8th August 2017 - In the absence of any real competition in the desktop/workstation CPU space, Intel has been been strolling along for the last 10 years at its own incremental pace, increasing cores but not really increasing (well actually sometimes decreasing) clock speed. Might be good for rendering, but large constraint network simulations, like cloth, benefit from higher clock speeds.
However finally in 2017 AMD has come out with Ryzen, a CPU competitive enough to lite a fire under Intel.
Ryzen is a desktop part with a lot of cores and so Intel has been forced to respond.
As a result Intel has had to step up its game and release higher clock speed, higher core desktop CPUs.
Welcome to the i9-79xx range and we feel the 14 core i9-7940 is probably the best # cores vs Turbo Boost trade off - see turbo chart below.
We'll be benchmarking this CPU later in the year and with an all core turbo of 3.8GHz we are expecting a 20+% performance increase over the current 14 core E5 v3 and v4 (all core turbo of 3.1/3.2GHz). As these are desktop CPUs there is also plenty of scope for an easy 10% overclock to add a bit of spice. Finally these chips have a new matrix cache architecture and with some of the L3 cache moving to L2 both in the desktop and workstation parts. It will be interesting to see if this is beneficial to typical Carbon simulations.
On paper these desktop CPUs look ideal for workstation use and it will be interesting to see the knock on effect on the E5 series v5 CPUs usage, which will be at slightly lower clock speeds, and much more expensive as usual.