Fig. 1: The computing architecture of fs CIM.

a Job execution flow chart of fs CIM computing services. The complete process involves the user submitting computing jobs, such as molecular docking, financial analysis, neural networks, and material design, to a cloud server. The cloud server then converts these jobs into an Ising model matrix and transmits it to the local computer of the CIM via Kaiwu SDK (25). The FPGA is connected to the local computer for calculation using the obtained matrix. b Sketch of the overall architecture of fs CIM. Device description. PPLN periodically poled lithium niobate, IM intensity modulator, EDFA erbium-doped fiber amplifier, SHG second harmonic generation, PZT piezoelectric transducer, AOM acoustic optical modulator, BHD balanced homodyne detection. c The evolution of the in-phase components of the N = 100 OPO pulses as a function of the circulation times in solving the Möbius Ladder graph