THROUGHPUT OPTIMIZATION AND RESOURCE ALLOCATION ON GPUS UNDER MULTI-APPLICATION EXECUTION

Platform heterogeneity prevails as a solution to the throughput and computational chal- lenges imposed by parallel applications and technology scaling. Specifically, Graphics Processing Units (GPUs) are based on the Single Instruction Multiple Thread (SIMT) paradigm and they can offer tremendous speed-up for parallel applications. However, GPUs were designed to execute a single application at a time. In case of simultaneous multi-application execution, due to the GPUs’ massive multi-threading paradigm, ap- plications compete against each other using destructively the shared resources (caches and memory controllers) resulting in significant throughput degradation. In this thesis, a methodology for minimizing interference in shared resources and provide efficient con- current execution of multiple applications on GPUs is presented. Particularly, the pro- posed methodology (i) performs application classification; (ii) analyzes the per-class in- terference; (iii) finds the best matching between classes; and (iv) employs an efficient re- source allocation. Experimental results showed that the proposed approach increases the throughput of the system for two concurrent applications by an average of 36% compared to other optimization techniques, while for three concurrent applications the proposed approach achieved an average gain of 23%