Search papers, labs, and topics across Lattice.
Affiliation:
2
0
1
8
This work develops lightweight characterization methods to uncover per-chip compute topology and memory affinity and uses the discovered information to make existing fine-grained scheduling asymmetry-aware, considering not only how many resources are allocated but also which physical resources are assigned.
FIBER achieves up to 2.25x speedup in LLM serving by decoupling tensor computation from register ownership, revolutionizing GPU efficiency.
