On a two-node Apple M2 Ultra cluster, a distributed, unquantized DBRX model generates 6.1 tokens/sec, claimed to be 1.15x more cost-efficient than an 8xH100 NVIDIA system.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DC 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Towards Building Private LLMs: Exploring Multi-Node Expert Parallelism on Apple Silicon for Mixture-of-Experts Large Language Model
On a two-node Apple M2 Ultra cluster, a distributed, unquantized DBRX model generates 6.1 tokens/sec, claimed to be 1.15x more cost-efficient than an 8xH100 NVIDIA system.