AIConfPaper
← Search

Charith Mendis

6 accepted papers

2025

COGNATE: Acceleration of Sparse Tensor Programs on Emerging Hardware using Transfer Learning

Chamika Sudusinghe, Gerasimos Gerogiannis, Damitha Lenadora, Charles Block, Josep Torrellas, Charith Mendis
ICML 2025poster

Sparse tensor programs are essential in deep learning and graph analytics, driving the need for optimized processing. To meet this demand, specialized hardware accelerators are being developed. Optimizing these programs for accelerators is challenging for two reasons: program performance is highly s…

Cited by 0Source ↗PDF ↗Scholar ↗
2023

Learning Large Graph Property Prediction via Graph Segment Training

Kaidi Cao, Phitchaya Mangpo Phothilimthana, Sami Abu-El-Haija, Dustin Zelle, Yanqi Zhou, Charith Mendis +2
NeurIPS 2023poster

Learning to predict properties of large graphs is challenging because each prediction requires the knowledge of an entire graph, while the amount of memory available during training is bounded. Here we propose Graph Segment Training (GST), a general framework that utilizes a divide-and-conquer appro…

Cited by 13Source ↗PDF ↗code ↗Scholar ↗
2023

TpuGraphs: A Performance Prediction Dataset on Large Tensor Computational Graphs

Phitchaya Mangpo Phothilimthana, Sami Abu-El-Haija, Kaidi Cao, Bahare Fatemi, Michael Burrows, Charith Mendis +1
NeurIPS 2023poster

Precise hardware performance models play a crucial role in code optimizations. They can assist compilers in making heuristic decisions or aid autotuners in identifying the optimal configuration for a given program. For example, the autotuner for XLA, a machine learning compiler, discovered 10–20\% s…

Cited by 16Source ↗PDF ↗code ↗Scholar ↗
2019

Compiler Auto-Vectorization with Imitation Learning

Charith Mendis, Cambridge Yang, Yewen Pu, Dr.Saman Amarasinghe, Michael Carbin
NeurIPS 2019poster

Modern microprocessors are equipped with single instruction multiple data (SIMD) or vector instruction sets which allow compilers to exploit fine-grained data level parallelism. To exploit this parallelism, compilers employ auto-vectorization techniques to automatically convert scalar code into vect…

Cited by 52Source ↗PDF ↗code ↗Scholar ↗
2019

Ithemal: Accurate, Portable and Fast Basic Block Throughput Estimation using Deep Neural Networks

Charith Mendis, Alex Renda, Dr.Saman Amarasinghe, Michael Carbin
ICML 2019oral

Predicting the number of clock cycles a processor takes to execute a block of assembly instructions in steady state (the throughput) is important for both compiler designers and performance engineers. Building an analytical model to do so is especially complicated in modern x86-64 Complex Instructio…

Cited by 206Source ↗PDF ↗code ↗Scholar ↗
2016

Parallelizing WFST speech decoders

Charith Mendis, Jasha Droppo, Saeed Maleki, Madanlal Musuvathi, Todd Mytkowicz, Geoffrey Zweig
ICASSP 2016accepted

The performance-intensive part of a large-vocabulary continuous speech-recognition system is the Viterbi computation that determines the sequence of words that are most likely to generate the acoustic-state scores extracted from an input utterance. This paper presents an efficient parallel algorithm…

Cited by 0Source ↗Scholar ↗