@misc{sarkar2026compute, title={Compute-Optimal Is Not Cluster-Optimal: Systems-Aware Scaling for Sparse Mixture-of-Experts}, author={Soumajyoti Sarkar and Yuxin Tang and Sheng Zha}, year={2026}, eprint={2608.10605}, archivePrefix={arXiv}, primaryClass={cs.LG}, url={https://arxiv.org/abs/2608.10605} }