What should the fabric accelerate?
The idea is to share hardware across post-quantum cryptography algorithms while keeping data movement fast and memory boundaries secure. This is still at the architecture stage, so the blocks below are planned features, not implemented results.
Architecture scope
- Shared NTT/INTT engine and modular polynomial arithmetic.
- Keccak-f[1600] state hashing support.
- Dedicated X25519 elliptic curve datapath.
- ML-KEM-512/768/1024 and ML-DSA-44/65/87 algorithm targets.
Interfaces and security
APB4 control register interface, AXI4 master/slave data movement, secure scratchpad memory, internal key vault, TRNG/DRBG, and a security root-of-trust controller form the architectural roadmap.
What needs to happen before implementation
Before building it, I need to settle which algorithm parameter sets to support, define the threat model, and estimate the scheduling and memory traffic. Then I can verify correctness and constant-time behavior before moving on to physical design.
← Back to all projects