search

FastSIKE2019 / generic

SIKE
MIT License
2 stars 0 forks source link

Comparions with optimized implementation. #1

Open armfazh opened 4 years ago

armfazh commented 4 years ago

How does this code compare against the optimized (not the generic) implementation?

FastSIKE2019 commented 4 years ago

Thanks for your attention. Do you mean the implementation with the Assembly code for x64 platform? Yes, we just made a fair comparison in C code with the portable optimized implementation (not using the GMP library or the Assembly code). Recently, we have deployed our new SIKE algorithm on an FPGA platform and achieved favorable results. For example, for the SIKEp751, the total time for the encap and decap is only 13.2 ms for a Virtex-7 FPGA, about 2x better than the fastest x64 software implementation (24.9 ms for an Intel i7-6700 CPU) in the SIKE library (https://github.com/microsoft/PQCrypto-SIDH).

Exactly, I have to say that I'm more familar with the Verilog language than the Assembly language. Recently, we are looking for to cooperate with someone who is good at the Assembly language for AMD or ARM platform.

FastSIKE2019 commented 4 years ago

If you are interested in, please leave a message.