Intel uncloaks 'no excuses' smartphone for emerging markets Intel has announced a smartphone reference platform for emerging markets that's a "no excuses multimedia phone," according to the general manager of its mobile and communications group, Mike Bell. "We at Intel think that emerging-market customers shouldn't have to settle for a substandard experience," Bell said at an Intel …
Intel have never really understood low-margin chipsets so it is a huge corporate struggle to try to find a way to make any money in cash-strapped emerging markets. Even Nokia - cranking out billions of candy-bar phones - struggled to make a profit.
I think Intel misread the market quite badly. When you have absolutely nothing, then what a westerner might call substandard is pretty bloody fancy and high spec video over cellular is an expensive luxury.
That surely means the less-than-hottest specced parts will still rule in emerging markets because the prices are unbelievably low and Intel, yet again, will fail to ship into this market.
But what do I know? I only lived in the 3rd world for 30 years. Perhaps Mike Bell has a better understanding than I do and understands the market better.
Perhaps 'emerging markets' is a bit too loose of a description? If they are pitching it at Ethiopia then it probably misses the mark, but if its aimed at India then perhaps it might stand a chance of gaining market share and intel 'mindshare' (****ing hate that term sorry) as the country develops. It will be tough going up against arm but intel in theory can do it. Their motives may not be immediate profit but rather future profit.
Where the average pay in these emerging markets would be classed as slave labour anywhere else in the developed world unless the product is being sold for peanuts, less that £20 say, it is doubtful that they will sell at all.
In fact I think £20 to the average Bangladeshi, Indian, Peruvian, Angolan, Anfgani, Mongolian,Malian etc etc etc will make the phone out of reach for most. They would have limited sales for those with better paid jobs in the emerging market because the phone would be competing with other brands.
It is true that most people will have little money to spend on phones, but let me offer two observations:
1) Average figures are of no use in analysing these markets: India alone has around 1.3 billion people. If you focus only on the upper 20% then you've still got a market not far off the size of the US.
2) In the developing world marginally more people have mobile phones than have access to safe drinking water (although that's not currently full on mobile internet or smartphones). The people of these countries are already demonstrating that they want or need these devices, whether we approve or not.
I think Intel recognise that the value of the basic candy bar handset are gone (Goodbye, Nokia!) and that even for the poor there's significant utility from a basic smartphone. Education, the economy, democratcy, public services - all things which can be improved by better data and telecomms, and which are well provided by smartphones in the absence of Western luxuries of fixed lines, dependable power, and cheap desktop computing.
Intel is probably going to do far more for the developing world by trying to make money than the UK government will by giving away £50 billion quid over the life of this government.
Those upper 20% are desperately trying to not be served with any products tainted with "emerging market" stigmas. They will want the same hipster toys the Westerners have. Apple/Samsung etc. If a product can't fly in the West, then it won't fly for these people.
The bottom/middle end are well served by existing cheaper ARM chipsets with all the capability to serve the bottom end.
Where is the gaping market opening that a huge organisation like Intel can find a toe hold?
"Intel is probably going to do far more for the developing world..." They're just trying to do a small part of what Huawei et al are already doing,
Intel have tried to scale this mountain many times before and have not succeeded. The market and competitors are now both a lot tougher. There is nothing in this new Intel offering that is enough to change the game.
Batteries are still pants. Pink polka dot pants.
"I can watch the lord of the rings trilogy on my awesome new iclone at 1080p"
"For seven minutes"
And i have access to a charger in my car, a dock at home and a usb cable at work... yet as soon as i make phone calls on it it still dies on me! What if power was less available and more expensive? Lets hope the new chips are less thirsty...
Doesn't really matter. As your comments illustrate, there many ways to charge a battery, and in developing countries then the unreliable mains (if even available) is best used to charge batteries rather than try and power a desktop PC whilst praying that the power doesn't fail or fluctuate. Hand crank chargers, disposable batteries, solar chargers etc etc, they can all work.
On the subject of chip energy use, that's exactly where Intel has been putting a lot of effort. Given the demands from non IC power use (so GPS, Tx/Rx, screen etc) the improvements will be modest, but my guess is that in the developing world they may not be prioritising streaming of light entertainment media.
You might want to take a look over on Anandtech for an article today comparing the latest Atom variant with ARM, which doesn't really support this claim, even if historically it was true. I might also say that my (work) Intel smartphone shows no material performance differences from my SGS2 in basics like battery life and heat generated.
As I understand it the CPU isn't that big a cost itself in the total bill of materials (BOM) of a phone, compared to say the screen or the battery. However it can have big cost side effects if it is hard to integrate, skews the board design and worst-case requires a bigger battery.
ARM don't sell CPUs, they sell designs for cores that manufacturers integrate with as much other stuff as possible onto one piece of silicon. This is convenient and lowers the total BOM. Are Intel trying to sell a standalone low-power CPU? Regardless of how power-efficient or not it is, in a tiny embedded device that could end up being more expensive to use than the ARM option, even if Intel gave it away for free.
Battery is cheap enough, and all the phone and chip makers know that SoC is better than discrete components. The Orange San Diego is a very passable first stab, using the Atom (Medfield) and relatively speaking low end graphics.
The main difference between Intel and ARM is simply that ARM sell the IP, and let others design onto and around that, and ARM also expect others to manufacture those designs. WIth Intel they want to go the whole hog of design and build, so you're paying for their fab, you're playing on their rules, and if you want anything different you can go swing. Potentially the Intel solution can be as cheap as ARM, and cheaper still if they don't make different versions for every phone.
If you were a second rank smartphone maker (or lower) making thin pickings from your own ARM based phones, then it might be worth taking Intel's shilling, sacking your ARM-skilled design team and accepting that you're going forward as a commodity phone maker on low margins. (HTC? Nokia?). The leaders will probably continue to want to use ARM because they can innovate, at least for the next few years.
Longer term, about 2017-2020 we're looking at 10nm processes that will put the power of a current desktop into a mobile phone, and then I think the game might start to shift in Intel's favour, as people start to want their phone to do the things that we currently expect of a full sized computer. But even then, the excitement may not be on the CPU, but on the GPU, in which case we're looking at Nvidia and Imagination. AMD's purchase of ATI may prove to have been an inspired move, but executed a decade too early.
Second and third tier smartphone manufactures do not design their own SOC. They buy them ready made from Qualcomm et al.
Snapdragons already use Radeon* cores for graphics. Qualcomm bought the mobile graphics division of AMD (formerly ATi) in 2008.
*Adreno actually but that's just an anagram.
> Potentially the Intel solution can be as cheap as ARM
There are ARM chips that sell for as little as 0.50 (50cents) in lots of 1000. Granted these are really slow and limited but good enough for switches and stuff.
> we're looking at 10nm processes
maybe, and so are ARM licencees. I see comparisons between last year's ARM and next year's Intel purportedly showing how Intel will win. Do they think ARM will twiddle thumbs while others catch up ?
That is the major reason for requiring an Intel processor after all.
And Intel are used to selling that privilege at a premium price.
If it's a computer (that allows you to make phone calls and you can put in your pocket) perhaps quite a lot.
If it's an appliance that you might get to do a few other jobs then not so much.
I think "instruction set compatibility" is rather a long way down most peoples wish list for a mobile phone in developing countries.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
