This move could also help ARM manufacturers even the playing field with Intel. Most manufactures are still working at 28nm (some even at 45nm) using SOI (Silicon on Insulator) in comparison to Intel’s 22nm FinDET Trigate Atom CPU. Qualcomm is one of the exceptions with their own ARM SoCs showing off what can be done at 20nm.
Intel has reportedly delayed the mass production of their 14nm product which was originally scheduled to be ramping up for the end of 2014. Now some sources are claiming that 14nm will not be ready until end of Q1 2015 which is in line with the new dates that TSMC is claiming for their own 14/16nm products. It will make things very interesting in the mobile world starting in Q3 2015 as the latest products begin to hit the stores.
The people that are most likely to benefit from this move are Apple customers considering the fact that they are much more likely to move to the smaller process node in keeping with their plans for smaller, lighter and thinner devices. NVIDIA, on the other hand, might not jump at the new process simply because it has been their pattern to stick with the older and more mature nodes. That having been said, with the recent problems at 20nm, NVIDIA might decide to go with the newer process simply to avoid issues with their Tegra line up.
As of this writing TSMC is expected to be ready for volume production of 14/16nm products in Q1 2015 with Samsung right behind them. Intel is also expected to drop 14nm on the world around the same time frame. If both of these products hit around the same time we will be looking at a real fight between x86/64 and RISC for the first time since the mid-1980s. It will be interesting to see if x86 can win the fight like it did back then or if the more nimble RISC architecture and design from ARM will pull ahead…
Tell us what you think in our Forum