Roshan Dathathri

Computer Science · Physical Sciences

15h-index816citations39works0.22yr avg
OpenAlex↗ORCID↗Google Scholar↗

Accepting Students?

No reports yet. Know if this professor is taking students?

Research Topics

Graph Theory and Algorithms(26), Parallel Computing and Optimization Techniques(16), Advanced Graph Neural Networks(14), Cloud Computing and Resource Management(9), Interconnection Networks and Systems(6)

Publications39 total

MSCCL++ ASPLOS'26 Artifact
Zenodo (CERN European Organization for Nuclear Research)·2026Open Access
MSCCL++ ASPLOS'26 Artifact
Zenodo (CERN European Organization for Nuclear Research)·2026Open Access
Flashlight: PyTorch Compiler Extensions to Accelerate Attention Variants
arXiv (Cornell University)·2025Open Access
MSCCL++: Rethinking GPU Communication Abstractions for AI Inference
ArXiv.org·2025Open Access
Kimbap: A Node-Property Map System for Distributed Graph Analytics
2024· 1 citedOpen Access
Optimizing Graph Transformer Networks with Graph-based Techniques
arXiv (Cornell University)·2021Open Access
Sandslash
2021· 39 citedOpen Access
CuSP
ACM SIGOPS Operating Systems Review·2021· 6 cited
Distributed Training of Embeddings using Graph Analytics
2021Open Access
Sandslash: A Two-Level Framework for Efficient Graph Pattern Mining
arXiv (Cornell University)·2020· 1 citedOpen Access
A Study of APIs for Graph Analytics Workloads
2020· 1 cited
Evaluation of Graph Analytics Frameworks Using the GAP Benchmark Suite
2020· 25 cited
Compiler and runtime systems for homomorphic encryption and graph processing on distributed and heterogeneous architectures
2020Open Access
A Study of Graph Analytics for Massive Datasets on Distributed Multi-GPUs
2020· 15 cited
Single machine graph analytics on massive datasets using Intel optane DC persistent memory
Proceedings of the VLDB Endowment·2020· 11 citedOpen Access
Pangolin
Proceedings of the VLDB Endowment·2020· 90 cited
An Adaptive Load Balancer For Graph Analytical Applications on GPUs
arXiv (Cornell University)·2019· 2 citedOpen Access
Pangolin: An Efficient and Flexible Graph Mining System on CPU and GPU
arXiv (Cornell University)·2019· 6 citedOpen Access
Pangolin: An Efficient and Flexible Graph Pattern Mining System on CPU and GPU
arXiv (Cornell University)·2019· 2 citedOpen Access
Distributed Word2Vec using Graph Analytics Frameworks.
arXiv (Cornell University)·2019Open Access
page 1 of 2Next →

Frequent Co-authors

Gurbinder Gill(25), Keshav Pingali(25), Loc Hoang(19), Olli Saarikivi(6), Xuhao Chen(6), Vishwesh Jatala(5), Saeed Maleki(4), Todd Mytkowicz(4), Hoang-Vu Dang(4), Marc Snir(4), Alex Brooks(3), Nikoli Dryden(3), Uday Bondhugula(3), Bozhi You(3), V. Krishna Nandivada(3), Abhinav Jangda(3), Sreevatsa Anantharamu(3), Hao Chen(2), Kim Laine(2), Kristin Lauter(2)