Search papers, labs, and topics across Lattice.
Hanyang University
1
0
ShuntServe achieves up to 1.42x higher throughput and 31.9% cost efficiency improvements by leveraging heterogeneous GPU clusters for LLM serving.
