Source: Tom's hardware UK – Keywords: overclocking, intel
Categories: Hardware
An Overview of the Core 2 Steppings
So far, Intel has implemented several steppings for its Core 2 processors. When Intel first introduced the Conroe core, it started out using the steppings B0 through B2, with B2 being the most common. B0 and B1 versions, on the other hand, were usually only used in early samples or sent out as test samples to hardware developers (such as motherboard companies).
Among overclockers, the B2 stepping is the clear favorite. In the majority of cases, it offers the greatest overclocking potential while allowing stable operation.
Intel’s Core 2 Duo line of CPUs featuring the B2 stepping is nowhere near as energy efficient as AMD’s Athlon 64 processors. As our measurements showed, a Core 2 Duo E6300 (B2) draws about 12 Watts when idling, with Speedstep enabled. AMD’s smallest dual core processor on the other hand, the Athlon 63 X2 3800+ EE only requires 6.14 Watts when running at its slowest, thanks to its Cool & Quiet functionality.
To lower the Core 2’s power requirements with the processor idle, the core received an update called the L2 stepping. The tweak worked, as this processor only draws 8.62 Watts when idle. However, the processor is unpopular with buyers wishing to overclock their processors, since the stable frequencies that can be attained with this version are much lower than with the B2 versions.
We were able to push our Core 2 Duo (B2 stepping) as high as 3 GHz without having to raise the core voltage. By comparison, our L2 version of the same processor only barely reached 2.8 GHz.
The quad-core processors we that we overclocked basically consisted of two full Core 2 Duo models with the B3 stepping, each with 4MB of L2 cache. Interestingly, there have been no dual-core processors with this stepping so far. Yields for a quad-core processor are never 100%. It follows, then, that a certain number of CPU cores from each production batch are likely to have L2 caches that are only partly functional.
Since Intel only sells quad-core processors with 8MB of L2 cache, this begs the question, what does the company do with the remainder of these cores? It’s entirely possible that there is no technological difference between the B2 and the B3 steppings, and that Intel chose this designation simply because there are two dies in one package. This would also explain why the overclocking results are identical to those of the dual-core models with the B2 stepping.
- Previous page The New Core 2 Models with FSB 1333
- Next page An Overclocking Powerhouse: The new...
- Do More Cores Beat More Clock Speed?
- Extreme FSB: Taking the E6750 Beyond 4 GHz
- Tom's Hardware's 2007 CPU Charts
- AMD's Smart Strike: Athlon X2 BE-2350
- Energy Efficiency Duel: Intel Left Out In The Cold
- Which is the Best Mainstream CPU?
- The Gigahertz Battle: How Do Today's CPUs Stack Up?
- Dual Core Notebook CPUs Explored
- Does AMD's Athlon 64 X2 6000+ Have Any Kick Left?
- Quad-Core Xeon Clovertown Rolls Into DP Servers


I hope Intel doesn't start 2 get a monopoly...
You're probably right to worry, but I read that AMD has taken a big upswing this month... hopefully it'll sort itself out.

I even fear the game monopoly