[curves] Comparing high-speed / high-security curve implementations

Samuel Neves sneves at dei.uc.pt
Wed Apr 23 13:45:01 PDT 2014


On 04/23/2014 09:05 PM, Trevor Perrin wrote:
> Thanks, do you have Sandy Bridge or Haswell numbers, since that's what
> I have for others?
Quick and dirty measurements give me 139K Haswell cycles and 154K Sandy Bridge cycles. I measured the djb variant, which
was the fastest constant-time version reported in the paper.


More information about the Curves mailing list