Elwin Stelzer Eliazer wrote:
After your email below, i tried "-multi" and also "time". I also
loaded the CPU and took some measurements.
But to my surprise, i am getting the same performance numbers, for
"time ./openssl speed dsa", in all the three following cases.
Are you sure the engine offloads DSA ? Cavium site talks about RSA, DH,
AES, 3DES, RC4, MD5, SHA1, but says nothing about DSA.
The -multi parameter specifies the number of processes the benchmark
will start the "100000" value you tried seems very high. Start with 2
and multiply by 2 until the performance numbers stabilize. The optimal
number might depend on the key size. Also you can make your tests faster
by specifying only the key size you're interested in, for instance "time
./openssl speed rsa1024" will only test 1024 bit (whereas rsa tests for
512, 1024, 2048 and 4096).
Lastly the performance numbers you will get with hardware offloading
will be skewed if you don't add the "-elapsed" flag. Otherwhise the
numbers use the CPU time to make speed calculation. When only using
software CPU time is more or less equal to elapsed time. That's not the
case with hardware offloading (and that's the point :P).
Sincerely,
--
Mathias Brossard
______________________________________________________________________
OpenSSL Project http://www.openssl.org
User Support Mailing List openssl-users@openssl.org
Automated List Manager [EMAIL PROTECTED]
