backtop


Print 28 comment(s) - last by torpor.. on Mar 23 at 6:39 PM


Anand Chandrasekher   (Source: Intel)

Intel, distantly behind ARM in the ultra-mobile space has had plenty of talk, but today appears no closer to gaining any ground in tablets or smart phones.  (Source: HotHardware)

No major smart phone has an Intel processor in it today. And only a handful of tablets carry last-generation power-hungry Atom CPUs (like the ExoPC and HP Slate). Best-selling models, like the iPad exclusively use ARM.  (Source: VILLA)
Senior vice president of Intel's ultra mobile unit company peaces out

A few more months have rolled by and Intel Corp. appears no closer to regaining any market share in the tablet and smartphone market.  Even as ARM CPU manufacturers like Qualcomm, Texas Instruments, NVIDIA, and Samsung sell millions of units, Intel has yet to ship a single modern smartphone CPU.

Now Intel has been dealt another setback in its mobile campaign.  Its senior vice president of mobility has quit, offering little explanation for his sudden departure.

I. Anand Chandrasekher -- One of Intel's Best and Brightest

Clearly Intel wanted desperately to begin to take advantage of the booming smartphone chip business.  That hope was evidenced by its decision to throw one of its brightest stars at the effort.

That star was Anand Chandraskher.  Mr. Chandrasekher had effectively given the world Wi-Fi, leading the team who first pushed the standard with Intel's Centrino wireless chipset.  Even at Intel, it's hard to find someone in the realm of mobility who had made a bigger difference than Mr. Chanrasekher.

Hopes were high that he'd be able to right the ship and steer Intel's struggling ultra mobile division to good fortunes.

II. Where's the Product?  The Curious Case of Intel's Ultra Mobile CPUs

Smartphones took off beginning in 2002 with the introduction of RIM's Blackberries.  And they gained an even higher profile with Apple's iPhone launch.  And today, Google's Android OS is the world's top smartphone OS selling tens of millions of units a quarter.

Tablets, are a more recent craze almost exclusively driven by Apple's iPad, which launched in early 2010.  Today Android tablets are slowly gaining a small market share as well.

There's one thing all of these devices have in common -- they don't use Intel processors.

That's a nightmarish situation for Intel, as the company has long maintained a nearly 90 percent market share in the desktop, laptop, and server space.

Despite having a mobile line of processors -- Atom -- Intel has struggled to shrink these processors' power requirements to levels competitive with ARM designs.  Vocally, however, it has remained confident.  Even at this year's Mobile World Conference 2011 (MWC 2011) in Barcelona, Spain Mr. Chandrasekher was bragging about Intel's upcoming mobile CPU offerings, Medfield (smartphones) and Moorestown (tablets), claiming they will beat ARM chips in both active power performance and processing power.

Along the way Mr. Chandrasekher and his colleagues have made many similarly boastful claims.  But for all the bravado, third parties have been unable to validate Intel's statements as the company's next-gen smartphone and tablet CPUs have essentially become vaporware.

While samples reportedly have finally shipped to hardware makers, there's much uncertainty about whether Intel Medfield-powered smartphones and Moorestown tablets will be able to meet ARM CPU designs' power and processing performance, much less beat it.

III. The Chief Quits

Amid this tense atmosphere for Intel, Mr. Chandrasekher appears to have tossed in the towel.  He announced his departure from the unit and the company with a brief email, writing, "I have done what I wanted at Intel and I felt it was time to explore other opportunities."

The announcement is a huge loss for Intel at a time when it is behind its competitors and has little room for error.

Moorestown, one of Intel's new tablet chips, has faced delays.  It has been shown off powering a number of demo platforms over the last year, but has yet to ship actual product.

States Linley Gwennap, principal analyst at The Linley Group, a chip consulting firm, in aCNET interview on Mr. Chandrasekher's departure, "Moorestown was a complete flop. Intel is still struggling to get traction in tablets and particularly smartphones. Atom [Silverthorne] is in a few tablets that run Windows, but Windows tablets are not very popular, except in a few vertical applications."

