| Popis: |
Quantum computers use quantum physics phenomena to create specialized hardware that can efficiently execute algorithms operating on entangled superposed data. That hardware must be attached to and controlled by a conventional host computer. However, it can be argued that the main benefit thus far has been from reformulating problems to make use of entangled superpositions rather than from use of exotic physics mechanisms to perform the computation – such reformulations often have produced more efficient algorithms for conventional computers. Parallel bit pattern computing does not simulate quantum computing, but provides a way to use non-quantum, bit-level, massively-parallel, SIMD hardware to efficiently execute a broad class of algorithms leveraging superposition and entanglement. Just as quantum hardware needs a conventional host, so to does parallel bit pattern hardware. Thus, the current work presents Tangled: a simple proof-of-concept conventional processor design incorporating a tightly-coupled interface to an integrated parallel bit pattern co-processor (Qat). The feasibility of this type of interface between conventional and quantum-inspired computation was investigated by construction of an instruction set, building complete Verilog designs for pipelined implementations, and by observing the effectiveness of the interface in executing simple quantum-inspired algorithms involving operations on entangled, superposed, values. |
