[Zlib-devel] patch-in-progress: vectorized adler32 calculation

[John Bowler]

From: Stefan Fuhrmann
>Throughput:
>SSSE3: ~3.5 bytes / clock tick
>SSE2:  ~3 bytes / clock tick
>MMX:  ~1.5 bytes / clock tick
>C-Code: ~1 byte / clock tick
>
>IOW, for x86 processors released since 2003 (SSE2), the checksum portion
>is reduced from ~15% to ~5% or better of the inflate runtime.

Did you measure that last figure?  CPU clock tick calculations rarely translate into time figures in practical applications when the data is in large buffers, because sooner or later the CPU stalls waiting for data.  

John Bowler <jbowler at acm.org>