Intel ULP SoCs

After their latest report, I heard Intel said they will focus more on cloud, sell Xeons there.

Less so on mobile.

 

Ouch. There goes my dream of having a single computing device (that Win10 phone capable of running legacy x86 apps as a laptop).

 

This is probably happening as a Surface Phone down the road. It's just not happening in 2016. Maybe it'll be somewhere in 2017 when Intel's 5G Atom comes out.

Also, they're most certainly keeping their Atom X5/X7 tablet line. They're moving tons of those chips nowadays in hybrid Win10/Android tablets, not to mention Surface 3.


SoFIA was a terrible competitor to Mediatek's MT675x and MT673. Broxton probably didn't support LPDDR4 and its GPU must have been something mediocre too. It wasn't going anywhere when faced with e.g. Snapdragon 650 and up.

 

Why would a Surface Phone have to be x86? Isn't UWP supposed to support ARM executable too?

 

You going to run desktop Office on a phone?

 

Why isn't Windows 10 Mobile a huge success instead of an utter failure, then? This is quite the mystery...
Why was Windows RT another utter failure and the x86 Surface models huge successes?

Anything from Microsoft for the consumer market that was launched without windows x86 legacy has just been burning money.
Thinking this will change by miracle is naive IMO.

People are going to connect their phone to an external monitor, keyboard and mouse and make it their personal computer that carries around in their pocket.
Like windows Continuum, but made in a successful way (i.e. using hardware capable of running legacy software).

 

Hmm, I guess they could but will it be like running a $400 laptop or a $1000 laptop with Core i7?

 

IMO what is naive is believing the majority of consumers will abandon iOS or Android legacy to run Office on their phone, be it connected to a screen and a monitor. Do you think putting x86 on a phone will magically make phone-friendly apps available by some magic?

For pros, of course, that'd be a different story. And perhaps that's what MS will target. We'll see.

 

For most people's computer needs, Android and iOS are convenient appendages and not the critical device they need.
Be it professionally or personally, whenever a person needs to do something that requires focus and/or productivity (writing an extensive e-mail, redacting a document, editing videos, editing photos, organizing portfolios, etc.), they will always resort to a Personal Computer, which inherently means legacy-capable windows.

If people could turn the convenient appendage into the real thing, then they would shift their purchase intentions in a heartbeat.

 

We'll never know now though. Game over.

That's not to say there aren't attractive aspects to a phone that can be converted to do more robust personal computing tasks, requiring other input and output options. The computing resources are there. The only done deal is that it won't be x86 based.

 

That's not what I read from the news.
Intel cancelled Broxton because it probably wasn't competitive with Snapdragon 820 (probably not even S650), so they scrapped it in order to focus on 5G SoCs.

 

Mate you need to get out more, Most people use a PC/phone/tablet to access the internet, that is it, the end, case closed. This isn't 1995 anymore

 

What does this post have to do with anything within the post you quoted?

 

PC sales are plummeting because people are fine with their older PCs.

Or they do more with phones, have no problems doing things that might be better with bigger displays like watching a lot of video.

Sure for Office/office work, you need a full desktop or laptop at work.

But for personal use, most people are consuming content so mobile devices are "good enough" to more and more people. I can't see watching longer video content like TV shows or movies yet more and more people, especially the younger generation who are growing up with phones from the time they're in elementary school, are doing it.

For Intel, failure to get their chips into phones may have ended their growth potential.

 

?
You're stating android/IOS/windows mobile don't cut it for the most people, and they 'need' a PC, but this is not true
as wco81 saiz PC sales have been dropping quarter after quarter, yes partly due to PC's not getting much better than what they were a few years ago but also cause for most people they are finding they can function perfectly fine on an android/IOS device

btw saw this a few days ago on arstechnica, the fanless ARM A9X chip is pretty decent against the more expensive (fanned in the skylake's case) intel chip, esp in GFX benchmark. Like I've said I wouldn't be surprised if apple starts putting ARM chips in its laptops.
Multicore

 

I think you are making a mistake in assuming that intel seeing 5G as a potential growth area is the same as them returning to making x86 mobile SoCs. They've spent a decade and lost roughly ten billion USD trying to make x86 competitive in that market and failed. The text at the end is quite explicit that they focus on servers and communications (while cutting all low end client SoC development).
They are focusing on growth in areas where they can at least hope to make a worthwhile profit.

 

PC is dropping year after year.
IOS tablets are dropping year after year.
Android tablets are now dropping.