As for smartphones, Mr. Gwennap adds, "It remains to be seen whether Medfield (the next Atom for smartphones) will do any better, since Intel has not disclosed any details on that product yet. In addition to finding customers, the new [Ultra Mobility Group] management has a big task in figuring out how to integrate Atom with the ex-Infineon baseband products."

Other analysts were not much kinder.

Ashok Kumar, an analyst at Rodman & Renshaw writes, "The industry has gone right past them. They're just another player [in the smartphone and tablet markets]. There's no first among equals.  Legacy (Windows) compatibility doesn't matter in the post-PC era. All the blockbuster products that Apple has had are post-PC. Therein lies the conundrum for Microsoft and Intel."

IV. Intel Looking for a Winner

Intel is desperate to gain some sort of traction amid these struggles.  

As Mr. Chandraskeher announced his departure, David Perlmutter, executive vice president and Intel Architecture Group (IAG) general manager tried to rally the troops, writing, "Intel remains committed to this business. We continue to make the investments needed to ensure that the best user experience on smartphones and handhelds runs on Intel architecture, and to ship a phone this year."

Intel installed dual vice-presidents Dave Whalen and Mike Bell to co-manage the ultra-mobility group.

The question remains whether they can pick up where Mr. Chandrasekher left off and bring next-gen Atom tablet and smartphone CPUs to market.

And the pressure on Intel's mobility team grows monthly.  With ARM's tremendous success in the ultra-mobile space it's looking to establish a beachhead in mobile (laptop) sales in the near future.  Microsoft has already announced it will support ARM CPUs with its next desktop/laptop version of Windows.  With a multitude of ARM chipmakers announcing powerful, highly-clocked, multi-core designs, Intel must realize that if it can't take the fight to ARM, ARM is about to take the fight to it.



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

There's already a solution for this.
By arthur449 on 3/22/2011 10:59:53 AM , Rating: 3
Well, if they can't compete with in-house technology right now, they can bite the bullet and license ARM and name the product an Intel ARM CPU.

You know, like Intel did with Marvell for an SSD stop-gap.




RE: There's already a solution for this.
By Flunk on 3/22/2011 11:47:06 AM , Rating: 4
You mean like Intel used to do with the X-Scale before they sold it to Marvel?

Actually the X-Scale was a very competitive product back then. It's a shame they sold it. x86 everywhere doesn't make any sense for platforms without legacy code. Coders don't care what instruction set the CPUs use if they're coding in high level languages (and seriously, who isn't?).


RE: There's already a solution for this.
By Strunf on 3/22/2011 1:43:55 PM , Rating: 2
The problem isn't X86 it's the CPUs that Intel have on its portfolio that just aren't competitive enough in terms of power needs, Intel could even release a less powerful CPU for phones while needing the same power as the ARMs then they would have some market share by just playing with the prices... it was a bad move on their part to discard the X-Scale while not having something else already up and running to offer.


RE: There's already a solution for this.
By Flunk on 3/22/2011 3:17:25 PM , Rating: 2
You're missing the forest for the trees, the problem is x86. It's just too complex to create a CPU that runs in the required power envelope with acceptable performance.


By Strunf on 3/23/2011 8:49:04 AM , Rating: 2
That's just not true, the 486 could run X86 over 20 years ago, are you saying that the ARM cpus are less complex than that?... the single fact that X86 processors have become more efficient with each generation means that the "problem" doesn't come from the X86 but from the CPU design.

Even if the CPU it's X86, internally it's not running X86, it just has a "translator" that changes the X86 into whatever other code they use.
What you speak of is the RISC (ARM and others) vs CISC(Intel and others) comparison, but even the I7 is only a CISC on the looks, internally its just like any other RISC...


