Dwave’s 502 qubit adiabatic quantum computer was found to be 10,000 times faster than a regular quadcore workstation for some optimization problems and just a little faster for some other problems.
There was an attempt to approximate the computing power of the Dwave quantum computer by using multiples of performance against the Linpack capabilities of the workstation.
The classical [regular] computer used for most tests was powered by a cluster of seven Intel Xeon E5-2609 processors running at 2.4 GHz. Comparing to Linpack benchmark results from massively parallel computers using similar processors, the D-Wave Vesuvius chip obtained a (very approximate) Linpack rating of 420 GFlops.
For the optimization problems that are best matched to the capabilities of the Dwave system where it is 10,000 times faster then the Dwave system is approximately equal to a 4.2 petaflop supercomputing system.
If speed scaling seen between the 128 qubit to the 512 qubit system held between the 512 qubit and a 2048 qubit system, then the 2048 qubit system would be expected to be 1000 to 500,000 times faster for optimization problems.
The 2048 qubit system would then be expected to have approximately 4.2 to 2000 exaflops (2000 exaflops is 2 zettaflops) of performance for optimization problems that best match its hardware.