Story
Meta to Begin In-House AI Chip Production in September, Memo Shows
Summary
An internal memo reveals Meta Platforms plans to start manufacturing its custom AI chip in September as part of an aggressive push to double its computing capacity and reduce reliance on external suppliers like Nvidia.
Meta Platforms is set to begin manufacturing its first in-house artificial intelligence chip in September, according to an internal memo reviewed by Reuters. The move is a key part of the company's strategy to significantly expand its computing infrastructure and reduce its dependence on third-party chip suppliers.
In-House Silicon Initiative
The custom chip, codenamed "Iris," is part of a four-generation project for Meta Training and Inference Accelerators (MTIA) designed to power the AI features on its Facebook and Instagram platforms. The memo stated that testing of the chip was completed in just six weeks with no major issues, a positive development for an internal effort that began over five years ago.
Meta is collaborating with Broadcom on the chip's design and with Taiwan Semiconductor Manufacturing Co. (TSMC) for its production. This custom silicon is intended to augment, not replace, the large quantities of GPUs Meta purchases from market leaders Nvidia and AMD. The primary goals are to lower massive computing costs and gain more independence in its hardware supply chain.
Aggressive Infrastructure Expansion
The chip production supports an ambitious plan to scale Meta's computing power. According to the memo, the company has set aggressive targets for its infrastructure:
Ad- Deploy 7 gigawatts of computing power in the current year.
- Double that capacity to a total of 14 gigawatts by 2027.
To fuel this expansion, Meta expects to spend as much as $145 billion on AI infrastructure this year. The memo also noted that the company has secured long-term, multi-year supply agreements for critical components with partners including Samsung Electronics for memory chips and Sandisk for flash storage.
Market Implications
Meta's foray into custom chip production reflects a broader industry trend where tech giants, or "hyperscalers," develop their own silicon to gain a competitive edge in the AI arms race. This strategy allows for hardware tailored to specific workloads, potentially leading to better performance and cost efficiency.
The intense demand for AI components has created supply chain pressures and rising prices, a phenomenon some analysts have called "chipflation." By developing its own chips, Meta aims to mitigate some of these external risks and control a more significant part of its technology stack. Meta declined to comment on the memo's contents.