Sunday, December 7, 2025

Meta Acquires RISC-V Chip Startup Rivos for AI Push

Meta​ Reportedly pursues Acquisition of RISC-V AI GPU Firm Rivos

Meta ⁣is in talks to acquire Rivos,⁢ a semiconductor startup specializing in RISC-V based GPU⁣ technology, according​ to sources ‍familiar with the matter. The​ potential acquisition would considerably bolster Meta’s internal silicon development​ teams working on its ambitious ​”personal superintelligence” initiatives and ⁣could position ​the company to reduce its reliance on Nvidia for AI processing⁣ power.

The move comes as‌ Meta accelerates its investment in custom silicon, aiming to create specialized chips optimized for artificial intelligence workloads. CEO Mark Zuckerberg has reportedly ⁤expressed dissatisfaction with‌ the pace of development on‍ the company’s MTIA (Meta Training and Inference Accelerator) project, with the‍ next generation chip still some time away‌ from completion. Acquiring ⁤Rivos would provide Meta with a team of experienced chip designers and⁢ a⁤ foundational RISC-V GPU architecture to accelerate its AI hardware roadmap.

Rivos was founded by a team of former Apple engineers and has been developing ‍high-performance, energy-efficient GPUs based on the open-source RISC-V instruction set architecture. This contrasts with Meta’s current reliance on Arm-based designs for its MTIA chips. In 2022, Apple filed ⁤a lawsuit against Rivos,‌ alleging that the startup had recruited 40 former‍ Apple employees and misappropriated‍ trade secrets. The legal dispute was settled in February 2024 after Rivos countersued,⁤ claiming Apple attempted to stifle⁢ employee movement to the​ new company. ‌

The financial terms of a potential acquisition have not been disclosed.However, industry ‍analysts suggest that‍ Rivos’ expertise in ⁤RISC-V GPU ‍design and its team of experienced engineers make it a valuable asset for Meta as it seeks to build a⁢ vertically ⁣integrated AI hardware stack. A ⁤accomplished acquisition could allow meta to exert greater control over its AI infrastructure and possibly lower costs associated with procuring GPUs from external vendors ‍like Nvidia.

You may also like

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.