By fteoath64 on 3/23/2011 1:53:20 AM , Rating: 2
Intel's problems during then was that ATOM was to replace Xscale as people inside Intel thinks X86 in a low power way would somehow outshine ARM architecture. While ATOM was a nice implementation, it did NOT evolve as much as it should considering the time it had.
It also had a nasty "problem" in competing with the low-end Celeron and later C2D chips for small-formfactor ultraportables. The unfortunately thing was its performance compared to mainstream chips were just a fraction. This started the NetBook revolution which Intel hopes was going to be short-lived. It was not and the tablet evolved from it not from X86 but with ARM. Horror of horrors.

Still Intel can still make a really great ATOM and a beefed up chipset but it has to try way harder than before. A9 dualcores and quadcores are just being released so time is of essence ....
Sent from Q6600 @ 1.2Ghz ....


RE: There's already a solution for this.
By ET on 3/23/2011 2:56:33 AM , Rating: 2
They can, and might even do decently in that market, but this doesn't play to Intel's strengths, and I don't think Intel wants to be an also run.

x86 still has a benefit in this day and age. If Intel can get into phones now x86 can still provide it an edge. Imagine a phone that you can use with a phone UI on the road, then dock and use it as a Windows PC to run your favourite software. I think such an idea has some attraction, and could prolong the PC age as we know it -- in another form factor, but with the same software and capabilities.

In the long run, I think that x86 could reach parity with ARM solutions in terms of performance / power. As the ARM chips grow more complex, x86 overhead percent will grow lower. However, the window of opportunity for x86 added value might close by the time Intel makes the Atom competitive.


By Da W on 3/23/2011 8:49:59 AM , Rating: 2
Given time X86 will be usable in phones. The problem is that ARM is moving a lot faster than Intel (and AMD for that matter) expected.


Not really sure why this is such a big deal...
By hans030390 on 3/22/11, Rating: 0
By Aloonatic on 3/22/2011 12:17:27 PM , Rating: 5
I think that Intel are worried about the way the consumer market is moving, rather than ARM specifically trying to get into the desktop market.

We've already seen a shift in computing within the home moving away from the PC in the home office/study, to the laptop. Then more affordable notebooks & netbooks came along, taking their piece of the pie too. Now we have tablets coming along to take money from peoples' wallets (and purses) as well.

I don't think that desktops and powerful, desktop replacing laptops are going away, per sa, but the share of the market that Intel can dominate with it's traditional processors is shrinking, while consumers move to lower power, portable computing, whether anyone here thinks that that makes sense or not.

Intel would be mad to not recognise this, and Atom kinda did a job for a fair while, buying Intel some time, but you (and Intel) have gotta ask yourself what sort of processors are going to be in the devices that people are walking out of the shops with in 5 years time. Will the majority be processors of the kind where Intel dominates now, or are they more likely to be the kind or processors where ARM are at? If Intel were to focus wholly on the high power, high performance desktop and laptop processor market where they dominate now, they may well find themselves as being more of a niche processor manufacturer as fewer and fewer homes demand that sort of processing power (thinking about it, how many home users really need i7 levels of performance?) and businesses realise that they don't need them either, for general office work, let alone the rapidly growing mobile device market.


By tamalero on 3/22/2011 2:49:28 PM , Rating: 2
I second this motion.
People are moving away from big bulky desktop counterparts and whants efficient mobile and easy to use devices that can be used almost anywhere.
Intel as minimum reach in this section and thus since its the fastest growing market (while normal computers are kinda stagnant) not surprised that intel gets the creeps.
Since its a market Intel also cannot control at their will with hidden dirty monopolistic tricks ala DELL & HP.


RE: Not really sure why this is such a big deal...
By mmp121 on 3/22/2011 12:27:23 PM , Rating: 2
Its the first serious attempt to use a mobile device in a laptop form factor. Same goes for all the tablets in the market. They are all ARM based. Tablets are where ATOM was supposed to end up. Intel needs to get its ducks in a row or the mobile / portable market will become mainstream, and they will be left with the niche server / and a shrinking desktop market.


