In this thesis, a hardware architecture named KAFA(KAist Fuzzy Accelerator) is designed and implemented for the fuzzy information processing based on fuzzy logic and fuzzy set theory. It provides not only various fuzzy inference method but also various fuzzy set operations. It is designed in consideration of simplicity, extensibility and generality. To utilize a massive parallelism which is inherent in fuzzy information processing, the architecture has SIMD structure. The KAFA consists of one Main Controller and 128 FPE(Fuzzy Processing Element)s. Current ASIC technology could compact eight or sixteen FPEs on a chip. The KAFA has very simple structure and high performance.
