@article{memoryandbandwidthareallyouneedfor, title = {Memory and Bandwidth are All You Need for Fully Sharded Data Parallel}, author = {Jiangtao Wang and Jan Ebert and Oleg Filatov and Stefan Kesselheim}, year = {2025}, eprint = {2504.03655}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2504.03655v1}, }