Charith Mendis
Massachusetts Institute of Technology
35 Papers
105 Citations
Charith Mendis is an academic researcher from Massachusetts Institute of Technology. The author has contributed to research in topics: Computer science & Compiler. The author has an hindex of 10, co-authored 18 publications. Previous affiliations of Charith Mendis include Google & University of Illinois at Urbana–Champaign.
Chat about Author
Papers
Making caches work for graph analytics
Yunming Zhang,Vladimir Kiriansky,Charith Mendis,Saman Amarasinghe,Matei Zaharia +4 more
- 01 Jul 2017
TL;DR: Cagra achieves speedups of up to 5× for PageRank, Collaborative Filtering, Label Propagation and Betweenness Centrality over the best published results from state-of-the-art graph frameworks, including GraphMat, Ligra and GridGraph.
110
•Posted Content
Ithemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks
TL;DR: Ithemal as discussed by the authors uses a hierarchical LSTM-based approach to predict throughput based on the opcodes and operands of instructions in a basic block, which is more accurate than state-of-the-art hand-written tools currently used in compiler backends and static machine code analyzers.
64
•Proceedings Article
Ithemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks
Charith Mendis,Alex Renda,Saman Amarasinghe,Michael Carbin +3 more
- 24 May 2019
VeGen: a vectorizer generator for SIMD and beyond
Yishen Chen,Charith Mendis,Michael Carbin,Saman Amarasinghe +3 more
- 19 Apr 2021
TL;DR: This paper introduces Lane Level Parallelism (LLP), which captures the model of parallelism implemented by both SIMD and non-SIMD vector instructions, and presents VeGen, a vectorizer generator that automatically generates a vectorization pass to uncover target-architecture-specific LLP in programs while using only instruction semantics as input.
39
Helium: lifting high-performance stencil kernels from stripped x86 binaries to halide DSL code
Charith Mendis,Jeffrey Bosboom,Kevin Wu,Shoaib Kamil,Jonathan Ragan-Kelley,Sylvain Paris,Qin Zhao,Saman Amarasinghe +7 more
- 03 Jun 2015
TL;DR: This paper shows how to automatically lift performance-critical stencil kernels from a stripped x86 binary and generate the corresponding code in the high-level domain-specific language Halide, and shows that new optimized versions of these kernels can replace the originals to rejuvenate the application for newer hardware.
39