Carrizo-L For Christmas, Full Carrizo For The New Year

The Crazed Child · October 9, 2014

Great process for the GPU portion, not so much for high CPU clock speeds.

We saw exactly that with Kaveri. Even the small IPC boost wasn't enough to counteract the lower stock clocks compared to Richland.

LukaP · October 9, 2014

This is all going out the window. AMD and Intel are both designing heterogeneous chips which use system memory, and you can only keep the engines fed if you never run out of fuel, but you need to be able to do rapid-fire small chunks too (thank you caching).

And about 2666 RAM, this is why we need to fight for lower CAS latencies. If you increase CL by 2 you lose bandwidth and access speed.

mordorson · October 10, 2014

Zen is supposedly 16nm FinFET, but even that is unconfirmed.

In the mean time, Intel are shipping 14nm, with 10nm going into large scale production by the time Zen roles around

Zen is actually supposed to be 14nm, 16nm is designed more for GPUs.