Linked by Thom Holwerda on Thu 22nd Apr 2010 09:22 UTC, submitted by REM2000
In the News Well, you can classify this under the double-you-tee-eff header. There is rampant speculation in London's financial district that Apple may be planning to buy ARM, the processor design company many of us have a soft spot for. Shares of ARM went upwards quickly when the speculation started, making it the biggest winner of London's FTSE.
Order by: Score:
The classic acquisition problem
by alcibiades on Thu 22nd Apr 2010 09:32 UTC
alcibiades
Member since:
2005-10-12

The classic acquisition problem is, after you buy it, what do you do differently, and why? And if it was such a great idea to do that, why was the target not doing it anyway?

So what exactly will they do when they have it? They cannot cut off sales to other competitive companies, or they will end up with a hugely loss making subsidiary, owing to lack of sales mass. They cannot move to less favorable terms, since probably, to retain the outside business, given their reputation for arbitrariness and unpredictability, they are likely going to have to make the terms more favorable rather than less.

Do we really think there is overhead to be eliminated by the merger? There's lots of fat in Cupertino, but that is not going to be affected by this.

In short, it seems like a completely pointless acquisition. Not that this will stop Apple from doing it of course. Once the investment bankers get into a company, the social dynamics of it make it almost impossible to change your mind and get off the train. So if they are in, we can expect this to go ahead, and for Apple to lose huge amounts of money and management energy on it.

Reply Score: 3

Cody Evans Member since:
2009-08-14

Whatever loses ARM causes by lost sales, will be made up for by what Apple does with ARM. Kicking all other smartphones off of ARM will make the iPhone the ONLY smartphone for sale.

But a question remains, what about the non-phone electronics that use ARM?

Of course, if they do this, you can bet that all other phone makers will cry foul to the DoJ.

Reply Score: 1

fatjoe Member since:
2010-01-12

I believe Apple is accountable for far, far less than 5% av ARM sales.

Beside, the EU would never let this happen.

Reply Score: 1

Lobotomik Member since:
2006-01-03

I'd say far, far less than 1%. The ARM is into pretty much every phone in existance, smart or dumb, as well as media players, navigators, POS terminals, printers, bluetooth chips, gaming consoles, photo cameras, you name it.

Apple is VERY SMALL fish in ARM's market.

The range of products that makes ARM successful is unsustainable with the three-chip-designs-per-year rhythm and the low volume.

Edited 2010-04-23 08:15 UTC

Reply Score: 2

glarepate Member since:
2006-01-04

http://www.pcmag.com/article2/0,2817,2362956,00.asp

Apple still could launch an unsolicited takeover of ARM. But in comments to the U.K's Guardian paper, ARM chief executive Warren East appeared to deny that the two companies were in talks.

"Exciting though it is to have the share price pushed up by these rumours, common sense tells us that our standard business model is an excellent way for technology companies to gain access to our technology. Nobody has to buy the company," East told the paper.

Reply Score: 2

I'm a bit of an Apple fanboi...
by thavith_osn on Thu 22nd Apr 2010 10:00 UTC
thavith_osn
Member since:
2005-07-11

...but I don't quite see why they would do this either.

I don't like the idea that ARM is only used by Apple, so if that is their reason, then it's a very bad one (IMHO). Just like forcing Adobe out of the iPhone dev market (and others). I understand some of the reasons, but let the market vote with their $$$.

Apple needs to keep doing what it has always done, build great products and if they are good enough, let the audience pay, if not, then let them go else where.

Locking people in or out of a tech is a bad idea...

If however, they just want to have more control over what gets made and how it's made, then that is a another story. Apple had a lot of trouble in the past with Moto, and maybe they see similar things with Intel (though, I haven't heard anything if that is the case, other than Atom not being up to scratch in the past).

Either way, I can't see any point to this unless they are worried someone else will snatch them up and lock them out.

Reply Score: 4

what-the-eff indeed
by Stratoukos on Thu 22nd Apr 2010 10:05 UTC
Stratoukos
Member since:
2009-02-11

I don't really have a problem with Apple's closed platform, since whoever cares can buy a different device, but holy shit this scares the bajesus out of me.

