Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
Affiliation:
2
0
2
1
PAFT, a phase-aware, dynamically adaptable GPU frequency tuning system that reduces energy consumption of training workloads with minimal performance overhead, achieving energy savings of up to 46% with an average overhead of 4%.
FOLD achieves up to 2.09x higher throughput than traditional methods while maintaining 93-97% recall in deduplicating massive datasets for language model training.