Thread by @fgrosshans

Now, Yoshihis Yamamoto on Physica of Quantum-toClassical Crossover and coherent Ising machines #LTQI

Yoshihisa Yamamoto : The goal is to build acelerators for combinatorial an continuous optimization problems. A traditonal quantum computer essentially uses a Grover algorithm and gets a quadratic improvement, but scaling essentailly remains exponential. #LTQI

Yoshihisa Yamamoto: One needs to use hueristic quantum machines, like a quantum neural network. He compares it to quantum annealers in arxiv:1805.05217 arxiv.org/abs/1805.05217 #LTQI

Yoshihisa Yamamoto: The quantum neural network is implemented in nonlinear optics. It allows delocalized wavefonction ofet 2000—10⁶ “spins”, quick, works at oom temperature and couples to ptical fibres.
#LTQI

Yoshihisa Yamamoto: The artificial spins are actually optical parametric oscillators (OPOs). Spin ↓↑ corresponds to squeezed coherent state at φ=0 or π. #LTQI

Yoshihisa Yamamoto uses measurement feedback for spin-spin coupling. Pulses are in a loop with an OPA, partial measurements and a feedback through optical pulse injection #LTQI

Yoshihisa Yamamoto: The omputation is 3-step: quantum parallel search based on correlated quantum walk beyond threshold; spont. sym. break makes decision at threshold, exponential amplitude amplification above threshold #LTQI

Yoshihisa Yamamoto: The trajectory of the machine is an example of “quantum darwinism”, where the system selects the quantum state which entagles least through environment #LTQI

Yoshihisa Yamamoto: This is an example of Dissipative Quantum Computation, by opposition to the more traditional Unitary quantum computation model #LTQI