I mean the only reasonable explanation for this acquisition is that they want to keep future ARM designs for themselves. I'm pretty sure that this won't affect chip builders that have licensed existing reference designs, but if this go through I can easily see them saying "ARM is ours, go to Atom. Have fun with your two-and-a-half-hours battery y'all" in two years or so.

There is alway a chance that they intend to keep ARM's existing business model and this move is about shaping ARM's future, but I put my money on the first scenario.

Of course it could all be just speculation, but still it's cringe-worthy.

Edited 2010-04-22 10:05 UTC

Reply Score: 7

RE: what-the-eff indeed
by lemur2 on Thu 22nd Apr 2010 10:29 UTC in reply to "what-the-eff indeed"
lemur2 Member since:
2007-02-17

"ARM is ours, go to Atom. Have fun with your two-and-a-half-hours battery y'all" in two years or so.


http://en.wikipedia.org/wiki/VIA_Nano

http://www.via.com.tw/en/products/processors/nano/

Reply Score: 3

RE[2]: what-the-eff indeed
by darknexus on Thu 22nd Apr 2010 12:56 UTC in reply to "RE: what-the-eff indeed"
darknexus Member since:
2008-07-15

Oh, please. If you want to say something, say it instead of just providing us links to Via's advertising. If you're trying to say that Via Nano gets better battery life than Atom then I hate to disappoint you, but I must. Nano's battery life, when you put everything together, is roughly the same as Atom N280. It's even outstripped a bit by the Atom N450 with a good chipset, and doesn't perform as well as the N450 in most cases anyway though it does run a little cooler. Nah, for embedded devices if ARM is no longer viable in the future, people are more likely to go to MIPS than anything based on X86.

Reply Score: 6

RE[3]: what-the-eff indeed
by twitterfire on Thu 22nd Apr 2010 13:24 UTC in reply to "RE[2]: what-the-eff indeed"
twitterfire Member since:
2008-09-11

Oh, please. If you want to say something, say it instead of just providing us links to Via's advertising. If you're trying to say that Via Nano gets better battery life than Atom then I hate to disappoint you, but I must. Nano's battery life, when you put everything together, is roughly the same as Atom N280. It's even outstripped a bit by the Atom N450 with a good chipset, and doesn't perform as well as the N450 in most cases anyway though it does run a little cooler. Nah, for embedded devices if ARM is no longer viable in the future, people are more likely to go to MIPS than anything based on X86.


Does it occurred to you that Via runs circles around any ARM CPU? You don't need the actual power of a Nano or Atom in a mobile phone. Via and Intel can cut some cpu power to improve on battery life.

Reply Score: 3

RE[4]: what-the-eff indeed
by bhtooefr on Fri 23rd Apr 2010 00:02 UTC in reply to "RE[3]: what-the-eff indeed"
bhtooefr Member since:
2009-02-19

How so?

Keep in mind, Intel's TRYING to cut Atom down to ARM-level power consumption, and struggling to stay just AS fast as the latest ARMs, and having drastically higher idle power consumption.

Reply Score: 1

RE[2]: what-the-eff indeed
by henderson101 on Fri 23rd Apr 2010 10:23 UTC in reply to "RE: what-the-eff indeed"
henderson101 Member since:
2006-05-30

""ARM is ours, go to Atom. Have fun with your two-and-a-half-hours battery y'all" in two years or so.


http://en.wikipedia.org/wiki/VIA_Nano

http://www.via.com.tw/en/products/processors/nano/
"

And now please, find a comparison* of power consumption between ARM and VIA chips. I suspect, ARM will beat them by a significant margin.

* As I did look, didn't find much info don't have hours to waste proving one way or the other and you seem to be so good at regurgitating these types of things.

Reply Score: 1

RE[3]: what-the-eff indeed
by henderson101 on Sun 25th Apr 2010 20:57 UTC in reply to "RE[2]: what-the-eff indeed"
henderson101 Member since:
2006-05-30

And now please, find a comparison* of power consumption between ARM and VIA chips. I suspect, ARM will beat them by a significant margin.


As I suspected - the reply was lacking. It's fun to pretend stuff, but reality is, VIA don't have anything that can touch ARM at the moment.