By mmp121 on 3/22/2011 12:28:09 PM , Rating: 2
I was referring to the Motorola Atix in that comment, sorry.


By Ushio01 on 3/22/2011 3:57:02 PM , Rating: 3
Here's an example of why Intel is worried the Nokia N8 it can connect with wifi to the router and with bluetooth to a keyboard and mouse, hdmi to a 1080P monitor, 3.5mm TRS to speaker's, usb to an external harddrive and still plug into the mains with the Nokia propriety charger.

It completly replaces the physical tower of a desktop and then look at the SoC in the ipad2 and then realise that the Sony NGP due to be released later this year doubles that and with the internet how many consumers still buy boxed software that needs x86 cpu's.


How is losing a manager a tragedy?
By name99 on 3/22/2011 1:58:11 PM , Rating: 5
How exactly is this guy "one of Intel's best and brightest" or "a wireless guru"?
I did a quick web search of him and, as far as I can tell, he's been a manager all his life. If there's a single product he ever actually DESIGNED or even worked on in an engineering capacity, the web has not heard of it.

Looks to me like what we have here is a manager who was a "genius" for as long as he was in the right place at the right time, then when life became tough, it turns out that there never was much genius there. In other words the standard US valorization of management as the people that make things happen, screw the engineers.

Look, you, I and everyone else in the world know why Atom is a sad POS that cannot function in the market that it is supposed to be targeting --- the phone and tablet market. The problem is a MANAGEMENT decision, to retain every idiotic design feature of the x86 platform that was ever thought of, even though this would have been the perfect place to ditch every such feature that made no sense.
And it's not like this is unique in Intel history. Remember that previous great triumph of marketing ideas over engineering common sense, the Pentium 4? Guess who was in charge of that effort.

As far as I can tell, Intel is losing someone who has made a number of bad calls (probably because he has limited technical expertise) and his replacement may well do a whole lot better; though it's a tough problem, given how much time Intel has pissed away on Atom. Maybe they simply have to accept another two years of being on the outside while they design a new mobile chip, this time one that is actually designed for the FUTURE not the past. (Hint, it has ONE mode, ONE, not x86 and 286 and SMM and the rest of that crap; and that one mode is x86-64, with NO x87 and NO MMX; heck no segments. Include a lightweight version of AVX that maybe does two ops per cycle rather than eight --- but which has the capacity to grow with time.)




RE: How is losing a manager a tragedy?
By bah12 on 3/22/2011 5:21:58 PM , Rating: 3
I agree very sensationalistic of the author. Especially this part.
quote:
The announcement is a huge loss for Intel at a time when it is behind its competitors and has little room for error.
No it isn't! It is, after all, his damn division that is failing to perform. Where is Mick's GM execs heads need to roll attitude, huh? What GM execs don't perform and they need to get the boot, but this guy is somehow special because he has successfully ran his division into obscurity.
quote:
That star was Anand Chandraskher. Mr. Chandrasekher had effectively given the world Wi-Fi,

And what is with the BS that Centtrino ushered in wifi? LOL wifi was widely used in laptops long before the first centrino design. Now centrino did bring good emprovements with regards to power, but I distinctly remember an Anandtech article discounting the glory of Intel's centrino marketing related to wifi, since it was already wide spread in laptops at the time, and was not the holy grail that Intel's PR was selling. Better yes, but revolutionary hardly, and wifi was well on it's way with or without centrino.


