Amar Phanishayee

Bellevue Hospital Center — United States

Computer Science · Physical Sciences

26h-index4.2kcitations87works3.02yr avg
OpenAlex↗ORCID↗Google Scholar↗Lab page↗

Accepting Students?

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

Research Topics

Advanced Neural Network Applications(26), Advanced Data Storage Technologies(19), Parallel Computing and Optimization Techniques(18), Cloud Computing and Resource Management(18), IoT and Edge/Fog Computing(14)

Publications87 total

Scaling Llama 3 Training with Efficient Parallelism Strategies
2025· 2 citedOpen Access
Forecasting GPU Performance for Deep Learning Training and Inference
2025· 12 citedOpen Access
Integrated Hardware Architecture and Device Placement Search
arXiv (Cornell University)·2024Open Access
Forecasting GPU Performance for Deep Learning Training and Inference
arXiv (Cornell University)·2024Open Access
Workload-Aware Hardware Accelerator Mining for Distributed Deep Learning Training
arXiv (Cornell University)·2024Open Access
Blox: A Modular Toolkit for Deep Learning Schedulers
2024· 6 cited
DéjàVu: KV-cache Streaming for Fast, Fault-tolerant Generative LLM Serving
arXiv (Cornell University)·2024· 1 citedOpen Access
Blox: A Modular Toolkit for Deep Learning Schedulers
arXiv (Cornell University)·2023Open Access
Packrat: Automatic Reconfiguration for Latency Minimization in CPU-based DNN Serving
arXiv (Cornell University)·2023Open Access
A Study on the Intersection of GPU Utilization and CNN Inference
arXiv (Cornell University)·2022· 4 citedOpen Access
Harmony
Proceedings of the VLDB Endowment·2022· 17 cited
Harmony: Overcoming the Hurdles of GPU Memory Capacity to Train Massive DNN Models on Commodity Servers
arXiv (Cornell University)·2022· 1 citedOpen Access
Piper: Multidimensional Planner for DNN Parallelization
Neural Information Processing Systems·2021· 12 cited
Efficient large-scale language model training on GPU clusters using megatron-LM
2021· 38 citedOpen Access
Synergy: Resource Sensitive DNN Scheduling in Multi-Tenant Clusters
arXiv (Cornell University)·2021· 1 citedOpen Access
Memory-Efficient Pipeline-Parallel DNN Training
2021· 3 cited
Boosting the Throughput and Accelerator Utilization of Specialized CNN Inference Beyond Increasing Batch Size
International Conference on Machine Learning·2021· 7 cited
Doing more with less
2021· 5 citedOpen Access
Efficient Large-Scale Language Model Training on GPU Clusters
arXiv (Cornell University)·2021· 31 citedOpen Access
CheckFreq: Frequent, Fine-Grained DNN Checkpointing.
File and Storage Technologies·2021· 25 cited
page 1 of 5Next →

Frequent Co-authors

David G. Andersen(12), Deepak Narayanan(9), Vijay Vasudevan(9), Gregory R. Ganger(7), Matei Zaharia(7), Shivaram Venkataraman(7), Elie Krevat(6), Garth A. Gibson(6), Nikhil R. Devanur(5), Ratul Mahajan(5), Hiral Shah(4), Gennady Pekhimenko(4), Phillip B. Gibbons(3), Jason Franklin(3), Michael Kaminsky(3), Lawrence Tan(3), Hongyu Zhu(3), Myeongjae Jeon(3), Junjie Qian(3), Wencong Xiao(3)