Reply Score: 1

RE: what-the-eff indeed
by werpu on Thu 22nd Apr 2010 11:35 UTC in reply to "what-the-eff indeed"
werpu Member since:
2006-01-18

If they buy arm the SOC designers will either go for low power powerpc or mips. ARM is not the only game in town with low power consumption, they just have been the best ones for years now.

Reply Score: 6

RE[2]: what-the-eff indeed
by twitterfire on Thu 22nd Apr 2010 13:06 UTC in reply to "RE: what-the-eff indeed"
twitterfire Member since:
2008-09-11

If they buy arm the SOC designers will either go for low power powerpc or mips. ARM is not the only game in town with low power consumption, they just have been the best ones for years now.


Even if some powerpc and mips are present in some embedded devices, those are old technology and don't have anywhere near the same performance per watt as ARM.

If Apple is to buy ARM, maybe we will see some new architectures coming up (maybe based on ARM) or we will see some new and improved mips and powerpc based designs as the current designs can't keep pace with ARM.

Reply Score: 2

RE[3]: what-the-eff indeed
by fatjoe on Thu 22nd Apr 2010 14:49 UTC in reply to "RE[2]: what-the-eff indeed"
fatjoe Member since:
2010-01-12

Actually, I think the original ARM design is older than the PowerPC.

In any case, all three architecture have been updated very recently, and more is to come [e.g. ARM Cortex-4/5/6 is coming this year]. ARM is still the market leader, but they will loose that position very very quick if Apple buys them.

Reply Score: 1

RE[3]: what-the-eff indeed
by nicolas.det on Thu 22nd Apr 2010 20:44 UTC in reply to "RE[2]: what-the-eff indeed"
nicolas.det Member since:
2009-10-03

"If they buy arm the SOC designers will either go for low power powerpc or mips. ARM is not the only game in town with low power consumption, they just have been the best ones for years now.


Even if some powerpc and mips are present in some embedded devices, those are old technology and don't have anywhere near the same performance per watt as ARM.
"

In my opinion, PowerPC core, in general, are just better. PowerPC is better: design, core, ISA, architecture...

The problem is that the major SoC vendor did not succes to build a sense-full chipset for embedded devices, notebook or desktop computer.
They succeed for game station (PS3, WII), network devices (router, firewall)...

As an example, a good 603e 600Mhz core (or LP G4) 512Ko cache with DDR controller, USB2, SDIO, I2S, I2C, NAND controller, real PM, 45nm fabric, coherent bus, display mit 2D/3D accelerator would really beat them all.


Regards,

Edited 2010-04-22 20:50 UTC

Reply Score: 2

RE[4]: what-the-eff indeed
by viton on Fri 23rd Apr 2010 11:07 UTC in reply to "RE[3]: what-the-eff indeed"
viton Member since:
2005-08-09

In my opinion, PowerPC core, in general, are just better. PowerPC is better: design, core, ISA, architecture...

Wrong. PPC is plagued by bloated ISA and dumb design decisions: you can view PPC CPU as logically separated processors (FX, FPU, VMX) feed from single instruction queue. For example the only way to convert float value in FP register to fixed-point register is to store and load it from memory. This causes 40 cycles pipeline stall known as LHS in console cores (360/ps3).
In p7 they tried to fix it in VSX block what is capable of FP and Int operations, but this is a half-baked solution what duplicated a lot of functionality of VMX and FPU.

u-arch of "console" cores is totally brain damaged. For example load signed data from memory is not supported in hardware (it is microcoded and leads to 21 cycles pipeline stall). Shift operations are microcoded too. Some libraries like zlib, heavily rely on shift instructions.

And that's not all. In power effeciency and code density PPC is not a competitor to ARM.

On the other hand take the power6 core for example. While individual blocks are unique state-of-art high performance design Intel can only dream of, overall result is rather dissapointing.

Reply Score: 2

RE[2]: what-the-eff indeed
by Downix on Thu 22nd Apr 2010 15:08 UTC in reply to "RE: what-the-eff indeed"
Downix Member since:
2007-08-21

You forgot SPARC as well, which is also a licensable design and has a few GPL'd forms available.

Reply Score: 1

RE[2]: what-the-eff indeed
by fithisux on Thu 22nd Apr 2010 17:01 UTC in reply to "RE: what-the-eff indeed"
fithisux Member since:
2006-01-22

Loongson 4ever!!!

