Originally posted by: rogue1979
I'd have to disagree with a 15% increase across the board. I have both a t-bred and a Barton both running in the 2200-2300MHz range. The difference across the board is more like 3% at the same clock speed. While you could find a benchmark that benefits more from the Bartons extra L2 cache, there are also a few that don't show anything. 3Dmark shows the same 3% increase, gaming doesn't show a big increase. My advice to anyone out there is to stick with the t-bred. Until the recent mobile versions of the Barton, the average overclock was 2200MHz at best. So a 2300MHz t-bred is gonna be about 1-2% faster than a 2200MHz Barton because of the 5% increase in clock speed.
Originally posted by: SickBeast
Originally posted by: rogue1979
I'd have to disagree with a 15% increase across the board. I have both a t-bred and a Barton both running in the 2200-2300MHz range. The difference across the board is more like 3% at the same clock speed. While you could find a benchmark that benefits more from the Bartons extra L2 cache, there are also a few that don't show anything. 3Dmark shows the same 3% increase, gaming doesn't show a big increase. My advice to anyone out there is to stick with the t-bred. Until the recent mobile versions of the Barton, the average overclock was 2200MHz at best. So a 2300MHz t-bred is gonna be about 1-2% faster than a 2200MHz Barton because of the 5% increase in clock speed.
I got the 15% figure because he is currently running at 2300mhz, and could potentially be running at 2600mhz with the additional 256kb of cache. You're right, as identical clock speeds, the barton is only marginally faster than the tbred. Seriously look at the gaming benches tho, Quake3 is a good example. The cache makes a substantial difference in those cases.