Hmmmm.....my 920 batch 3837B317 couldn't even reach a stable 4GHz by any combination of BCLK or voltage manipulations (I went as high as 1.48ish vcore) that I've tried on my ASUS P6T6 WS Revolution. I've settled for 3.8GHz (19*200) as I can get that prime stable, but that requires a vcore of 1.4v. So there must be significant heterogeneity within batches as far as overclocking potential (or I just suck wind at overclocking, or both).