The performance of public key algorithms regularly comes up with customers – they are after all the most costly operations.
wolfSSL has you covered with the Single Precision (SP) big number maths library that brings you the best implementations.
But how does it compare to OpenSSL?
Intel x86 64-bit assembly optimised implementations are as fast or faster than OpenSSL for RSA 2048-bit, EC P-256 and X25519.
One major drawback with OpenSSL is the lack of choice if not using assembly code. The C code is significantly slower, especially the elliptic curve operations which can be 15-40 times slower!
wolfSSL has the option to compile in optimised C code. This code is significantly faster and can be configured to meet your resource profile. Lots of memory, the default build is optimised for maximum speed and handily beats OpenSSL C code. Not so much memory, try the small build, it has similar performance to OpensSSL but half the size!
If you don’t have space for specialised code then good news is coming! Later this year a new implementation of MP, our big number math API, will be released that will be faster than the OpenSSL C code and it’s still smaller.
We can give you the performance you need and we want to help you!