More about Bulldozer

Further on the new AMD Bulldozer architecture, one will see that it is actually made up of blocks of 2 cores. Each of these is called a 'Bulldozer module', and each shares an L2 cache, an FPU, and more. When looking at it from this perspective, this will help people to understand why it is most efficient to *try* to keep the workload on every other core, if possible. Not only for TurboCore, but also due to the intrinsic properties of the platform.