The 2.5+GHz A9 is interesting, not because it is on 28nm, but because of the speed it operates at. This A9 is built on the 28nm HPP process, and is aimed at wired application, not wireless, so it sounds like leakage isn’t exactly the highest priority on this SoC. In light of the high clocks, that is expected, but what wasn’t is the .85v operating voltage. In any case, should a manufacturer want to burn the wattage, doubling the frequency of current A9 designs should be quite achievable.
If you are worried about running out of headroom on 28nm, the second announcement, an A9 taped out on 20nm, should help you sleep better. It may not amount to much as a product, but it is still a Technical Qualification Vehicle (TQV), but it will shake out a lot of bugs. The GloFo 20nm process is a long way off. There should be plug and play ARM Ax designs ready for GloFo’s new process from day 1.
In the end, we don’t have anything you can run out and buy, just moving the stakes in the sand a bit farther. A9s get what appears to be a speed record, GloFo gets a halo, and 20nm inches ever closer. This is how progress is made.S|A
Latest posts by Charlie Demerjian (see all)
- More on Intel’s 10nm process problems - Sep 17, 2018
- Intel puts out another 14nm 2020 server platform - Sep 11, 2018
- Why Can’t Intel Supply Enough 14nm Xeons? - Sep 10, 2018
- Intel can’t supply 14nm Xeons, HPE directly recommends AMD Epyc - Sep 7, 2018
- AMD reintroduces the Athlon name with two CPUs - Sep 6, 2018