Reply Score: 2

RE: what-the-eff indeed
by kristoph on Thu 22nd Apr 2010 15:15 UTC in reply to "what-the-eff indeed"
kristoph Member since:
2006-01-01

There are billions of ARM chips shipped each year. Even if Apple buys ARM they will continue to license the technology; they'll just keep the latest and greatest designs for Apple devices keeping Apple 6-12 months ahead of the competition.

Reply Score: 1

RE: what-the-eff indeed
by Earl Colby pottinger on Thu 22nd Apr 2010 18:21 UTC in reply to "what-the-eff indeed"
Earl Colby pottinger Member since:
2005-07-06

The problem with your statement is I have a Toshiba NB205, and it has a 280 Atom CPU (that dual cpu to those who do not know) and I am getting about 10 hours life out of the batteries today in Windows 7 and about 8-9 hours of life in Haiku. The SSD I installed may help explain the run times, but I clearly do not need to move to ARM to use this machine all day.

Intel is not made of fools, expect the Atom series plus support chips to get better in the future.

Reply Score: 3

It makes some sense
by mrhasbean on Thu 22nd Apr 2010 10:11 UTC
mrhasbean
Member since:
2006-04-03

If they buy ARM they will control the direction of future development of the platform and they get to make $$$ off every other device out there that uses the processors. They can also utilise the technologies owned by ARM to create custom processors that are only for their own devices, and it gives them a price advantage on the processors for their own devices. Lots of pluses for Apple.

Reply Score: 1

Comment by Kroc
by Kroc on Thu 22nd Apr 2010 10:28 UTC
Kroc
Member since:
2005-11-10

I wonder if shares in MIPS went up as well?

Reply Score: 2

Comment by Kroc
by Kroc on Thu 22nd Apr 2010 10:31 UTC
Kroc
Member since:
2005-11-10

Vertical integration baby. It was the difference between a PC costing $2000 and a C64 costing $595. In Apple’s case it would be the difference between 15% margin and 30%–50% margin.

Reply Score: 1

RE: Comment by Kroc
by Slambert666 on Thu 22nd Apr 2010 12:08 UTC in reply to "Comment by Kroc"
Slambert666 Member since:
2008-10-30

Vertical integration baby. It was the difference between a PC costing $2000 and a C64 costing $595. In Apple’s case it would be the difference between 15% margin and 30%–50% margin.


Apple does not make 15% on anything...
In apples case it would be the difference between 50% margin and 50.1% margin.
The only reasonable explanation for buying ARM would have to be restricting competitors access to new designs and this way keep the 50% rolling for longer (there is a sucker borne every minute after all).

Reply Score: 3

RE[2]: Comment by Kroc
by JAlexoid on Thu 22nd Apr 2010 16:35 UTC in reply to "RE: Comment by Kroc"
JAlexoid Member since:
2009-05-19

Apple does not make 15% on anything...
In apples case it would be the difference between 50% margin and 50.1% margin.
The only reasonable explanation for buying ARM would have to be restricting competitors access to new designs and this way keep the 50% rolling for longer (there is a sucker borne every minute after all).


Their last financial statement shows that their margins are closer to 25%, witch is still quite high.

Reply Score: 2

RE: Comment by Kroc
by Earl Colby pottinger on Thu 22nd Apr 2010 18:25 UTC in reply to "Comment by Kroc"
Earl Colby pottinger Member since:
2005-07-06

C64 dropped well below $200 while Commodore was still a profitable business. Later when the price dropped into the hundred dollar range Commodore was hurting.

But yes, vertical integration can really drive the costs down and/or profits up.

Reply Score: 2

Puzzled
by ameasures on Thu 22nd Apr 2010 11:16 UTC
ameasures
Member since:
2006-01-09

Apple already bought PA Semi a while back.

ARM has an IP portfolio but no fabrication capability (AFAIK).

There has to be something in the IP portfolio or the option of first dibs on new creations that are in progress.

Just now I had thought ARM were fairly fully priced.

An interesting development.

