Fujitsu claims 'major technical milestone' in quantum simulation

Fujitsu says it has developed the world's fastest quantum simulator capable of handling 36 qubit quantum circuits. The firm will use this to speed development of quantum applications, and said it is already planning a further simulator capable of handling more qubits. The unnamed simulator was built using a 64-node cluster of …

[Doctor Syntax]

What's the relative performance of running a quantum application on a classical simulator of a quantum computer and running a classical application for the same purpose on the same computer?

[Mike 137]

Well...

"What's the relative performance [...]"

64 CPUs each with 48 cores and 32 GB memory => 36 Qbits, so we can make an educated guess, can't we?

Apart from which, it seems that quantum application development is running ahead of quantum hardware to run it on. How soon (if at all) are they likely to join up?

[zuckzuckgo]

Directly from the article:

"IonQ noted that for a small number of qubits, a simulator can be faster than an actual quantum processor. However, the length of time and the amount of memory required for a simulator to process a quantum circuit increases exponentially with the number of qubits, while it only increases linearly for the quantum computer."

The unanswered question is whether the complexity of an actual quantum computer will also increase exponentially with the number of qubits supported. They are still working on that one.

[zuckzuckgo]

To clarify my own post: The unanswered question is whether, as qubits increase, does the complexity of a quantum computer grow faster than the computational benefits.

[Anonymous Coward]

In other news

I've heard the fairies at the bottom of the garden have perfected the worlds fastest simulation of 36 unicorns.