On the Internet Super Hero blog today, there's this new post with a sort of retraction of some of the previous comments they've made on the mysqlnd drivers - really more proof that "mysqlnd can't do wonders".

Lastly, I've been blogging about positive performance related feedback on mysqlnd. Whenever I did so, I got some positive feedback - for obvious reasons: I was posting good marketing news. [...] As much as I'd love to, I can't stand behind a simple statement like "mysqlnd makes your applications faster". You must not assume that each and every application will profit from mysqlnd.

His target of choice was WordPress which he ran through a testing cycle with JMeter (a load and performance testing tool). In the end, though, using the mysqlnd drivers and the regular drivers didn't show much of a difference at all. To figure out why, he uses XDebug, APC and KCachegrind to work through where the issues might lie. Unfortunately:

I was not able to see any mysql_*-function in the call graphs. The reason is that the mysql_*-functions do not contribute to the overall runtime of this application in a major way.