Meta (META) Stock Climbs on Custom AI Chip Deployment Strategy
π META shares climbed up to 2% before settling with a gain after reports of accelerated deployment for custom AI chips Arke and Astrid.
π¬ The third-generation MTIA 450 (Arke) chip is currently undergoing testing with performance metrics matching design simulations within 2-3% variance.
β‘ Meta's custom silicon aims to reduce dependence on Nvidia hardware while improving energy efficiency and lowering infrastructure expenses for AI inference.
π A fourth-generation chip named Astrid is projected to complete its design phase in approximately 30 days with deployment targeted for late 2027.
π° Meta VP Yee Jiun Song stated that operating at current scale makes absorbing a 30% cost elevation from Nvidia financially impractical, validating the custom chip strategy.
π€ The proprietary processors are being developed through partnerships with Broadcom for design and TSMC for manufacturing operations.
π§ Engineering teams tested the Arke processor running Meta's proprietary AI models alongside third-party tools from DeepSeek and Alibaba successfully.
π Meta terminated a previous Olympus chip initiative scheduled for 2028-2029 due to cost considerations, focusing now exclusively on inference operations.
π Wall Street maintains a Strong Buy consensus on META with a mean 12-month price objective of $758.03 indicating 13.5% appreciation potential.
- Shares climbed up to 2% before settling with a gain following reports of accelerated deployment for custom AI chips Arke and Astrid.
- The third-generation MTIA 450 (Arke) chip is currently undergoing testing with performance metrics matching design simulations within 2-3% variance.
- Meta's custom silicon aims to reduce dependence on Nvidia hardware while improving energy efficiency and lowering infrastructure expenses for AI inference.
- A fourth-generation chip named Astrid is projected to complete its design phase in approximately 30 days with deployment targeted for late 2027.
- Meta VP Yee Jiun Song stated that operating at current scale makes absorbing a 30% cost elevation from Nvidia financially impractical, validating the custom chip strategy.
- Engineering teams tested the Arke processor running Meta's proprietary AI models alongside third-party tools from DeepSeek and Alibaba successfully.
- Wall Street maintains a Strong Buy consensus on META with a mean 12-month price objective of $758.03 indicating 13.5% appreciation potential.