Closer Look at AMD FX-8350
For our tests we received AMD’s most interesting product – the top CPU in the refreshed FX family with Vishera design – FX-8350. You can check out its formal characteristics in the following CPU-Z screenshot:
I would like to remind you that this processor consists of four dual-core modules. Each module has its own shared 2 MB of L2 cache-memory, and the entire processor has a shared 8 MB L3 cache. Like in their predecessors, the L3 cache and the dual-channel memory controller in Vishera based processors belongs to the North Bridge, which works at its own frequency. For FX-8350 it is set at 2.2 GHz.
The nominal frequency for the FX-8350 is set at 4.0 GHz, but the processor supports Turbo Core technology, which can automatically overclock it to 4.2 GHz. This technology has become much more aggressive in the new processors than it was before. Now they can hit the maximum frequency even if all the computing cores are utilized. In other words, Turbo Core can now boost the performance not only in low-threaded environments, but also during maximum thread count.
I would like to draw your attention to another interesting thing. FX-8350 processor has OR-C0 stepping, while the previous generation mass production processors used OR-B2 stepping. It is a clear indication that AMD views their new processor design not as an individual product, but merely as another revision of the original Orochi core.