasks, but for various reasons they're still perilously slow at some tasks vs CPUs e.g. complex single-threaded workloads. Which one is 'faster'?

[plexabit]

As a scientist, I ask: how are its floating-point computations over the cloud?

Yeah, I'm asking if it will run Crysis.

[lumireleon]

am no tech savvy scientist working on blackholes or molecular spectroscopy but this idea of quantum computing will be applied to bitcoin mining, video trans-coding and rendering a complex scene in Maya or 3DS which takes an i3 with a GTX 970 nearly a month to complete.

[lumireleon]

With Quantum computing think of the advances, in targeted advertising google can exploit.... yay!

the google quantum computer will bypass all the browser add-blockers and show BY-FORCE

[QuorTek]

We have all this Tech.. with Quantum computing still being in the breeding pen then it is all a matter of time as we have all previous tech already invented and put into use... now it is just for integrate Quantum computing into all the sockets and replace everything with that.

[lumireleon]

The major problem with the current quantum computers is the below zero cooling unlike silicon candy which needs a cheap copper heatsink.

[valhar2000]

The major problem with the current quantum computers is the below zero cooling unlike silicon candy which needs a cheap copper heatsink.

This is why IBM's idea of making a quantum computer accessible over the Internet is a good one. Not many organizations will be able to purchase and maintain the infrastructure needed to run one of these things, but they may be able to afford to rent time on one of these machines to perform specific tasks that other computers cannot feasibly carry out.

[Gerrard]

I might be wrong, but if they were to get a 128-bit+ version (yeah, some time away I know, but I am not sure if you can parrallel smaller ones up somehow...), wouldn't that make some encryption pointless? I may be remembering it wrong and it has probably moved on since then, but some encryption relies on multiplying two prime numbers as keys together for a code. With quantum computing, it can simultaneously try every combination of two numbers to find the keys for that code.

[DanceswithUnix]

I might be wrong, but if they were to get a 128-bit+ version (yeah, some time away I know, but I am not sure if you can parrallel smaller ones up somehow...), wouldn't that make some encryption pointless? I may be remembering it wrong and it has probably moved on since then, but some encryption relies on multiplying two prime numbers as keys together for a code. With quantum computing, it can simultaneously try every combination of two numbers to find the keys for that code.

New quantum encryption schemes are being worked on for if that time comes.

But yes, if someone can take a copy of your encrypted drive, then in ten years time they might be able to decrypt it. That might be important for some information with a long shelf life.

[watercooled]

Shor's algorithm is most potentially damaging one which as I understand it can attack public key encryption algorithms based on the discrete logarithm, integer factorisation problems and their elliptic curve equivalents in a reasonable amount of time on a big enough quantum computer, although I'm not sure what 'big enough' actually means if I'm honest - I assume current systems aren't capable of it but I'd like to know what would be necessary to pull this off.

Gorver's algorithm can attack symmetric key ciphers like AES but is far less damaging and fairly easily mitigated without any major changes to the cipher or its performance; the attack effectively reduces the complexity of the attack to x^n/2 i.e. a 128 bit key becomes as hard to attack as a 64 bit key on a conventional computer, which would be feasible on a fast enough quantum computer. However you can just stick to 256 bit keys increasing the complexity to 2^128 again.

I'd like to learn some more basics about quantum computing - I often hear speed comparisons (i.e. equivalent speed rather than algorithmic speed, if that makes sense) between quantum and conventional computers, I guess they're referring to something like clock speed. Also I'm not sure how the 'size' fits in to things e.g. the quibit count. I mean I accept them as concepts but I'd like to know how they all fit together - most of the 'explanations' I've seen don't come close to this sort of detail.