Meta Platforms plans to manufacture an AI chip starting September, aiming to boost its computing power to 14 gigawatts next year, according to an internal memo.
New Delhi, India Jul 10, 2026 ALN: Meta Platforms, the parent company of social media giants Facebook and Instagram, is set to embark on a significant technological initiative by starting the production of an artificial intelligence (AI) chip in September. This move is part of a broader strategy aimed at enhancing the company's overall computing capacity, with a target of achieving 14 gigawatts of computing power by the year 2027, as revealed in an internal memo reviewed by Reuters. The memo highlights the ambitious plans of a company that has been increasingly investing in AI technology to maintain its competitive edge in the rapidly evolving tech landscape.
The chip, internally referred to by its code name “Iris,” is a product of Meta's four-generation project known as Meta Training and Inference Accelerators (MTIA). This initiative aims to design chips in-house, utilizing custom-built silicon that is specifically tailored to meet the unique demands of Meta's AI applications. The decision to develop proprietary chips underscores a growing trend among major tech firms to gain more control over their hardware resources, particularly as the demand for advanced AI capabilities continues to surge.
The testing phase for the Iris chip was notably swift, taking only six weeks to complete without encountering any major issues. This rapid progress is a promising sign for Meta, especially considering that its previous efforts in developing in-house chips have faced challenges and delays since the initiative began over five years ago. The ability to quickly test and validate new technology is crucial in the fast-paced world of AI, where advancements can significantly alter competitive dynamics.
Meta’s collaboration with industry leaders such as Broadcom and Taiwan Semiconductor Manufacturing Co. (TSMC) is indicative of the company's strategic approach to chip development. By partnering with these established firms, Meta aims to leverage their expertise in chip design and manufacturing, which could lead to more efficient production processes and cost reductions. This is particularly important for Meta, as the company has been heavily reliant on external suppliers like Nvidia and Advanced Micro Devices (AMD) for graphics processing units (GPUs) that are essential for running AI applications.
The memo pointed out the difficulties Meta has faced in adopting the latest GPUs, describing the process as a “heavy lift” that has consumed valuable time and resources. This highlights the broader challenges that tech companies encounter as they scale their AI operations, particularly in securing the necessary hardware to support their ambitions. Following the report of its chip production plans, Meta’s shares experienced a brief decline before rebounding after the company announced developer access to an AI coding model, positioning itself as a competitor to firms like OpenAI and Anthropic.
Industry analysts have noted that independence from external chip suppliers is critical for companies like Meta that aspire to be leaders in AI technology. Mike Gualtieri, a vice president and principal analyst at research firm Forrester, emphasized that becoming an AI titan necessitates a self-sufficient chip production capability. This sentiment is echoed across the tech industry, where companies are increasingly recognizing that reliance on third-party suppliers can hinder innovation and competitiveness.
Meta's commitment to chip development is part of a larger trend among major tech companies, including hyperscalers and firms like SpaceX, which are also investing in custom chip designs to enhance their computing capabilities. The rapid pace of technological advancements in AI and machine learning has created a pressing need for tailored hardware solutions that can efficiently handle complex computational tasks.
In addition to its chip production plans, Meta has ambitious goals for expanding its computing infrastructure. The company aims to deploy a total of seven gigawatts of computing power this year, having already added one gigawatt in the first half of the year and projecting an additional 5.5 gigawatts by year-end. To put this into perspective, one gigawatt of energy can power approximately 800,000 homes, illustrating the scale of Meta's infrastructure ambitions.
Looking ahead, Meta plans to double its computing capacity again by 2027, aiming for a total of 14 gigawatts. The company’s projected expenditure on AI infrastructure for this year is expected to reach as high as $145 billion, a substantial investment that reflects the broader trend among Big Tech companies, which are collectively projected to spend over $700 billion on AI technologies in the coming years.
To support its ambitious infrastructure expansion, Meta has secured long-term, multi-year supply agreements with key suppliers. These agreements include partnerships with Samsung Electronics for memory chips, Sandisk for flash storage, and Sumitomo Electric for fiber-optic equipment. Such long-term contracts are becoming increasingly vital for companies looking to expand their data center capabilities, especially in light of ongoing supply chain challenges and a global shortage of memory chips that have forced some companies, including Apple, to raise prices.
While Sandisk has chosen not to comment on its partnership with Meta, both Samsung Electronics and Sumitomo Electric have not responded to inquiries regarding their agreements. The importance of these partnerships cannot be overstated, as the demand for components like memory and AI chips has skyrocketed amid the rapid expansion of data centers, driven by the insatiable appetite of AI applications for computing power.
This surge in demand has led to a phenomenon referred to as “chipflation,” where rising prices for chips and other components have become a macroeconomic concern. Analysts at Morgan Stanley have highlighted that the escalating costs associated with chips are influencing broader economic trends, as companies strive to keep pace with the growing needs of AI technology.
In conclusion, Meta's decision to begin AI chip production marks a pivotal moment in its ongoing effort to enhance its computing capabilities and reduce reliance on external suppliers. As the company prepares to roll out the Iris chip and expand its infrastructure, it is positioning itself to compete more effectively in the AI landscape. The implications of these developments extend beyond Meta, as they reflect larger trends in the tech industry where companies are increasingly investing in custom hardware solutions to meet the demands of advanced AI applications.
To learn more about the latest developments in Corporate & Industry Insights, stay updated with our exclusive reports and analyses on AiLensNews.
