Minimizing Energy and Computation in Long-Running Software

Long-running software may operate on hardware platforms with limited energy resources such as batteries or photovoltaic, high-performance that consume a large amount of energy. Since systems be subject to failures, checkpointing is often used assure the reliability application. introduces additional computation time and consumption, we study how checkpoint intervals need selected so minimize cost function includes execution Expressions for both program’s consumption are derived failure probability per instruction. A first principle based analysis yields interval minimizes linear combination average program, in terms classical “Lambert function”. The sensitivity importance attributed also derived. results illustrated numerical examples regarding programs various lengths showing relation between time, one weighted sum two. In addition, our applied popular benchmark, posted publicly accessible web site, together optimization have developed.