I've been reading a couple of papers regarding qPCA and its dequantisation. In the course of my reading, it appears to me that one of the ingredients that makes this dequantisation meaningful is the requirement for a qRAM for state preparation when it comes to using qPCA for analysing classical datasets. The other ingredient being that the problem in question is low-dimensional. While one potential way to avoid dequantisation is tackling a high dimensional problem, the other seems to be removing qRAM from the picture. In light of this answer in an older question of mine, I wonder if this so-called quantum data that is sampled directly from a physical system requires a qRAM or any state preparation. In summary, my question is as follows -

Can quantum computation (in particular qSVT-type problems) be done without a qRAM-based state preparation? If so, are there any caveats?



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.