2 ms·
They kind of say this, but we really need better quantum benchmarks. How does something like quantum volume analogize to address space or FLOPs or something som
by nonameiguess 3y ago
They kind of say this, but we really need better quantum benchmarks. How does something like quantum volume analogize to address space or FLOPs or something somewhat intelligle in classical computing? What were the space and cost requirements to achieve this growth? I was under the impression years back that a major obstacle to quantum computing was keeping qubits from decohering required immense space and extreme cooling compared to classical computers. How has qubit density per dollar spent on the computer changed over this same span of time? Did the technology actually get better or did they just throw more money at it?
- gaze 3y agoWithout error correction, these benchmarks are going to be a bit weird and fluffy. It’s a bit like racing a half-built car. There’s the other bit of complexity that the connectivity matters a great deal. Supercomputers can be connected on a grid, torus, fat tree, etc. and the different connectivity will lend to different problem solving capabilities in a somewhat subtle way. Hence different benchmarks highlight different capabilities, and correspondingly, each company will push a benchmark that makes their machine look the best.