Partitioned state encoding for low power in FPGAs
Luis Mengibar, Luis Entrena, Michael G. Lorenz, E. San Millán
Abstract
Luis Mengibar, Luis Entrena, Michael G. Lorenz, E. San Millán
Abstract
The problem of finite state machine (FSM) encoding for low power in field-programmable gate arrays (FPGAs) is addressed. In this technology, one-hot encoding is typically recommended for large FSMs and binary encoding for small FSMs. A partitioned encoding approach is proposed which uses a combination of both binary encoding and zero-one-hot encoding with intermediate code size. Experimental results demonstrate that the proposed encoding approach can produce significant power savings.
OpenAlex reports 24 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The problem of finite state machine (FSM) encoding for low power in field-programmable gate arrays (FPGAs) is addressed. In this technology, one-hot encoding is typically recommended for large FSMs and binary encoding for small FSMs. A partitioned encoding approach is proposed which uses a combination of both binary encoding and zero-one-hot encoding with intermediate code size. Experimental results demonstrate that the proposed encoding approach can produce significant power savings.
Key concepts: Encoding (memory), Field-programmable gate array, Finite-state machine, Binary number, Computer science, State (computer science), Power (physics), Code (set theory)