The architecture is different. Conroe shares it's level 2 cache with each core, while the Presler has two independent l2 caches that communicate with each other using the memory controller and the system bus thus increasing latency.
Conroe also processes 128-bit SSE instructions in a single clock cycle, while Presler can only process half every clock cycle.
Conroe is also able to process four instructions per clock cycle. Presler does only three.
That's it in a nutshell. There is much more to the Conroe then I could sum up in a few words, so I hope this helps.