メタがMTIA 300トレーニングチップを発表、内蔵NIC搭載

By co-designing MTIA’s communication library, HCCL, alongside the chip we’ve taken a fundamentally different approach to chip design and made communication a first-class citizen.

with MTIA 300, the first of the MTIA family optimized for training recommendation and ranking models

We’re sharing how MTIA 300’s built-in NIC chiplets allow it to meet the communication needs associated with training recommendation models with superior performance over general-purpose GPUs.

MTIA 300 is the first of Meta’s family of in-house training and inference accelerators optimized for training ranking and recommendation models.

recommendation models are bottlenecked by a need for fast and efficient communication between the accelerators that train them

With MTIA 300, the network interface lives inside the chip package itself

Our communication library, HCCL, was co-designed with MTIA 300.

the first of the MTIA family optimized for training recommendation and ranking models

MTIA 300’s total communication time is 3.9 times faster than the equivalent GPU cluster.