You may have missed the notice that Microsoft now includes the "XPERF" toolkit within both the WDK as well as the Platform SDK.   Previously used only internally within Microsoft, the tools are now finally available to everyone.   An example of the power of these tools...   In less than 5 minutes, I witnessed a performance engineer identify a single node within a cluster curiously running at 66% CPU utilization but showing no Task Manager entry consuming more than 1 percent CPU utilization.   XPERF enabled the engineer to quickly identify the unclassified CPU utilization culprit as the OS kernel process and as a specific DPC responding to a persistent USB device interrupt.   A very impressive scenario when you consider the task of tracking down a problem source-point amongst many compute nodes each contributing potentially negative characteristics to the overall system performance profile.  

Learn more about Windows Performance Toolkit and how to use it here and use the MSDN forums to see how others are using these tools.