Meta launches compute division with Daniel Gross leading capacity strategy, targeting tens of gigawatts
Jan 12, 2026
Key Points
- Meta launches a dedicated compute division targeting tens of gigawatts this decade and hundreds long-term, splitting leadership between Santosh Nandy on technical operations and Daniel Gross on capacity strategy.
- Gross, who ran the Andromeda Cluster at Lerer Hippeau providing GPU access to portfolio companies including Perplexity and Cursor, brings proven infrastructure management experience to Meta's buildout.
- Meta hired Dina McCormick, former deputy national security adviser, as president and vice chairman to secure government-backed financing for infrastructure, signaling a shift toward state partnerships for critical AI capacity.
Summary
Meta is launching a dedicated compute division targeting tens of gigawatts of capacity this decade and hundreds of gigawatts longer term. Leadership splits between Santosh Nandy, who oversees technical architecture, silicon, and data center operations, and Daniel Gross, who handles capacity strategy, supplier relationships, industry analysis, and business modeling.
Gross previously ran the Andromeda Cluster at Lerer Hippeau, an early accelerator-backed GPU cluster that became a resource advantage for portfolio companies including Perplexity, Cursor, Replicate, and Chroma. The cluster either leased dedicated GPU instances from cloud providers or built shared infrastructure, but his track record suggests Meta's capacity play is in experienced hands.
Meta also hired Dina McCormick, former deputy national security adviser, as president and vice chairman. She will lead government and sovereign partnerships for financing and deploying infrastructure. Her appointment signals a shift from purely commercial infrastructure deals toward direct state involvement in financing capacity buildout. This positions Meta to tap government-backed capital for infrastructure increasingly viewed as critical to AI development, a move that mirrors competitive pressures from OpenAI and xAI to secure long-term compute supply.