Key Moments
- Meta Platforms (NASDAQ:META) shares fell 2.6% Thursday morning as investors focused on the scale of its planned AI infrastructure spending.
- The company intends to start producing its in-house “Iris” AI data center chip in September as part of a four-generation Meta Training and Inference Accelerators (MTIA) roadmap.
- An internal memo indicated Meta could spend up to $145 billion on AI infrastructure this year, aiming to deploy seven gigawatts of capacity and double that by 2027.
Wall Street Reacts to Meta’s AI Ambitions
Meta Platforms (NASDAQ:META) traded lower Thursday morning, with the stock down 2.6%, as investors weighed the company’s aggressive artificial intelligence buildout against the near-term pressure it could put on profitability. The market appeared more focused on the magnitude of Meta’s planned capital expenditures than on its latest hardware progress.
In-House AI Chip “Iris” Set to Begin Production
Meta outlined plans to begin manufacturing its first proprietary AI chip in September, aiming to eventually reduce its computing costs. According to an internal memo reviewed by Reuters, the chip – code-named “Iris” – is a key element of a four-generation Meta Training and Inference Accelerators (MTIA) program developed entirely in-house.
The company is collaborating with Broadcom on the chip’s design and relying on Taiwan Semiconductor Manufacturing Co. (TSMC) for fabrication. The strategic objective is to significantly cut Meta’s reliance on high-priced third-party providers such as Nvidia and Advanced Micro Devices.
Massive AI Infrastructure Budget Alarms Investors
The memo also highlighted the extraordinary financial commitment required to support Meta’s AI ambitions. The company plans to deploy seven gigawatts of computing infrastructure this year and intends to double that capacity by 2027.
To fund this expansion, Meta expects to spend as much as $145 billion on AI infrastructure this year alone. That figure represents a substantial portion of the roughly $700 billion in capital expenditures projected for the broader Big Tech industry, and it was this level of spending that appeared to pressure the stock.
Historically, investors have tended to punish large technology companies when heavy upfront infrastructure investment threatens near-term margins, especially before there is clear, quantifiable revenue tied to that spending.
| AI Initiative | Detail |
|---|---|
| First in-house AI chip | “Iris” data center chip, production starting in September |
| MTIA program | Four-generation Meta Training and Inference Accelerators, designed in-house |
| Infrastructure deployment (this year) | Seven gigawatts of computing capacity |
| Capacity target by 2027 | Double current seven-gigawatt level |
| AI infrastructure spending (this year) | Up to $145 billion |
Supply Deals and “Chipflation” Add to Cost Pressures
In parallel with its infrastructure rollout, Meta is securing long-term supply agreements with Samsung Electronics, Sandisk, and Sumitomo Electric to ensure access to critical components for its expansion.
These deals are being struck at a time when major technology players are simultaneously racing to scale their data center capacity, driving up prices for memory and AI chips. Morgan Stanley analysts recently warned that this price surge has led to “chipflation,” elevating infrastructure costs to the point where they have become a broader macroeconomic concern. According to the analysts, this trend could further compress Meta’s margins in the coming quarters.
TradingPedia.com is a financial media specialized in providing daily news and education covering Forex, equities and commodities. Our academies for traders cover 
