Why a single 6 core instead of dual 6 cores or dual quads? Seems like that'd be better, at least from a number-crunching point of view.
My workflow is not CPU bound. I rarely pegged four cores for any length of time with my prior workstation, mainly because the bus, I/O channel and disk drives are always the primary bottleneck in my work. Slow I/O causes frequent latent periods where the CPU does nothing while waiting for I/O to unblock. Reading, encoding and then writing hours of video content back to disk is an extremely I/O bound process. This is why I invested heavily in high-capacity, fast storage.
The next biggest bottleneck is memory, as in not enough to load all desired applications at once, each having their own appropriate caching in place for ideal performance. Invariably, things quickly begin to swap to slow scratch storage. Therefore, I again invested heavily in high-density, high-speed DRAM, leaving open slots to double down again if necessary.
When editing video, the third worst bottleneck is the CPU, as the encoding step does take advantage of all available cores. With the upcoming increase in storage and memory performance, there should also be fewer latent periods, but I don't think encoding will peg all 12 virtual cores. To hedge my bet, I did invest in the fastest 6-core processor available, and I am willing to take advantage of its unlocked crystal to over-clock a bit, but only if necessary.
When playing games, these bottlenecks change in importance rather radically. The GPU suddenly becomes the most critical component to maximize frame rates at high resolution pixel densities. Very few games require more than two CPU cores during game play, so the CPU is a distant second or third-place bottleneck when gaming.
It's a calculated risk, but instead of funding a more expensive 8 or 16-core Xeon platform, I decided to invest more heavily in faster storage, greater memory capacity and strong gaming GPU hardware.
