Meta's self-developed chip has come to light and is said to have entered the small-scale deployment stage.

①Meta Platforms is testing its first self-developed chip for training AI systems, which has begun small-scale deployment, with plans to increase production for large-scale use if testing goes well; ②Meta's self-developed chip is a dedicated accelerator produced in collaboration with Taiwan Semiconductor, which is more energy-efficient than the GPUs typically used for AI workloads.

cls.cn4minutes ago

Nvidia Decline Is a 'Buying Opportunity,' as Wells Fargo Says GTC 'Can't Come Fast Enough'

Seeking Alpha38minutes ago

Apple Leads Decline in Global Smartwatch Shipments in 2024: Counterpoint

Seeking Alpha47minutes ago

Trump Set To Meet Top Executives From Apple, JPMorgan And Other Giants In Business Roundtable Amid Economic Concerns And Market Downturn: Report

Benzinga09:29 ET

Report: Meta collaborates with Taiwan Semiconductor to produce training chips.

Currently, this chip has entered the small-scale deployment phase. If testing goes smoothly, Meta plans to expand production for widespread application. Company Executives at Meta have indicated that they plan to start using self-developed chips for training by 2026, initially applying them to recommendation systems, then expanding to generative AI products such as the Meta AI chatbot.

wallstreetcn09:18 ET

Canaccord Genuity Maintains Buy on Tesla, Maintains $404 Price Target

Benzinga09:03 ET · Ratings