http://liliputing.com/2016/09/breaking-band-microsofts-fitness-tracker-may-do...



ZDNet’s Mary Jo Foley reports her sources say there won’t be a new Microsoft Band this year, and she’s “heard that Microsoft disbanded the group of individuals who were trying to get the Band to run Windows 10.” So it’s possible that we won’t see a major software update for current-gen hardware either.

The Band 2 has a list price of $250, but it’s currently on sale for $75 off. It features a curved AMOLED display, an optical heart rate sensor, motion sensors, ambient light, UV, galvanic skin response, and skin temperature sensors as well as a barometer and microphone.

When it went on sale at 3x the price with only half the features of existing ARM units, we predicted an early demise for this product.   Not early though, two years of being out of date and out of step later ......




Now we are waiting for this abortion to die "early" as well .....

https://www.engadget.com/2016/05/19/chromebooks-beat-mac-sales-in-early-2016/



According to research firm IDC, more Google Chromebooks were sold in the first quarter of 2016 than all of Apple's Mac line. The milestone marked the first time Google's Chrome OS moved more units than OS X in the United States.

Although IDC doesn't usually separate Chrome OS or Chromebooks from the PC category, the group did confirm the numbers to The Verge, saying "Chrome OS overtook Mac OS in the US in terms of shipments for the first time in 1Q16." The firm noted that Macs sold about 1.76 million units in the first quarter of 2016, meaning Chromebooks sold somewhere in the neighborhood of 2 million units or more. Overall, however, PC shipments are on the decline.

So, Chromebook sales just lapped Apple's entire Macbook line (all of it, counted together).  

Next stop, Windows PC space.

Going forward the Chromebook space will include the new Chrome Pixel phones, so Google will be swinging both mobile and laptop volumes together as they go after MS, which is "supposedly" swinging  all their laptop, PC and Win 10 phone volumes together, along with whatever old stuff they can load Win 10 on to that will accept it.


It will be a giant game of liar's poker, since MS likes them funny numbers so well .....

http://www.theregister.co.uk/2015/08/19/microsofts_walled_garden_is_no_pleasu...

Pirate MEP: Microsoft's walled garden is no consumer pleasure park
Redmond hijacks control over what users can and can't do with their own PCs



Microsoft is trying to create its own “walled garden”, much to the detriment of consumers, Pirate Party MEP Julia Reda has told El Reg.

Reda was referring to the new unified Microsoft services policy, which came into force at the beginning of this month and covers almost all of the tech goliath’s consumer services.

In it, Microsoft says: “We may automatically check your version of the software and download software updates or configuration changes, including those that prevent you from accessing the services or using unauthorised hardware peripheral devices.”

In other words, we’ll block any unapproved hardware or software as and when we feel like it.

“Users are now getting accustomed to an idea that – as in other consumer electronics like gaming consoles or smartphones – they do not have full control over what their computers can and cannot do. The great thing about general-purpose computers is that they can be programmed to perform any task we want them to. This feature is what makes computers a tool for empowerment,” she said.

 
http://liliputing.com/2016/09/leaked-intel-roadmap-shows-chips-2017-2018-coff...

A new Liar's Poker hand, this time from Intel

What is amazing about this thing is that just earlier this week Intel leaked a theory that they were going to 10nm and then to 7nm just like the phone boys were actually about doing right now, but taking several extra years too long to do so compared to the phone boys, yes, but going there all the same.

(despite the fact that 14nm showed Intel a decreasing effectivity wall in the sky that actualy made Intel's 14nm no faster than the old 22nm)



So, what this chart puppy actually shows is 14nm sticking around through 2018/2020.   This actually makes some practical sense as Intel has already converted all their 22nm production lines to 14nm lines and the chipsets produced these lines produce simply cannot process data any faster than the problematical 14nm stuff they are selling right now.

So, any practical speed ups shown in 2018 will ONLY APPLY to Microsoft Win 10 making hardware calls to the hardware accelerator functions that Intel and AMD are quietly building into their daughterboard assemblies.

Since this is all a big game of Liar's Poker from Intel's side of things, I would say this is the likely the real, true card and the earlier card thrown was smoky brown "unbridled optimism".

http://semiengineering.com/10nm-versus-7nm/

http://www.pcper.com/news/Processors/GlobalFoundries-Will-Allegedly-Skip-10nm...


Let's talk NANOMETER for a bit.

Global Foundries and IBM dropped a bombshell yesterday, they have Production Line #8 at Global Foundries up and running on the new IBM patented 7nm process (possibly using a non-silicone hybrid technology) for Global Foundries 7nm production.

Yes indeedy, they are ready to take orders and make production runs on that first 7nm line.           right .....

IBM and Global Foundries have been working very closely together as IBM sold their new technology production development plant to Global Foundries last year with a commitment to bring on 7nm as part of the sale.  They have done so at this point in time.

Where do the other players sit right now?   Qualcomm likes the Samsung 10nm process just fine right now because they have NO HEATING ISSUES and they like that, a lot.   Let somebody else go scout out them Indian uprising villages, Qualcomm has gotten enough arrows in the back lately and wants no more embarrassing "technical leadership issues", ever.   Qualcomm also sez "Let Global and IBM produce 7nm for a full season, then we will talk changeover".

Samsung has their own take on 7nm in progress going at the moment, but is quite happy with selling their "known world-class winner" 10nm set up to the world right now.   Remember, Samsung is still part of the IBM/Global/Samsung technology group, so they got access to whatever Global is building for 7nm right now.    Plus, Samsung is all caught up with their exploding batteries right now and Samsung Mgmt. is somewhat risk adverse at the moment.

TSMC is sitting at 10nm on four of their own lines and is looking to license somebody's proven 7nm technology to be starting their up 7nm changeover efforts up in 2017 (after the new EUV processes prove out to be reliable and stable).   TSMC licenses only known winner technology, TSMC takes no risks on the new stuff and are content to be part of a year late to the party.  

TSMC and Samsung currently own the Apple A-11 SOC generation production contracts for the next two years, BTW ..... all at 10nm lithography.

Intel is completely lost in their own massive clouds of brown butt vapor, so nobody knows really where they are right now except that "always 1-2 years late to the party" is currently a pretty good general Intel planning statement as of late.  

"Non-competitive" is another, harsher way of saying it about Intel.



Intel is choking and sputtering .....       Intel is now saying their own "limited 10nm capability will be put out for rent to the highest bidder", intending to find some overeager somebody out there in ARM land who will partner with them to help (pay for) finishing out developing a 10nm production process "the Intel way".    

NOTE PLEASE:   Intel will NOT use their newest 10nm process on their own bread and butter PC CPUs
(that sounds pretty bad, Intel --- low yields and overheating issues again ???).

So far Intel has no takers ......    (people remember Rockchip and Altera entirely too well)

Intel lovers say Intel is totally working on non-silicon and is putting absolutely zero dollars down into silicon land ever again .....  This might be a smart move for a strongly shrinking company like Intel.   Hire you a few new non-silicon gurus away from IBM and ditch them thousands of home grown silicon experts that you already have in house.  Shrink some more, that's the ticket.    

You can save billions RIGHT NOW by cutting all them useless silicon dead heads, you know.    

Jest circle them wagons with your contractual buddy Microsoft, and don't do anything for the next year or so.             Trust good 'ol Microsoft to carry you through to safely.


                           

Part of what causes this confusing mixed bag on future progress is that Apple isn't up in front cracking the whip any more.   Apple made this choice when they sold ARM Holdings to the Japanese Softbank Consortium as Apple now sees all the major players (all the big guys is now including Huawei and Mediatek) as designing their own chipsets now with only a few of the smaller guys actually buying the box stock ARM designs any more.

Apple used to define the state of the art each new year, but that era ended this past year when Apple ceeded their old tech crown to Samsung and Qualcomm.

So you got Intel and Apple dropping back down into the pack (Intel going to the very back grouping of the "slow pack", BTW)  with IBM and Global taking the tech leadership role for a while.   Samsung/Qualcomm is still wearing the current tech crown for the upcoming year off of last year's joint heat pipe 10nm initiatives anyway.

Pundits are now saying 7nm may require a total product redesign of your phone product as 2.2 volts is all that these chipsets will need as they may durn well not even be silicon based any more .....

Initial designs at 7nm may well still be held to 5 volts as a change-over stage, mainly because all the sub suppliers still need to develop reliable 2.2 volt product offerings for batteries and radios etc. etc. etc. and ramp these offerings up into significant production before 2.2 volt component costs can come down.

Big picture time, 20nm is now the low cost leader as it requires no FinFet and requires no multiple pass lithography runs, 14nm is now mostly a "has been intermediate node" and all the latecomers to current high tech progress are jumping directly down to 10nm directly as they can only afford to redesign one (1) lithography move right now and they feel they need to keep 5 volt tech right now as well.

We have a global recession looming in the next 3 years, and all the tech companies are making their plans accordingly.

Look for ARM Holdings to evolve rapidly during this dead time now that Apple's heavy thumb isn't holding them back any more.

The 10nm and 7 nm references are conductor and input/output connector sizes going into transistors, right? Board traces , etc,,

Did you guys catch this little nugget in the text above?

"Somewhat extraordinarily, due to incredibly tight stacking (30nm transistor pitch), IBM claims a surface area reduction of "close to 50 percent" over today's 10nm processes. All told, IBM and its partners are targeting "at least a 50 percent power/performance improvement for the next generation of systems"—that is, moving from 10nm down to 7nm. The difference over 14nm, which is the current state of the art for commercially shipping products, will be even more pronounced."

These are significant improvements, improvements that would move you straight to the front of the pack it would.    Both Samsung and Qualcomm are kinda sneaky sometimes and both of them do like to do a "world shaking product announcement" especially if they can get a real, "pay me back 3x overall" competitive advantage out of it .....

Sammy and Qualcomm both may be trying to play the other one right now, in other words.  Or else perhaps they need to go forward with another agreement for Samsung to make chipsets for both of them at the new 7nm process node.


This puts Intel really and truly way... way... way... way... behind the industry leaders, you know that, right?