QML: "Quantum Data loader" instead of QRAM?



In last year‘s conference "Quantum For Business 2019" Iordanis Kerenidis gave a nice talk about quantum machine learning.

At about time 27:10 he mentions a "Quantum Data loader" as an alternativ for QRAM. To me it sounds very concrete but I can not find any reference for it. Has somebody any details about it or did I get this wrong? I have looked at arxiv papers 1911.01117 and 1812.03584, that he probably referred to in his talk. But these seem to use QRAM explicitely.


Posted 2020-04-06T19:17:02.900

Reputation: 33


Hi and welcome to Quantum Computing SE. Something on qRAM which may be useful: https://quantumcomputing.stackexchange.com/questions/9413/what-is-the-state-of-the-art-on-quantum-random-access-memories

– Martin Vesely – 2020-04-06T21:52:00.497

thank you for the link – jens7 – 2020-04-07T15:49:18.757



He generally refers to QRAM. But there are other ways to load data in a quantum computer, generally referred as quantum embeddings or quantum feature maps. They are generally more near-term oriented than QRAM. Yet with a QRAM, you can design an algorithm with the goal of obtaining a possible expression of a speedup against a classical version. This article, and this one are examples/references you can learn about quantum embeddings and quantum feature maps.


Posted 2020-04-06T19:17:02.900

Reputation: 4 213

great, thank you for the references – jens7 – 2020-04-07T15:48:05.010