Reply Score: 3

RE: Puzzled
by cb88 on Thu 22nd Apr 2010 15:23 UTC in reply to "Puzzled"
cb88 Member since:
2009-04-23

I doubt Apple has enough money to buy ARM .... ARM would have nothing to gain from this.

Reply Score: 0

RE[2]: Puzzled
by JAlexoid on Thu 22nd Apr 2010 16:37 UTC in reply to "RE: Puzzled"
JAlexoid Member since:
2009-05-19

I doubt Apple has enough money to buy ARM .... ARM would have nothing to gain from this.

Apple does have, but ARM will not be sold without a bidding war. Or if it's magically sells itself to Apple, there is going to be a pricing war from all sides at once. Apple might even face "supply issues" with Samsung.

Reply Score: 2

RE[2]: Puzzled
by flynn on Thu 22nd Apr 2010 16:46 UTC in reply to "RE: Puzzled"
flynn Member since:
2009-03-19

I doubt Apple has enough money to buy ARM


You do realize that Apple makes as much money in a single week as ARM does in an entire year, right?

Reply Score: 1

RE[3]: Puzzled
by cb88 on Thu 22nd Apr 2010 18:24 UTC in reply to "RE[2]: Puzzled"
cb88 Member since:
2009-04-23

Well I had it figured that since ARM makes a profit on every cellphone and most all embedded appliances that they would be deeper in dough than Apple... perhaps I was wrong.

I still hold to the opinion that Sparc is the direction companies should take (assuming they are pro open source) MIPS = proprietary even loonson, PPC well same story just worse, and x85 isn't even worth talking about... that leaves Sparc which is a pretty slick RISC arch that is still relatively competitive and probably cheap to modify (S1 core, T2, T2 and LEON for example)

Sparc is the only alternative where you don't end up forced to pay someone for a license. I'm pretty sure oracle can't mess it up for anyone either since Sparc is a IEEE standard that any company regardless can implement.

Reply Score: 1

RE[4]: Puzzled
by bhtooefr on Fri 23rd Apr 2010 00:05 UTC in reply to "RE[3]: Puzzled"
bhtooefr Member since:
2009-02-19

Well... SPARC International could always decide that they don't want to let people use SPARC for free any more, if SPARC catches on.

MIPS-like is probably the absolute safest way to go - just don't implement the unaligned load/store, and you're safe.

Reply Score: 1

RE[2]: Puzzled
by bousozoku on Thu 22nd Apr 2010 18:56 UTC in reply to "RE: Puzzled"
bousozoku Member since:
2006-01-23

I doubt Apple has enough money to buy ARM .... ARM would have nothing to gain from this.


They just stated that they have just over $41 billion in cash and no long term debt. So yes, they have the money, though it would take about 20 % of what they have.

Reply Score: 3

kill apple
by nbensa on Thu 22nd Apr 2010 11:50 UTC
nbensa
Member since:
2005-08-29

before they kill the industry

Reply Score: 2

Erm, reality bites
by flibble on Thu 22nd Apr 2010 11:51 UTC
flibble
Member since:
2007-05-19

This isn't going to happen.

Apple Iphone sales (8.8 mill in one quarter) represent such a tiny part of ARMs sales (~1 billion per quarter), that Apple success or failure should be pretty much irrelevant to the share value of ARM.

For Apple to gain a controlling stake in ARM, they'd need 50% (plus a little bit) of the shares, which is currently valued at 1.6 billion UK pounds. ARM currently turns (in 2009) of 100 million UK pounds. So assuming no growth in profit (big assumption) it'd be 16 years to see a return on that investment, big companies (and their investors) don't really like working on that timescale.

Also this entirely discounts the possibility of ARMs current licenses (100+ of the worlds largest electronics companies) banding together to form a consortium to outbid Apple and keep their current very nice license terms.

More likely Apple will either buy a license for the latest ARM cores and design their own SoC for new products or buy a company that already specialises in that. Much cheaper and will bring them the same results.

Reply Score: 6

Hopefully not
by error32 on Thu 22nd Apr 2010 12:02 UTC
error32
Member since:
2008-12-10

I don't think this will happens when I look at the calculations in some other comments. However we know Apple can do some out of ordinary things so who knows. Yet I hope they won't do it because I can't believe this would be alright for an architecture like ARM the way we use it right now.

