Using Machine Learning Techniques to Detect Parallel Patterns of Multi-threaded Applications

Abstract

Multicore hardware and software are becoming increasingly more complex. The programmability problem of multicore software has led to the use of parallel patterns. Parallel patterns reduce the effort and time required to develop multicore software by effectively capturing its thread communication and data sharing characteristics. Hence, detecting the parallel pattern used in a multi-threaded application is crucial for performance improvements and enables many architectural optimizations; however, this topic has not been widely studied. We apply machine learning techniques in a novel approach to automatically detect parallel patterns and compare these techniques in terms of accuracy and speed. We experimentally validate the detection ability of our techniques on benchmarks including PARSEC and Rodinia. Our experiments show that the k-nearest neighbor, decision trees, and naive Bayes classifier are the most accurate techniques. Overall, decision trees are the fastest technique with the lowest characterization overhead producing the best combination of detection results. We also show the usefulness of the proposed techniques on synthetic benchmark generation.

Modernizing parallel code with pattern analysis

“…Pattern detection has been exploited without subsequently tying back to the source code. Poovey [37] and Deniz [13] instrument parallel code with low-level counters and statistically correlate these to profiles of code representative of predefined patterns. The results are used to select architectural and low-level system software policies.…”

Machine Learning in Compiler Optimization

“…Other machine learning techniques such as Kernel Canonical Correlation Analysis and naive Bayes have also been used in prior works to predict stencil program configurations [84] or detect parallel patterns [85].…”

Automated Parallel Kernel Extraction from Dynamic Application Traces

“…Kernels can also be compiled at this point to run on a more efficient, compatible architecture that specializes in kernels like a GPU or TPU. Kernel classification through ML inference [23] can also direct compiler optimization [24] or the direct exchange of naive code for expertly tuned library calls such as FFTW. Automatically detecting kernels will also enable identifying emergent classes of kernels that have not yet been expertly categorized.…”