backtop


Print 17 comment(s) - last by chizow.. on Feb 2 at 12:48 PM

2012 looks bright for ARM, even as competitors drop off

ARM Holdings plc (LON:ARM) is an innovator who's been to hell and back.  Born from the 1984 Acorn RISC Machine project, ARM grew under the financial investment Apple, Inc. (AAPL) and its UK parent Acorn Computers.  But mounting financial troubles at Acorn -- and to a lesser extent -- Apple threatened to sink ARM Holdings as the 90s approached.  A bold spinoff ultimately saved the profitable unit.  Divestment of stakes in the spinoff saved Apple, although it was not enough to stop Acorn from sinking.  

I. Apple's Record Profits Mean Big Profits for ARM

The brave little castaway grew from a glorified R&D project into a giant of the electronics world.  From modest beginnings in hard drive and appliance microcontrollers, ARM would go on to define itself as the mobile processor architecture of choice.  Virtually every smartphone and tablet sold today carries a chip that either uses ARM's proprietary core designs, or at the very least licenses ARM's instruction set.  Today there are by far more ARM CPUs in the wild than any other kind of CPU -- including Intel Corp.'s (INTC) traditional PC chips.

2011 was a landmark year of ARM.  While ARM has yet to wholly knock Intel out of the game, it's done its best over the last few years to marginalize it, as people have turned away time-wise from the traditional personal computer and flocked to its mobile evolution, the smartphone.

Smartphones
ARM Powers every smartphone on the market today, and almost all the tablets, as well.  From Android to iPhone, they all use ARM.  [Image Source: Esquire]

2011 saw Android surge out to a big lead, and then slow slightly.  Towards the end of the year, in Q4 2011, ARM's old ex-investor Apple charged back with the iPhone 4S.  But all of these phones had at least one important thing in common -- they all use ARM CPUs and they all funneled royalties/license fees to ARM Holdings.

ARM Holdings reported its Q4 results on Monday afternoon, and once again it blew past analyst estimates.  It pulled in 137.8M £ ($218.4M USD) in revenue -- up 20.9 percent over last year's total of 113.95M £ ($180.61M USD).  The company's pre-tax profit was 69M £ ($109M USD), up 45 percent from 47.6M £ ($75.4M USD) in Q4 2010.  The net profit was 33.1M £ ($52.5M USD) up from 29.7M £ ($47.1M USD).

British Pounds
[Image Source: World of Stock]

Analysts estimated that the company would deliver a revenue of 123.6M £ ($195.9M USD) and a pre-tax profit of 55.8M £ ($88.4M USD).

Warren East, Chief Executive Officer cheers, "In Q4 and throughout 2011 ARM has seen strong licensing growth, driven by market-leading semiconductor companies increasing their commitment to ARM technology, and more new customers choosing ARM technology for the first time. We have also seen our royalty revenue continue to grow faster than industry revenues as the ARM Partnership gains share in our target markets..."

II. Record Run Expected to Continue as ARM Attacks New Markets

Smartphones proved the biggest boost for ARM Holdings.  In a phone conference Chief Financial Officer Tim Score said that a single smartphone could earn ARM as much revenue as five to ten feature (voice-only) phones.

Over 2.2 billion ARM CPUs shipped in Q4 2011, and the company shows no signs of slowing down.  It added 25 new licensees in Q4 2011.  Among them is Hewlett-Packard Comp. (HPQ) who is cooking up the world's first ARM-based servers.  ARM also looks to invade the laptop segment, with the launch of Microsoft Corp.'s (MSFTARM-friendly Windows 8.

Windows Laptops
ARM will power Windows laptops and tablets later this year.  After sluggish delivery by Intel, Microsoft finally opened the gates to ARM and its partners with the new OS.
[Image Source: DailyTech/Jason Mick]

Even as chipmakers like Intel have slashed their outlets due to sluggish sales expectations for 2012, ARM is posting bold predictions.  It expects revenue to rise to 126.2M £ ($200M USD) or more in calendar Q1 2012.  And it expects a total revenue of 542.6M £ ($860M USD)  for the the year.

The future looks truly bright for ARM as it maintains its ubiquitous dominance of the mobile space and begins to expand into new markets.

