And Intel shows to the world they have sort of a new sort of issue with their new smaller lithography foundry processes. How can having upsized your wafer to the size of an old LP record be a bad thing? Well, it costs more money to make the big 'un and as you go below 20 nm there are an
awful lot of little chips on that huge wafer.
This would be good if they were all 100% usable chipsets -- but as everyone has discovered going below 20 nm costs you in having to add 3-D features to every bend in the trace and that means significantly more scrap and lower yields as 3-D lithography is difficult to do.
And Intel has found themselves a brand new issue with going below 20 nm and that is having enough chips on order to even fill up those big wafers.See the empty stripes?
Look to see Intel to try to fill up those stripes by offering to
build somebody else's ARM chipsets for "good long term customers" like Apple.
Intel is arguably the worlds best at the manufacturing processes that build chipsets. They are NOT the world's best at designing them though, not in the mobile world anyway.
And when they get to the point they have to PARTIALLY IDLE their production lines due to no orders, then it is time for Intel to consider new avenues to fill up their production lines again.
===============================
Intel is currently pushing some old mainline chipset designs that have been die shrunk by 2x or better. They have mated these energy wasting designs with an energy nanny chipset that fakes them being low power consumption chipsets (playing off the old Antutu testing methodology).
They are currently producing these monsters at a 15-20% loss right now, but are still trying to get into enough phone and tablet products to get their assembly lines filled back up again.
Adding the new manufacturing losses of the new smaller lithography processes to the current 15% loss leader pricing of the tablet chipsets they are making now, how will Intel be able to fuel the move into hand phone chipsets (on the map for later this year)?
It will cost them MORE to go there than it does right now.
Intel has gotten some fiscal downchecks from Wall Street lately, and we are beginning to see why.
http://www.fudzilla.com/home/item/33688-intel-14nm-transition-in-troubleBy the nature of their manufacturing processes, they will be running at less than 50% of capacity after the next lithography shift if they don't at least DOUBLE or TRIPLE the size of their phone/tablet business.
To do this, they
must go down price-wise, all the way down into the Allwinner/Mediatek neighborhood. And do it while making a profit margin that can support their business.
Big big challenges to folks with no real mobile chipset designs yet and who are producing tablet chipsets at a 15-20% loss right now .....
..... for Intel to survive long term, a lot of the little guys are going to have to die. And they will NOT die willingly.
Intel is likely heading for some form of "dumping" charges tried in an oriental court room under some very biased Chinese home laws before very long.
If Intel gets legally blocked from shipping those loss leader chipsets to the Oriental plants that make the tablets and the phones, then they won't sell much product nor will they make any amount of real new money.If Intel can move their ego out of the way there are pathways to success that exist for them. Pick a state of the art ARM chipset design that has full Mali/Linaro/VR support and contract with Samsung to product it for Samsung cheaper than Samsung can make it. Or do the same thing with Apple.
Let Samsung/Apple's volume fill up your slack for a year or so, then concentrate during that time on adapting the ARM stuff into an Intel X36 compatible design that really IS energy efficient and low cost.
Intel is going to survive -- how smart they are about it remains to be seen.