Ahmad Raif Mohamed Noor Beg wrote:
If we are talking about a PC which uses x86 hardware (Intel, AMD etc), yes with the Gigahertz speed, using software will be faster than using hw accelerator, in this case FPGA but the original question was I believe usage in an embedded environment and using ARM processor. Now using a hw accelerator should make sense - right ?
k, then I'd consider designing it so it has sufficient buffer memory to process several messages entirely in its own space, and memory map that buffer memory into the host (ARM) address space. the host drops messages into it, along with instructions for what to do, and the FPGA picks these up and runs with it, encrypting or decrypting the messages into the same buffer memory (probably into a different address). this way the host overhead is minimal (copy message(s) to the buffer, then copy results back)
______________________________________________________________________ OpenSSL Project http://www.openssl.org User Support Mailing List openssl-users@openssl.org Automated List Manager majord...@openssl.org
