Key Moments
- NVIDIA (NASDAQ: NVDA) may alter its upcoming Feynman AI chip platform because of constrained 2 nanometer capacity at TSMC (NYSE: TSM), according to Taiwan’s Economic Daily News.
- TSMC’s advanced 2 nanometer production is reportedly fully booked until 2028 and potentially beyond, driven by strong demand from major AI customers including Nvidia and Meta.
- Feynman, unveiled in 2025 with a planned 2028 launch, is slated to follow Nvidia’s Vera Rubin platform, which is expected to start shipping later this year.
Capacity Strain at TSMC Puts Pressure on Nvidia Roadmap
Investing.com – NVIDIA (NASDAQ: NVDA) may be forced to rework the design of its next-generation Feynman artificial intelligence chip platform as it grapples with limited manufacturing availability at leading contract chipmaker TSMC (NYSE: TSM). The potential shift was reported by Taiwan’s Economic Daily News on Monday.
According to the report, TSMC is seeing exceptional demand for its advanced 2 nanometer process technology from major AI players such as Nvidia and Meta. This surge has pushed the foundry’s 2 nanometer capacity to be fully allocated through 2028 and possibly beyond.
Pricing and Capacity Dynamics at the World’s Largest Foundry
The report indicated that Nvidia’s Feynman platform could require a redesign in light of the tight capacity situation. TSMC is also expected to raise prices, reflecting both its constrained production capacity and the rapid acceleration in AI-driven demand.
The world’s largest contract chip manufacturer has seen substantial gains from the expansion of AI-related workloads over the past three years, the report noted.
Nvidia’s AI Platform Roadmap: Vera Rubin and Feynman
Feynman is described as a next-generation AI processing platform that Nvidia unveiled in 2025, with a targeted launch in 2028. It is positioned as the successor to Nvidia’s Vera Rubin platform, which is expected to begin shipping later this year.
| Platform | Company | Key Timing | Notes |
|---|---|---|---|
| Vera Rubin | NVIDIA (NASDAQ: NVDA) | Set to begin shipping later this year | Predecessor to Feynman platform |
| Feynman | NVIDIA (NASDAQ: NVDA) | Unveiled in 2025; planned release in 2028 | Next-generation AI processing platform that may face redesign due to TSMC capacity limits |
| 2 nanometer process | TSMC (NYSE: TSM) | Capacity fully booked until 2028 and potentially beyond | High demand from AI majors including Nvidia and Meta |





