Meta commits to partnering with Broadcom to build 1GW custom AI accelerator; agreement extended to 2029. Meta’s initial deployment is 1GW, potentially expanding to several gigawatts in the future. Broadcom shares rise over 3% after hours.
Contents of the Cooperation Agreement
According to a report by Gold Pattern Connect APP, on April 14, 2026 (Tuesday), Meta Platforms (META.US) and Broadcom (AVGO.US) announced a comprehensive agreement to extend their existing cooperation on designing Meta’s in-house developed AI accelerators through 2029.
This long-term agreement marks a further deepening of Meta’s strategic partnership with Broadcom in the self-developed AI infrastructure track, with Broadcom continuing to provide Meta with custom chip design and manufacturing support.
1 GW Custom Chip Deployment
According to the agreement, Meta initially plans to deploy 1 gigawatt (GW) of training and inference accelerators, with the possibility of further expansion to a “multi-GW chip” scale based on Broadcom technology in the future.
A 1 GW deployment is equivalent to the computing power of hundreds of thousands of high-end AI accelerator cards, enough to support the massive AI training and inference needs across Meta’s platforms such as Facebook, Instagram, and WhatsApp. This plan highlights Meta’s determination to make massive capital expenditures in generative AI and recommendation systems.
| Duration of Cooperation | Extended through 2029 |
| Initial Deployment Scale | 1 GW training and inference accelerators |
| Future Plans | Potential expansion to several GW |
| Cooperation Focus | Custom design of Meta's in-house AI accelerators |
Strategic Significance Analysis
Meta’s move aims to reduce its reliance on external GPU vendors while optimizing cost and performance alignment through custom chips. As a world-leading semiconductor design company with deep technical expertise in high-performance network chips and AI accelerators, Broadcom’s closer partnership with Meta is expected to accelerate Meta’s self-sufficiency in AI infrastructure.
For Broadcom, securing Meta’s long-term major orders further solidifies its leading position in the AI custom chip market and is expected to drive stable and high-margin revenue growth in the coming years.
Immediate Market Reaction
After the news was announced, Broadcom’s shares rose more than 3% after-hours, showing the market’s high recognition of this long-term cooperation agreement. Meta’s stock price had already risen 4.41% in regular trading hours on the same day. In combination with the agreement, investors believe that Meta’s execution in AI capital expenditure is further validated.
This collaboration has also injected confidence into the entire AI industry chain, with related sectors such as semiconductor equipment, optical modules, and liquid cooling expected to continue to benefit.
Outlook
In the short term, the stock prices of Meta and Broadcom will be affected by the details of the agreement implementation, Meta’s capital expenditure rhythm, and the overall tech earnings season. In the medium to long term, the advancement of Meta’s in-house AI accelerators will have a profound impact on the industry’s competitive landscape, and Broadcom’s order certainty as a core partner will be significantly improved.
Investors can use the AI pattern recognition and quantitative signal features of the Gold Pattern Connect APP to track the progress of cooperation between Meta and Broadcom, as well as related industry chain developments in real-time.
Editorial Summary
Meta and Broadcom have extended their cooperation on AI accelerators to 2029, planning an initial deployment of 1 GW of custom chips, with future possibilities reaching several GW, underscoring Meta’s commitment to long-term investment in AI infrastructure. This agreement not only reduces Meta’s dependence on external GPUs but also brings Broadcom stable, high-value orders. The positive market response to this cooperation further validates the accelerating trend of in-house development and customization of AI computing power. Investors should continue to pay attention to the pace of agreement implementation and capital expenditure realization, and rationally allocate amidst the ongoing AI theme uptrend.
Frequently Asked Questions
Q: What are the core contents of the Meta–Broadcom cooperation agreement?
A: The two parties will extend their existing cooperation on designing Meta’s in-house AI accelerators through 2029. Meta initially plans to deploy 1 GW of training and inference accelerators, with the possibility of further expansion to multi-GW scale. This long-term agreement aims to increase the self-sufficiency rate of Meta’s AI infrastructure.Q: How large is the 1 GW accelerator deployment?
A: 1 GW is equivalent to the computing power of hundreds of thousands of high-end AI accelerator cards, sufficient to support the massive AI training and inference needs across various Meta platforms. This plan highlights Meta’s significant capital expenditure commitment in the field of generative AI.Q: What is the strategic significance of this cooperation for Broadcom?
A: As a leader in high-performance network chips and AI accelerator design, Broadcom securing Meta’s long-term major order will significantly enhance its order visibility and revenue stability, further strengthening its leading position in the AI custom chip market.Q: How did the market react after the news was announced?
A: Following the announcement, Broadcom’s after-hours stock price rose more than 3%, while Meta’s regular trading for the day already increased by 4.41%. The market believes that Meta’s execution capability regarding AI capital expenditures has been validated, and the related semiconductor industry chain is expected to continue to benefit.Q: How should investors view this cooperation?
A: In the short term, focus on the details of the agreement’s implementation and the pace of Meta’s capital expenditures; in the medium to long term, the advancement of Meta’s self-developed AI accelerators and Broadcom’s order growth potential are promising for the industry’s competitive landscape. It is advised to use the Gold Pattern Connect APP’s AI bulletin feature to monitor related developments and use trading volume and technical indicators to capture investment timing.
Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.
You may also like
Hollywood Plans "Film Production Reshoring"! US Economic "Soft Landing" Welcomes a $249.1 Billion Film Incentive Blueprint
A study shows that federal incentives for film and television production will bring $249.1 billion in revenue to the U.S. economy by 2035 and add 143,500 full-time jobs. The Motion Picture Association has been working with Hollywood unions to launch a campaign for national incentives to better compete with markets such as the United Kingdom and Australia.
Tonight, a "dovish rate hike"?
The Federal Reserve is almost certain to raise interest rates tonight, but the key issue is "what will be said after the hike." Citi characterizes this move as a "fine-tuning" adjustment, suggesting there is no inevitable future rate hike, yet warns that if Chairman Powell does not provide clear forward guidance, it will trigger significant market volatility. Goldman Sachs bluntly stated that there is insufficient economic foundation for this rate hike, with inflation being merely a one-off factor, and expects this to be a "signal-less rate hike."
JPMorgan: "Open source disruption" and "AI safety" are not issues, there is still room for capital expenditure in the next two years, semiconductor equipment will become the "new bottleneck"
JPMorgan believes that open-source models are not a threat, regulatory disruptions are only short-term, and cloud vendors’ leverage remains low—the fundamentals of computing power investment have not changed. It forecasts that the capital expenditure of the seven major tech giants will soar from $443 billion in 2025 to $1.577 trillion in 2027, with semiconductor equipment becoming the core bottleneck of the supply chain and a new round of price increases expected in wafer foundry and advanced packaging.
Micron executive: Storage determines AI limits, substantial new capacity will come after 2028
Micron executive Sumit Sadana stated that memory bandwidth and capacity have become the core factors determining the performance ceiling of AI systems. In the face of structural imbalances between supply and demand, Micron expects to increase capital expenditure to over $45 billion in fiscal year 2027, but due to process complexity, substantive new capacity will not be released until 2028. Additionally, long-term supply agreements are reshaping the industry's business model, and humanoid robots will trigger the next wave of massive demand.
Crypto prices
More
