GREENER: A Tool for Improving Energy Efficiency of Register Files
share
Graphics Processing Units (GPUs) maintain a large register file to increase thread block occupancy, hence to improve the thread level parallelism (TLP). However, register files in the GPU dissipate a significant portion of the total leakage power. Leakage power of the register file can be reduced by putting the registers into low power (SLEEP or OFF) state. However, one challenge in doing so is the lack of precise register access information for each instruction at run-time. This paper proposes GREENER (GPU REgister file ENErgy Reducer): a tool for minimizing leakage energy of the register file of GPUs. GREENER employs a compile-time analysis to estimate the run-time register access information. The result of the analysis is used to determine the power state of the registers (ON, SLEEP, or OFF) after each instruction. We propose a new power optimized assembly instruction format that allows GREENER to encode the power state of the registers with its instruction. Further, GREENER transforms a given input assembly language to a power optimized assembly. The optimized assembly, along with a run-time optimization to update the power state of a register during execution, results in significant power reduction. We implemented GREENER in GPGPU-Sim simulator and used GPUWattch framework to measure the leakage power of register file. We evaluated the effectiveness of GREENER on 21 kernels from CUDASDK, GPGPU-SIM, Parboil, and Rodinia benchmarks suites. We observe an average reduction of register leakage energy by 69.2 maximum reduction of 88.41 slowdown on average).
READ FULL TEXT