While Burn Test is useful as a quick preliminary "Easy Button" stability check, it is unsuitable for long term CPU stability testing or thermal testing.
As the author of the Core 2 Quad and Duo Temperature Guide - http://www.tomshardware.com/forum/221745-29-core-quad-temperature-guide
- here at Tom's, allow me to explain why Prime95 Small FFT's is the standard for stability and thermal testing in the overclocking community.
Prime95 Small FFT's maintains the steady-state 100% workload required for consistent CPU testing, and does not oscillate between high and low workloads, nor does it load all registers with all one's, which produces unrealistically high temperatures @ 115% workload.
You can get a similar explanation from the author of Real Temp, Kevin Glynn, (unclewebb
) over at XtremeSystems - http://www.xtremesystems.org/forums/showthread.php?t=179044&page=113
- where there's a 113 page thread detailing the ongoing development of Real Temp, much of which revolves around CPU load testing.
The proof is in observing CPU and Core thermal signatures, which react directly to workload, as shown below using SpeedFan "Charts".
Prime95 Small FFT's:
I hope this helps you to understand the differences, and why Prime95 Small FFT's remains the
CPU testing standard.