Fabless chip startup Cornami, whose architecture is optimized for Fully Homomorphic Encryption and quantum threats, raises a $68M Series C led by Vision Fund 2
Context & Ripple Effects
Cornami's $68M Series C lands in the middle of a run of outsized rounds for specialized fabless silicon: Astera Labs raised a $50M Series C at a $950M valuation, Cerebras pulled in a $250M Series F at $4B for wafer-scale AI chips, and D-Matrix took a $44M Series A weeks earlier for its chiplet-based Nighthawk architecture. Vision Fund 2 — which invested $35.2B and led or co-led the most rounds of any investor in 2021 — is writing this check.
The thesis is deliberately forward-dated: Cornami's architecture targets fully homomorphic encryption and resistance to quantum attacks. The related coverage shows the quantum side of that threat model is itself being funded — Quantum Machines raised $50M for quantum control hardware, and more recently QuamCore raised a $26M Series A with plans to move into chip fabrication.
First-order effects
- Cornami gets $68M to push an FHE-optimized, quantum-resistant architecture from design toward manufacturable silicon, while Vision Fund 2 adds another semiconductor position to a portfolio assembled at its record 2021 pace.
Second-order effects
- Every mega-round in this niche raises the differentiation bar for the others: D-Matrix's chiplet approach and Cerebras' wafer-scale bet now compete with Cornami for the same specialized-silicon capital pool and engineering talent.
- Funding on the quantum side — Quantum Machines' control hardware and QuamCore's move into fabrication — strengthens the very threat model Cornami is built against, pulling enterprise security budgets toward hardware-level cryptography earlier.
Third-order effects
- If post-quantum and FHE workloads become procurement requirements rather than research topics, cryptographic capability shifts from a software layer to a silicon specification, opening an accelerator category alongside AI chips.
- The pattern also concentrates architectural risk in mega-funds: with Vision Fund 2 later reported down 19% against the $49B it invested and posting a $2.1B quarterly loss, late-stage bets on unproven chip architectures mean a handful of concentrated investors effectively decide which computing paradigms get to exist.
The trend: Specialized fabless silicon startups are drawing mega-rounds as encryption and AI workloads pull compute beyond general-purpose processors, with concentrated mega-fund capital choosing which architectures survive.