Hi,

Many of our users are curious about how the CyaSSL embedded SSL library will perform on a specific hardware device or in a specific environment.  Because of the wide variety of different platforms and compilers used today in embedded, enterprise, and cloud-based environments, it is hard for us to give generic performance calculations across the board.

To help our users and customers in determining SSL performance for CyaSSL / CTaoCrypt, we provide a benchmark application which is bundled with CyaSSL.  CyaSSL uses the CTaoCrypt cryptography library for all crypto operations by default.  Because the underlying crypto is a very performance-critical aspect of SSL/TLS, our benchmark application runs performance tests on CTaoCrypt’s algorithms. 

The benchmark utility is located in ctaocrypt/benchmark.  Typical output may look like:

./ctaocrypt/benchmark/benchmark
AES      5 megs took 0.033 seconds, 149.84 MB/s
ARC4     5 megs took 0.017 seconds, 297.23 MB/s
HC128    5 megs took 0.004 seconds, 1185.10 MB/s
RABBIT   5 megs took 0.011 seconds, 453.14 MB/s
3DES     5 megs took 0.236 seconds,  21.18 MB/s

MD5      5 megs took 0.011 seconds, 444.13 MB/s
SHA      5 megs took 0.019 seconds, 261.73 MB/s
SHA-256  5 megs took 0.041 seconds, 122.98 MB/s
SHA-512  5 megs took 0.023 seconds, 213.44 MB/s

RSA 2048 encryption took   0.11 milliseconds, avg over 100 iterations
RSA 2048 decryption took   2.36 milliseconds, avg over 100 iterations
DH  2048 key generation    0.88 milliseconds, avg over 100 iterations
DH  2048 key agreement     0.92 milliseconds, avg over 100 iterations

This is especially useful for comparing the public key speed before and after changing the math library. You can test the results using the normal math library (./configure), the fastmath library (./configure --enable-fastmath), and the fasthugemath library (./configure --enable-fasthugemath).

As always, please contact us with any questions or comments.

Thanks,
Team yaSSL