4 ms·
It was unclear to me what the impossibility is that is referred to by the article, but it appears to be the exponential nature of the required storage for the s
by ckastner 9y ago
It was unclear to me what the impossibility is that is referred to by the article, but it appears to be the exponential nature of the required storage for the simulation:
> Simulating anything more than 49 qubits on a classical computer was thought impossible because of the limited memory classical computers have compared to their quantum counterparts. Because of the difference between classical and quantum computing, each time you add a qubit to a simulation it increases the necessary memory exponentially in the classical simulation.
If true, then the jump from a simulation of 49 to 56 was clearly not impossible in the first place.
Following some links, the term "impossibility" is first introduced by the New Scientist, but it's not backed up by anything substantial, as far as I can tell.
- zellyn 9y agoWell, they explained it a couple of paragraphs later. The previous record, a 45-bit quantum simulation, took 500 Terabytes of memory. I assume there's some sort of terrible exponential increase that would have made 49-bit simulations require four orders of magnitude more, or something impractical. But this one did 56 bits with just 4.5 terabytes of memory. All seems pretty clear.
- ckastner 9y agoThat's my point: impractical as it may be at the current time, I could not find an argument as to why it should generally be impossible.
- red75prime 9y ago> the limited memory classical computers have compared to their quantum counterparts It is a bit misleading. Holevo’s theorem states that it is impossible to reliably extract more than n bits from n qubit system.