RE: How is losing a manager a tragedy?
By dsumanik on 3/23/2011 9:06:54 AM , Rating: 2
Actually centrino was just a marketing ploy from what i recall...in order to get a centrino sticker on your laptop you had to have a genuine intel processor, chipset, and wifi card

it was the combination of the three which earned a laptop the genuine centrino moniker


By bah12 on 3/23/2011 10:24:36 AM , Rating: 2
Correct which is why this paragraph is absurd. Sorry I should have quoted the whole line. Centrino had a wifi aspec to it, but first pushed the wifi standard? That is just BS. Obviously the guy was admired by Mick, and has been a friend to AT, but lets call a spade a spade. His division is crap and is greatest accomplishment was a marketing ploy (aka Centrino).

quote:
That star was Anand Chandraskher. Mr. Chandrasekher had effectively given the world Wi-Fi, leading the team who first pushed the standard with Intel's Centrino wireless chipset.


Winner!
By thrust2night on 3/22/2011 12:06:58 PM , Rating: 4
quote:
IV. Intel Looking for a Winner


If Intel is really looking for a winner, they should hire Charlie Sheen.




RE: Winner!
By Brandon Hill (blog) on 3/22/2011 12:22:02 PM , Rating: 2
Bi-Winning!


RE: Winner!
By BioHazardous on 3/22/2011 12:31:52 PM , Rating: 2
They made the mistake of jumping the gun and hiring will.i.am instead of winning with Charlie Sheen.


RE: Winner!
By brandonicus on 3/22/2011 3:38:23 PM , Rating: 2
Intel's new chips featuring tiger's blood! These chips are tired of pretending they aren't special. They're tired of pretending they're not totally bitchin’ rock stars from Mars.


Eh
By BruceLeet on 3/22/2011 2:12:42 PM , Rating: 1
He probably sold out Intel, gave away company tech/secrets and had some money transferred to an offshore bank account. Now he goes back to India where he will live lavishly.

Makes you go hmm when you don't get a bonus, or ask for one...isn't that what all top tier execs do?




RE: Eh
By bah12 on 3/22/2011 5:28:01 PM , Rating: 2
Doubt it, dollars to doughnuts he was canned. Bonus? why would this guy deserve a bonus, his division has failed miserably to keep pace in the mobile market. Why Mick is making him to some kind of marter is beyond me. Especially when he has been harsh on other industries for allowing bad management to stick around.

Now I don't know the inside scoop, maybe his department was not getting the funding needed for him to be competitive, but part of what makes a good manager is his ability to get that funding and push his vision. Even if he had a competitive vision, he still failed as an effective manager in getting that vision funded and produced.


RE: Eh
By KOOLTIME on 3/23/2011 5:46:02 PM , Rating: 2
No boss will say, you all stay and work, i'll quite so you dont lose your jobs due to budget cuts ?? since when has that ever happened?? the bosses cut people in dept to maintain their dept budgets so they can keep their jobs, doesn't matter how many under them lose their jobs or how good of an employee they may be, they wont risk losing their jobs over someone else's no matter what. long as they maintain budget and keep their jobs going, thats how it works.


You're all wrong
By Shadowmaster625 on 3/23/2011 8:40:44 AM , Rating: 4
The problem with Intel's Atom is that they are charging $64 a chip, in the case of the N450. Obviously, that is obscene. Intel could easily take an N450, clock it down, lower the voltage, and make it consume 1 watt and still perform better than any ARM cpu. But Intel charges a premium every time they go ULV. So now we're talking about potentially a $90 chip. No smartphone maker is going to bend over for that. Intel cannot compete because Intel will not accept less than 60% margins. Especially when very little smartphone hardware would be integrated into that expensive chip. It takes many generations of integration to catch up to ARM in the mobile space, and Intel never got started with that effort because they could never accept the minimal if not nonexistent margins required for them to get their foot in the door.




Soooo old . . .
By kake on 3/22/2011 4:26:12 PM , Rating: 1
Oh Anand, how you've aged!




"Mac OS X is like living in a farmhouse in the country with no locks, and Windows is living in a house with bars on the windows in the bad part of town." -- Charlie Miller














botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki