I too wonder about the doom and gloom we keep hearing. If tablets will go anywhere beyond their current "toy" status it will only be because they can do enough of the PC tasks o take over those roles. I mean, beyond just reading the paper or watching YouTube videos.
The Haswell development seems to be right on target, as far as I can tell.
Agree with shikantaza.
However, sales are sales, and most people use computers for ENTERTAINMENT or INFORMATION not for creating content, which PC's do every well. So tablets will win in volume...as most people have waited for years to get computer technology that only communicates, does not really "compute" and up until recently, their PHONES were not web communication tools.
The three C's of Communications, Control, and Computing are costly for systems that have to be optimized for all three C's. Focus on one C and its cheaper.
I have confidence Intel will continue to do well, and surmount the religious biases that tempt people to use ARM. Both have their places, and Intel is well-motivated to add features for fine-grain power control while supporting that ol' installed software base.
I'm troubled by predictions of gloom and doom based on tablets taking over the world. Not any time soon for serious users - until there is a substitute for a keyboard and high-resolution mouse, desktops and laptops will remain important after the tablet blip subsides.
Remember how small cell phones were for a while? A useful computer can only be made so small before it becomes an impediment.
I have a tablet; it's great for games and email. It's close to useless for anything else, no matter what Apple, Amazon, Microsoft, or Samsung say.
I saw a presentation at USC by the guys from Indiana(?) who did the on-chip regulator research. It's gall-danged efficient - well over 90% - and switches between capacitive and inductive energy storage to maintain peak efficiency across multiple load ranges. There are apparently multiple regulators on chip, so if you idle a processor, you can turn it's power supply off as well. Hard to do that size-efficiently off-chip.
Are we talking about Merrifield Atom in the new Haswell architecture in 22nm (Silvermont - 2014) and 14nm (2015)
Silvermont will appear in several flavours.
-- Merrifield will target SPs (Mot/Google, Lenovo, Nokia, Samsung, etc.
-- Bay Trail is targeted at Ts while Avoton will focus on microsevers
Rick, about which Atom processor are you and analysts commenting about?
-- In 2014 with 14nm Atom, with fully integrated Wi-Fi in digital domain (fully scalable) Intel might conduct a scorched earth market penetration
-- with many OEMs using it, including Samsung, Google/Motorola, Lenovo, Nokia, etc., etc.