Reply Score: 2

makes no sense
by alcibiades on Thu 22nd Apr 2010 12:06 UTC
alcibiades
Member since:
2005-10-12

Kroc, you're not thinking it through. Vertical integration must mean dropping prices. So why is it worth paying billions to get a supplier to drop prices to you, and only you?

Then, in other comments, people talk about changing the direction. In what direction? Why has not ARM done it already, if its so obvious?

Then other commenters talk about holding back the good stuff for themselves for a couple years. Why would you do that? Just put ARM out of business. Drive everyone to the competition. Why did you buy it, to do that?

Reply Score: 3

Comment by Panajev
by Panajev on Thu 22nd Apr 2010 12:58 UTC
Panajev
Member since:
2008-01-09

I think Intel would LOVE it if Apple purchased ARM because it would be an incredible opportunity for x86 to take a good portion of those ARM licensees which would be quite worried about anti-competitive practices by Apple (focused on vendor lock-in... see latest SDK terms debate) and would give Intel a big reason to inves lots of R&D dollars into mobile x86 again... (like they did in the Desktop and workstations space when they pushed the Pentium Pro revolution against big RISC).

Also, what happened during the ACE consortium vs Intel days (Pentium days or so)... you bet that Intel will go to F.U.D. warp-speed to worry vendors about the future of ARM.

One last thing is about Apple and ARM... Apple has quite successfully detached the underlying HW (CPU, GPU, RAM, etc...) from the "experience" they sell to consumers.

In terms of HW, performance and features wise, nothing they sell is near the definition of cutting edge... they do focus on what the user directly interacts with (see MBP's keyboard, see the Mag-safe AC adapter plug, see their trackpads, their packaging, etc...).

Why would they need to be worried about ARM when they have gone to great lengths to separate HW from the applications (through a quite PC like abstraction layer... they kind of make it quite clear to devs that they can change CPU and GPU at any time if they so wish...)?

This move would make very little sense...

If Intel bought ARM to sabotage all other players, like Apple, and force them to use x86... anti-trust agencies all over the world would be on their butt quickly... and AMD or IBM don't have either the market position (AMD) or the interest (IBM) to become a an anti-competitive player in the mobile space.

Edited 2010-04-22 13:01 UTC

Reply Score: 1

RE: Comment by Panajev
by twitterfire on Thu 22nd Apr 2010 13:20 UTC in reply to "Comment by Panajev"
twitterfire Member since:
2008-09-11

Current Atoms are much more powerful than ARM but they draw much more power. I think that a Pentium 1 is enough to run Android, Meego, Maemo, Symbian or any other embedded OS. I think that Intel will need to cut some performance from the Atoms in order to emprove the battery life and make the CPU attractive for embedded devices. Something like an Pentium 1 based Atom on 22 nm will be very cheap, will have enough computing power and wouldn't draw much battery power.

Of course, when I say "Pentium 1" I don't mean the exact CPU design but something similar in simplicity of its design, computing power and features.

Reply Score: 2

RE[2]: Comment by Panajev
by viton on Thu 22nd Apr 2010 14:27 UTC in reply to "RE: Comment by Panajev"
viton Member since:
2005-08-09

Something like an Pentium 1 based Atom on 22 nm will be very cheap, will have enough computing power and wouldn't draw much battery power.

Do you really believe Intel guys didn't considered that?

Reply Score: 2

RE[2]: Comment by Panajev
by fatjoe on Thu 22nd Apr 2010 15:01 UTC in reply to "RE: Comment by Panajev"
fatjoe Member since:
2010-01-12

Something like an Pentium 1 based Atom on 22 nm will be very cheap, will have enough computing power and wouldn't draw much battery power.


Actually, no.

Att 22nm, the static CMOS leackage will be greater than the dynamic power. An ARM licensee told me just a few hours ago that they had moved from 90nm back to 0.18um for their ulrta low-power Cortex-M0 designs.

Reply Score: 1

Bye, bye ARM
by twitterfire on Thu 22nd Apr 2010 13:00 UTC
twitterfire
Member since:
2008-09-11

Bye, Bye Snapdragon, bye, bye Tegra 2, bye N900 and Droid. It was nice knowing you.

