Seti time fluctuations - prescott issue?

I have two prescott systems - not sure if this is limited to prescotts though...

One is a PIV3.0E in an IC7 which currently runs at 1:1 at 220fsb (3.33GHz). For two units I initially needed less than 2 hours (typically 1h:55min) - now alll of a sudden I am up to 2:48 or so for the same workload (no system changes).

Similar my Dell 9100 (2.8E) desktop replacement used to produce 2 units in 2h:24min (Dell=stock 200MHz)- now it's up to about 3:14.

Is this a new prescott phenomenon or are there possibly other reasons?

Could ambient temperature play a role (do prescotts or PIVs throttle performance when it's hot)?

Thanks

Yo

Whatcha got for cooling?

Koolance (water) for the destop, and it's not running that hot - the chip is capabble of 3.76GHz but now runs at 3.33

Air for the Dell and it gets warm but no crashes ever on either machine