It's sad to see that list of laptops all with low res screens.
That is the Intel influence.
I have acquaintances of acquaintances in upper management at Best Buy and various laptop manufacturing companies.
Intel gives half-off rebates for CPU/APU when paired with 1080p screens + etc.
Intel gives full-off rebates for CPU/APU when paired with 4K screens + etc.
Best Buy gets a rebate for simply selling more Intel-based machines than AMD.
The rebates never stopped. While they have gotten more and more complex in the ways they get achieved. So, they are practically not even rebates or classified as rebates.
---
This will effect Carrizo, immensely, because it is launching between Broadwell and Skylake. Unless, AMD upped the IPC to the point where we are looking at Netburst vs K8. Except, K8 is now super cheap and affordable over Netburst. There is no chance of Carrizo getting an affordable 1080p screen other than with the "FX" lineup.
It is all over the place, there is many issues. The AFS(Adaptive Frequency Scaling) is one of the few big ones. With AFS, Kaveri never runs at base clock speeds or boost clock speeds. Whatever is occuring in the Kaveri die. There is regular enough intervals of droops to cause the AFS system to malfunction.
====
Back to FDSOI
From LPS to:
2012 // 28-nm FDSOI Gen 1; 10% cost / 30% performance
2014 // 28-nm FDSOI Gen 2; <20% cost / >45% performance
2013 // 20-nm Planar; ~60% cost / ~40% performance
2014 // 14-nm FDSOI Gen 1; ~60% cost / >50% performance / 0.55x shrink
2015 // 14-nm FinFET; ~80% cost / >50% performance / 0.55x shrink
A second generation of 28nm FD-SOI, which implements source/drain engineering, delivers a 25% performance gain over the high-end 28nm high-K/metal gate (HKMG) bulk with slightly lower process wafer cost, or achieve the same performance as 20nm bulk but at the cost of 28nm HKMG.
28-nm UTBB FDSOI performance data is corrupted, only referencing Gen 1 FDSOI.
http://i.imgur.com/TBYwc5j.jpg
This makes more sense with the position of it now. It all makes sense now. The cost saving with 14nm FDSOI was saved but not the performance. I wonder why they want to hide 14-nm FDSOI going beyond 14-nm FinFETs.
Seems STM took somethings from olde AMD;
Continuous Technology Improvement
Introduction of improved transistor designs into high-
volume manufacturing on a quarterly basis
90-nm SOI -> 90-nm SOI + Strain Engineering -> 90-nm SOI + Strain Engineering + DSL -> 65-nm with everything