Principal Kernel Analysis: A Tractable Methodology to Simulate Scaled GPU Workloads
Cesar Avalos Baddouh, Mahmoud Khairy, Roland N. Green, Mathias Payer, Timothy G. Rogers
Abstract
Simulating all threads in a scaled GPU workload results in prohibitive simulation cost. Cycle-level simulation is orders of magnitude slower than native silicon, the only solution is to reduce the amount of work simulated while accurately representing the program.
Topics & Concepts
Computer scienceParallel computingKernel (algebra)WorkloadGeneral-purpose computing on graphics processing unitsComputational sciencePrincipal (computer security)Work (physics)Computer graphics (images)Operating systemGraphicsMechanical engineeringMathematicsCombinatoricsEngineeringParallel Computing and Optimization TechniquesDistributed and Parallel Computing SystemsAdvanced Data Storage Technologies