I am a Research Scientist at Meta working on ML Compilers.
I'm most interested in programming languages and systems for GPUs. My PhD dissertation (Princeton, 2024) is on VectorVisor—a vectorizing binary translator designed to compile existing, single-threaded WebAssembly programs and run many copies of them using a GPU.
Publications
-
VectorVisor: A Binary Translation Scheme for Throughput-Oriented GPU Acceleration
USENIX ATC '23, July 10th-12th, Boston, USA
Samuel Ginzburg, Mohammad Shahrad, Michael J. Freedman
-
Palette Load Balancing: Locality Hints for Serverless Functions
EuroSys '23, May 8th-12th, Rome, Italy
Mania Abdi, Samuel Ginzburg, Charles Lin, José M Faleiro, Íñigo Goiri, Gohar Chaudhry, Ricardo Bianchini, Daniel S. Berger, Rodrigo Fonseca
-
Serverless Isn’t Server-Less: Measuring and Exploiting Resource Variability on Cloud FaaS Platforms
Workshop on Serverless Computing (WoSC'20), December 7-11, 2020, Delft, Netherlands
Samuel Ginzburg, Michael J. Freedman
-
EnTAP: bringing faster and smarter functional annotation to non-model eukaryotic transcriptomes
Molecular Ecology Resources
Alexander J. Hart, Samuel Ginzburg, Muyang (Sam) Xu, Cera R. Fisher, Nasim Rahmatpour, Jeffry B. Mitton, Robin Paul, Jill L. Wegrzyn
-
Transcriptomic imprints of adaptation to fresh water: parallel evolution of osmoregulatory gene expression in the Alewife
Molecular Ecology
Jonathan P. Velotta, Jill L. Wegrzyn, Samuel Ginzburg, Lin Kang, Sergiusz Czesny, Rachel J. O'Neill, Stephen D. McCormick, Pawel Michalak, Eric T. Schultz