New Software Framework Doubles Processing Speeds, Halves Energy Use

Researchers at UC Riverside develop SHMT, a method that utilizes existing hardware for parallel processing, offering significant efficiency gains.

A new software framework called simultaneous and heterogeneous multithreading (SHMT) has been developed to double computer processing speeds and halve energy consumption.
SHMT enables parallel processing by using existing hardware components such as GPUs, AI accelerators, and digital signal processing units simultaneously.
The framework was tested on an embedded system platform, achieving a 1.96 times speedup and a 51% reduction in energy consumption without adding new processors.
The approach could reduce computer hardware costs, carbon emissions, and the need for freshwater in data processing centers.
Further research is needed to explore system implementation, hardware support, and the applications that would benefit most from SHMT.