Meta unveiled a new AI chip.
Meta revealed the details of its next-generation in-house AI accelerator chip on the 10th (local time).
Internally referred to as Artemis, the new chip is expected to help Meta reduce its reliance on Nvidia’s AI chips and cut overall energy costs.
Meta stated in a social media post, “The architecture of this chip is fundamentally focused on providing a balanced allocation of computing, memory bandwidth, and memory capacity for our ranking and recommendation models.”
The new Meta Training and Inference Accelerator (MTIA) chip is a component of the company’s more considerable custom silicon efforts, with additional hardware systems also under consideration.
The chip is deployed in data centers and contributes to AI application services.
Meta announced it is “implementing several programs to broaden the scope of MTIA to support generative AI workloads.”
Meta has also made significant investments in developing software needed to utilize the power of the infrastructure most efficiently, in addition to creating chips and hardware.
In particular, it has invested billions of dollars in purchasing Nvidia and other AI chips.
CEO Mark Zuckerberg revealed earlier this year that “the company plans to acquire approximately 350,000 flagship H100 chips from Nvidia, and intends, along with other suppliers, to accumulate 600,000 H100 chips by the end of the year.”
Most Commented