Sources: ARM, WSJ [Analyst estimates]



Comments     Threshold


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

That is one fat profit margin
By Mint on 2/1/2012 12:46:24 PM , Rating: 2
$109M pre-tax profit on only $218M revenue? I know it quite possible when you're fabless, but wow.

Still, when you compare that to Intel's revenue and profit, it's easy to see why they didn't care until now. Their pretax profit is over 40x as big.




RE: That is one fat profit margin
By Gondor on 2/1/2012 1:03:05 PM , Rating: 5
They are not only fabless, they don't produce anything tangible . They just license the ARM IP, be it complete chip design or just ISA.


By MrBlastman on 2/1/2012 1:09:06 PM , Rating: 5
As a general rule, I would ignore EBITA (Earnings before interest and taxes). You see, prior to the 2000 bubble crash, everyone was hyping EBITA for tech companies. It did nothing but get everyone in trouble, in the end.

Instead, focus on the net profit. Here, we see they have about a 24% margin by the numbers above. While that is ridiculously high, it really isn't if you compare it to say Apple which averages 24-25%, Microsoft at 33%, IBM at say 15% and Intel at 26%.

Nice numbers for sure--but in outer space? No, not really considering what others in the industry achieve.


RE: That is one fat profit margin
By sigmatau on 2/1/2012 2:22:01 PM , Rating: 1
Intel is dumb. They are taking way to long to enter the smartphone market.


RE: That is one fat profit margin
By Etsp on 2/1/12, Rating: 0
RE: That is one fat profit margin
By someguy123 on 2/1/2012 8:44:51 PM , Rating: 2
There's really nothing they could do other than shrink gradually while trying to create a working finfet process. With the move to 22nm and finfet they could very well have an x86 processor with a comparable power draw to ARM's designs. Right now they're still aiming for tablets first, but looking at 32nm atom's total power envelope of 3.5w medfield could very well push intel into ARM smartphone territory.


RE: That is one fat profit margin
By Hector2 on 2/1/2012 9:06:04 PM , Rating: 2
I agree. Intel has always been targeting their "real" entry with 22nm which was necessary to get the power & size way down. Until 22nm, they can't really compete with the small and power-stingy ARM. With 22nm, they're ahead of ARM.

They did the same thing with their first Celeron back in 1998. All Intel did was take the 2 L2 cache chips of the 300MHz Pentium 2 cartridge, then dropped the price way down to undercut AMD in the low end. People laughed at the performance without L2 cache but it got their foot in the door until they came out with their single chip version with integrated L2 cache. And it didn't hurt that the cache-less Celeron could be overclocked from 300MHz to 450MHz. It took the low end market over by storm.

These early 45nm & 32nm smartphones gets Intel's foot in the door to get manufacturing started, to get software built up, and to fine tune the product until 22nm is out. Similar to the Celeron going from 3 chips to 1, Intel smartphone entry started at 2 chips on 45nm, then to 1 SOC on 32nm, next to 22nm. It's just like a Tsunami. Early on, it's just a ripple, but when it hits shore, it's devastating.

Dumb ? Hardly. ARM just had record revenue & profit. That's great. Intel just had their 7th record-breaking quarter in a row and continue to smash both revenue & profit records. All that while spending tons of money on new Fabs.

The naysayers here don't know their tech history and don't understand the power of Moore's Law.


RE: That is one fat profit margin
By Fritzr on 2/1/2012 9:06:26 PM , Rating: 2
Let a FINFET licensee build an ARM chip and you could see some ultralow power consumption. Once the FINFET patent expires it will be available to anyone. Though by then it will likely be obsolete and vastly exceeded in efficiency :P

Intel needs to do what seems to be impossible. Build a chip with the power draw of an ARM that executes the full x64 instruction set.

Remember there was a time that 1000 nanometer process was considered the practical limit. These 'limits' are not barriers, they are research targets :)


RE: That is one fat profit margin
By Fritzr on 2/1/2012 9:09:08 PM , Rating: 2
Let a FINFET licensee build an ARM chip and you could see some ultralow power consumption. Once the FINFET patent expires it will be available to anyone. Though by then it will likely be obsolete and vastly exceeded in efficiency :P

Intel needs to do what seems to be impossible. Build a chip with the power draw of an ARM that executes the full x64 instruction set.