The market that wants tablets for consumption only is basically saturated now, similar to the PC market. However...

x86 Windows tablets are growing as well as x86 gaming PCs.

The market for content creation tablets is now a growth market and has been for the past couple of years. It's still a small fraction of the overall tablet market, so we'll have to see what the ceiling is for it.

iPad Pro did well, it'll be interesting to see if it continues to do well, but it certainly won't stem the fall of iOS tablet devices. It addresses the content creation market, but applications for it are extremely immature at the moment and only address the art creation market, atm. Many people bought it based on its promise, but many have also ended up continuing to do content creation on their Windows devices (specifically Surface Pros) while they hope applications for iPad Pro mature.

Regards,
SB

 

Exactly. The problem is not hardware capabilities per se, but software availability. (It's worth noting though that the iPad Pro vastly outsold the MacBook with both units having similar hardware.)
It is, as always for a new market, a bit of a chicken and egg problem. Still, the iPad Pro models have been well received, so the installed base will be counted in tens of millions this year, and Apple has demonstrated that they put their weight behind the concept. It seems like a decently safe market to enter for software developers. Will still take years to develop though, so a pioneering spirit could pay good dividends.

 

One of the major complaints I've continued to see with regards to the iPad Pro when it comes to content creation is content management and easy of interoperability. IE - they want a real file systems that is user accessible, access to user controlled external media, and ease of exporting and importing files (to and from Windows devices) without having to resort to cloud storage.

Apple at least thus far are very insistent on none of those things happening as it reduces control over their walled garden. We'll have to see if they come to their senses and at least offer up support for memory cards even if they still refuse to give access to physically connected external storage devices (via USB, for example).

Without that, it'll have a niche similar to how Mac continues to have a small niche in content creation, but Windows devices will likely dominate in the end. Similar to how Mac once ruled desktop publishing but desktop publishing is now mostly dominated by Windows PCs. Then again many have familiarity with iOS so maybe they'll be willing to put up with the extra headaches required to do content creation on the iPad Pro.

Regards,
SB

 

PC sales are dropping because people don't find a good enough reason to upgrade the ones they already have, and not because everyone is suddenly using ios and Android devices with a keyboard and mouse to perform important productivity tasks.

The performance graphs you showed have nothing to do with the issue at hand. Ultrabook users don't complain about the lack of performance for productivity of their own devices (again leading to them not upgrading for several years in a row). People who do need/want higher performance than an ultrabook then they actually need a >35W Core i7 and/or discrete graphics card that will do circles around a twice as expensive ipad pro (and with the widely available software for productivity or gaming too).
Right now, the A9X's excellent performance is more of a solution looking for a problem than anything else. You could argue that all that memory bandwidth and pixel fillrate is needed to drive that 2732x2048 resolution... but then again my Surface Pro 4 m3 with half the memory bandwidth and weaker GPU runs a display with a similar resolution plus a 1080p external display without a hitch.
Well, maybe the ipad pro runs FEM simulations on Solidworks faster than a Core M?
Oh wait no, there is no such thing for ios.


I used a TF700 for about 3 years. It even had a proper keyboard, mousepad and supported external mice too. Trying to do anything important in it was always a disaster. Android (and its free/$1 productivity apps) just doesn't cut it. You don't just need multi-tasking and a browser with multiple tabs. You need decent keyboards, resizable program windows, easy access to file system, seamless external multi-monitor monitor support, a taskbar telling you what is already opened up, etc.

Are there professional hipsters out there who refuse to let go of their ipads and write 3000 word articles in a capacitive screen, taking tens of hours to do it? Yes, there are. Anyone serious enough to do stuff on a deadline is still typing in a PC, though.

Intel won't be returning because they never stopped making x86 mobile SoCs. Apollo Lake for tablets is still coming out, as confirmed by @Ryan Smith :
https://twitter.com/RyanSmithAT/status/727296885459816448

As for the chances of Intel doing full SoCs with integrated 5G in the future, it's just what I took from the press release but I could be wrong because the wording in their statement isn't very clear (probably on purpose).
Regardless, either if Intel continues to update the Atom line or if they make a Cannon/Icelake variation with on-chip southbridge (not just on-package), I do think they'll be coming back with new SoCs carrying integrated 5G baseband. Even if they do continue to sell these chips at the premium they're used to and invent a new line of >$800 smartphones (superphones?).
I wouldn't mind if my next laptop was also my smartphone and vice-versa. I'd actually be saving money with that.