Meta Platforms, the force behind Facebook, has revealed its bold strategy to integrate its proprietary artificial intelligence (AI) chips, Artemis, into its data centers in the coming months.
This move aims to decrease reliance on Nvidia’s widespread H100 chips, addressing the rising costs associated with AI workloads.
Investing billions of pounds, Meta is committed to bolstering its computational capabilities for demanding generative AI initiatives across platforms like Facebook, Instagram, and WhatsApp.
These initiatives require significant investments in specialized chips and the restructuring of data center infrastructure.
Dylan Patel, founder of the silicon research group SemiAnalysis, anticipates substantial cost efficiencies for Meta.
The successful implementation of its in-house chip could potentially reduce annual energy expenditures by hundreds of millions and chip procurement costs by billions.
Despite these strides, Meta will persist in deploying Nvidia’s H100 GPUs in its data centers for the foreseeable future.
CEO Mark Zuckerberg outlined plans to have approximately 350,000 H100 processors operational by year-end.
