Everyone at Intel should watch this
https://www.youtube.com/watch?v=JpgV6rCn5-g
5 replies (most recent on top)
Volumes for IoT and memory can be huge, but the margin and profits slim compared to the fat margins Intel and its fast executives expect. It is laughable that anyone can believe that IoT and Memory can somehow help intels bottom line grow. Sure if the are willing to invest billions in the fabs and accept the small margins they can get volume, but with Intel's cost structure and mentality they will lose money worse than they did in tablets with their contra revenue, LOL.
The thing that is scary is that the board so far is going alone with it, could it be there is nothing else. Dang be happy for the ISP/VSP/ERP they are the lucky ones that got of the Titanic, LOL
Low margin IoT and commodity memory are not going to replace the lost PC unit volumes.
Here's a link to the slides:
http://www.hotchips.org/wp-content/uploads/hc_archives/hc25/HC25.15-keynote1-Chipdesign-epub/HC25.26.190-Keynote1-ChipDesignGame-Colwell-DARPA.pdf
I wish Bob was still at Intel.
Another great read is his book - "The Pentium Chronicles".
Required reading for engineering managers - I doubt the people that are left at Intel have even heard of it.
TSMC and Samsung are funded by Apple.
They have the money to invest in leading edge since Apple's customers drive the volume.
Apple is critically dependent on them, so therefore they will succeed or no one will succeed.
Everyone buys the same equipment.
Money can attract talent as well.
Interesting talk and some key takeaways
1) did the silicon technology matter? Intel claims to be ahead on process, but AMD was competitive for years on an inferior technology and Apple / Samsung produced amazing Aps processor that proved Intel with more advance process couldn't crack . Conclusion Intel spending billions on R&D and even more billions on manufacturing was wasting money. Or put it another way as Moore's Law end intel needs to completely transform and as such the model of more Moore on 3DNAND and IoT isn't the right strategy.
2) Technology doesn't matter, what matters is selling something the consumer wants to buy. Intel has been a epic fail. We've spent billions and billions on next generation of CPU yet total volume of PCs continue to fall, yet cell phone and IoT volume are growing and beyond the Aps processor nobody cares what node it is on. The solution and home run isn't in the silicon technology node.
Thus 1) and 2) says beyond the x86 monopoly intel really has nothing else... Moore's law was very convenient during the MegHz race from 25MegHz to about 3 years ago, but now nobody cares what frequency, how much memory is in a phone except how much video and pictures I can save,
Bad times indeed for inte, LOL
"Physics doesn't care what you believe in"
"Physics is the same in Taiwan and Hillsboro"
Believe it or not TSMC and Samsung have a greater urgency and need to deliver and larger volumes!