What happened
Nvidia invested $3.5 billion in MediaTek, a Taiwanese chipmaker, to help customers build custom AI chips that integrate with Nvidia's systems. Separately, Anthropic committed $35 billion to rent Nvidia-powered computing capacity from Lambda.
Why it matters
Nvidia aims to become the architecture of AI, not just its dominant chipmaker. By selling more of the factory around the GPU, its revenue opportunity per gigawatt of AI-factory capacity rises from about $18 billion with Hopper to $25 billion with Grace Blackwell and $40 billion with Vera Rubin.
What to watch
The strategy lets customers build chips that compete with Nvidia GPUs but connect via NVLink Fusion into Nvidia's data-center architecture. AWS plans to add 2 million Nvidia GPUs while linking its own Trainium chips more deeply, so Nvidia can profit either way.
Ask the AI about this article →
Summaries like this, in your inbox every morning.
Nvidia's investments and partnerships reflect a shift from selling chips to owning the entire AI factory. The $3.5 billion MediaTek investment and Anthropic's $35 billion commitment to Lambda highlight this strategy. Huang's quote, 'The input is electrons, the output is tokens,' frames AI data centers as factories that convert power into AI output.
The revenue-per-gigawatt figures (from $18 billion with Hopper to $40 billion with Vera Rubin) show that Nvidia expects to capture more value by selling the full stack around the GPU. MediaTek's custom chips can even rival Nvidia's, but they connect via NVLink Fusion, ensuring Nvidia still profits. AWS exemplifies this by adding 2 million Nvidia GPUs while integrating its own Trainium chips, so Nvidia participates either way.
This approach addresses a new bottleneck: money. After chips and power, financing AI capacity now becomes a constraint, which these deals aim to overcome.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
San Jose is positioning itself as a hub for physical AI (AI that operates in the real world, such as robotics)…
AMD's Q2 FY2026 revenue hit $11.54 billion, up 50.1% year over year, with Data Center revenue at $6.72 billion…
Nvidia Corp. CEO Jensen Huang said artificial general intelligence has arrived, following OpenAI's launch of G…
Saudi Arabia's state-backed AI company HUMAIN, led by CEO Tareq Amin, is positioning itself as a neutral hub f…
Alibaba's research division released Qwen-Drive 1.0, an AI model that handles spatial perception, traffic Q&A…
Ineffable Intelligence, the London-based AI startup founded by ex-Google DeepMind researcher David Silver, has…
