Google: We don't have a quantum computer yet, but we have a compiler
It's quantum, it's open source, it's on GitHub. Did we miss anything?
Posted in Emergent Tech, 24th October 2017 03:59 GMT
Google and rival Rigetti Computing have co-published what you could think of as an open source quantum compiler.
OpenFermion, available here, is more formally described by the Chocolate Factory as a “platform for translating problems in chemistry and materials science into quantum circuits that can be executed on existing platforms”.
Since there isn't a quantum machine to run OpenFermion problems yet, the target environments for now are conventional computers acting as quantum simulators – as described at the repo, it compiles and analyses “quantum algorithms that simulate fermionic systems” (emphasis added).
One day there will be quantum computers and Google reckons OpenFermion is ready for their debut, because it's “designed in a quantum programming framework agnostic way” to support packages that “compile quantum assembly language specifications for diverse hardware platforms”.
Google's rationale for the project is that – whether the compilation target is a simulator, or a quantum computer – running a chemistry problem demands “prohibitive amount of domain knowledge” (read: a PhD) in both chemistry, and quantum computing (make that two PhDs) (as described in the partners' arXiv paper).
That unattainable amount of know-how is needed just to get one problem up and running – the one OpenFermi's designed to solve, which Google describes thus:
“The primary function of OpenFermion is to encode the electronic structure problem in second quantisation defined by various basis sets and active spaces and then to transform those operators into spin Hamiltonians using various isomorphisms between qubit and fermion algebras.”
So no, don't ask for “hello world”.
Google's announcement follows a similar announcement by Microsoft in September, but Redmond gave way less detail and hasn't yet divulged its quantum software's name. ®