FIL/IPFS Hardware Acceleration

[dvo-orlando]

Hi all, I'm working on an academic project to implement an algorithm in hardware on an IC. I'm excited about distributed computing and FIL/IPFS specifically. Are there any algorithms within this field that would benefit from hardware-acceleration?
I'm thinking along the lines of accelerated CID hashing or snark-compression; small algorithms or processes that are executed many many times on the scale of a massive distributed network.

[cryptonemo]

Pinging @dignifiedquire

[dignifiedquire]

Algorithms that might be of interested to have accelerated from a Filecoin perspective

• Poseidon Hash function
• Sha256 Hash function (already very optimized with AMD cpus, but might be interesting for non AMD CPUs)
• Elliptic curve operations as required by bellperson (zkSNARK proving system)
• FFT & MultiExp as used by bellperson (currently accelerated on the GPU, might be possible to do faster)