Search papers, labs, and topics across Lattice.
1
0
3
Forget A100s for long-context LLMs – Salca achieves up to 74x better energy efficiency with a sparsity-aware hardware accelerator.
