Ising-CIM: A Reconfigurable and Scalable Compute Within Memory Analog Ising Accelerator for Solving Combinatorial Optimization Problems.
Shanshan XieSiddhartha Raman Sundara RamanCan NiMeizhi WangMengtian YangJaydeep P. KulkarniPublished in: IEEE J. Solid State Circuits (2022)
Keyphrases
- markov random field
- compute intensive
- low memory
- low cost
- limited memory
- higher order
- combinatorial problems
- memory space
- partition function
- signal processing
- memory requirements
- parallel implementation
- field programmable gate array
- combinatorial optimization problems
- main memory
- metaheuristic
- image processing
- neural network
- memory usage
- approximate inference
- general purpose
- evolutionary algorithm
- data conversion
- analog vlsi