Remember there was a time that 1000 nanometer process was considered the practical limit. These 'limits' are not barriers, they are research targets :)


RE: That is one fat profit margin
By Hector2 on 2/1/2012 9:31:18 PM , Rating: 2
Everyone is already working on FinFETs. Patents aren't the issue. Intel has the secret sauce for their process and is 2 years or more ahead of everyone else having FinFETs in production. That's a huge advantage. And a lot of what they have are trade secrets that aren't patented. Getting the right recipe is just the first step. Getting it out of the lab and into production is hard --- and very expensive.

With 22nm, Intel's power drops way down and the die size is very small, making it a lot cheaper. Yes, ARM on 22nm FinFET is lower power than Intel on 22nm FinFET, but so what if ARM has a battery life of 4 weeks and Intel is "only" 2 weeks ? Is that really a problem ? And when ARM is on FinFET (14nm at the earliest), Intel will be on 10nm or 8nm FinFET. Intel already has 14nm FinFET owrking in the lab.


2.2B ARM shipped is accurate???
By chizow on 2/2/2012 12:06:34 AM , Rating: 2
If so color me unimpressed with ARMs financials. 2.2B ARM chips shipped in Q4 and only $218M in revenue means they are licensing each chip for roughly $.10 a pop. You can't buy ANYTHING nowadays for a dime anymore, yet ARM happily licenses their chip for that much?

Honestly ARM needs to get some more skin in the game because right now, they are still on fragile footing. Tiny company could easily get gobbled up by a big fish like Samsung, I'm surprised no one has made a push yet.




RE: 2.2B ARM shipped is accurate???
By B3an on 2/2/2012 2:18:40 AM , Rating: 3
This is actually pretty normal in the world of these type of small mobile chips and competing companies. And this is one major reason why ARM do so well. The second is because they make chip designs far better than any American company can (Intel and especially AMD).

If ARM didn't license so cheap the chip design alone likely wouldn't make them have more CPU's out there than any other company, and be so attractive to so many hardware makers. This also gives them an advantage for Windows 8, because it's looking like ARM tablets will be way cheaper than any Intel powered alternatives. And being as ARM W8 tablets wont run x86 software, ARM need this large price advantage.


By chizow on 2/2/2012 12:48:36 PM , Rating: 2
I don't have a problem with their strategy, its obviously working great. I just have a problem with their return, assuming that figure is accurate of course.

Does no one else find the .10 cents per chip licensing fee incredibly low? I mean the technology that's changing the world in front of our eyes is being licensed out for a dime. Doesn't seem right.


RE: 2.2B ARM shipped is accurate???
By Calin on 2/2/2012 6:16:33 AM , Rating: 2
On the other hand, cheap ARM chips for cheap phones sell for maybe $10, out of which 90% or more is cost. Now, getting a tenth of the profits from what somebody makes is quite good in my oppinion


By chizow on 2/2/2012 12:45:21 PM , Rating: 2
I don't know where you're getting that 90% cost figure, I highly doubt its that high given firms like Intel and Nvidia are routinely posting 50% or higher margins on their silicon.

Obviously ARMs don't command the same price as these huge ASICs but they're also exponentially smaller and generally don't use the cutting fab processes, meaning further savings and yield benefits. For example, Tegra 3 is only 40nm, while the rest of the GPU industry is moving to 28nm now and the CPU industry is about to go to 22nm.


RE: 2.2B ARM shipped is accurate???
By A11 on 2/2/2012 6:40:52 AM , Rating: 2
"Over 2.2 billion ARM CPUs shipped in Q4 2011"

Somehow I doubt this figure is correct.

Apple and Samsung shipped what? ~100 million ARM units in total. Exactly where did the other 2.1 billion units end up?


By chizow on 2/2/2012 12:40:23 PM , Rating: 2
You have really mundane products like host controllers you'd find on HDDs or simple chips you'd find on home appliances nowadays. 2.2B is REALLY a crazy number though, but like I said, the amount they are getting on each part just seems really low.

I mean we thought the licensing #s on console hardware was low....


"I f***ing cannot play Halo 2 multiplayer. I cannot do it." -- Bungie Technical Lead Chris Butcher














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