It's not a tragedy after all. If Apple wants to buy ARM, let'em have it.

I see some alternatives:

ARM licensees may have the right to develop new CPUs architectures based on ARM. Nvidia, for example, can do it and might do it.

CPU design isn't that much of a big problem as it used to be. The likes of Nvidia, IBM, TI, VIA, Intel can design new and maybe better architectures to replace ARM.

Nvidia has attracted a lot of ex Transmeta employees and there are rumors that Nvidia will soon enter X86 market. Maybe their next x86 CPU will be a low power one designed for embedded applications.

Reply Score: 3

RE: Bye, bye ARM
by Fettarme H-Milch on Thu 22nd Apr 2010 13:08 UTC in reply to "Bye, bye ARM"
Fettarme H-Milch Member since:
2010-02-16

there are rumors that Nvidia will soon enter X86 market

No rumors. Just wishful thinking.
If you look at the current licensing spat between NVidia and Intel, I doubt that NVidia can obtain a x86 license.
Nvidia could buy Via, but that would be suicide. Via and Intel have a cross-licensing deal, meaning buying Via would result in access of Intel to Nvidia's graphics patents.

Reply Score: 1

RE[2]: Bye, bye ARM
by Bill Shooter of Bul on Thu 22nd Apr 2010 14:17 UTC in reply to "RE: Bye, bye ARM"
Bill Shooter of Bul Member since:
2006-07-14

They also cross license with AMD which has ATI. I don't think that would be a problem, unless either there was a huge difference between the cross licensing language, or Nvidia had patents that were more applicable to intels graphic designs.

Reply Score: 2

RE: Bye, bye ARM
by Downix on Thu 22nd Apr 2010 15:12 UTC in reply to "Bye, bye ARM"
Downix Member since:
2007-08-21

There's more fish in the sea than ARM. MIPS, SPARC, both licensable, but currently skipped due to cost. if ARM is locked off, both of these will become options again, far more potent options.

Reply Score: 1

RE[2]: Bye, bye ARM
by bhtooefr on Fri 23rd Apr 2010 00:09 UTC in reply to "RE: Bye, bye ARM"
bhtooefr Member since:
2009-02-19

Well, MIPS-I without load/store instructions is fair game, for free, owning nobody anything, tested in court.

Alternately, SPARC is dangerous, but is free right now. SPARC International could change that, though.

Reply Score: 1

Rumor started by Intel
by Fettarme H-Milch on Thu 22nd Apr 2010 13:04 UTC
Fettarme H-Milch
Member since:
2010-02-16

I bet Intel started that rumor to spread FUD about ARM in order to increase interest in Moorestown Atom CPUs and future iterations.

Reply Score: 2

RE: Rumor started by Intel
by merkoth on Thu 22nd Apr 2010 13:29 UTC in reply to "Rumor started by Intel"
merkoth Member since:
2006-09-22

By rocketing ARM's share price? Nice move!

Reply Score: 2

RE[2]: Rumor started by Intel
by Fettarme H-Milch on Thu 22nd Apr 2010 14:16 UTC in reply to "RE: Rumor started by Intel"
Fettarme H-Milch Member since:
2010-02-16

By rocketing ARM's share price? Nice move!

What does ARM get from increased share price? Higher value of the company on paper, but not increased market share.
Such rumor-based share boosts drop after a few days, market share stays longer. If smartphone and embedded hardware makers adopt Atom instead of ARM, Intel gains more in the long term.

Reply Score: 1

RE[3]: Rumor started by Intel
by merkoth on Thu 22nd Apr 2010 17:53 UTC in reply to "RE[2]: Rumor started by Intel"
merkoth Member since:
2006-09-22

Such rumor-based share boosts drop after a few days, market share stays longer. If smartphone and embedded hardware makers adopt Atom instead of ARM, Intel gains more in the long term.


And market share does actually change in those few days the rumor lasts? So hardware manufacturers are dropping ARM-based projects just in case Apple might buy ARM, right? Maybe they start using PPC just in case Apple also decides to buy Intel.

Reply Score: 2

Fettarme H-Milch Member since:
2010-02-16

Some executives who only considered ARM may look into Atom, esp. as this rumor conveniently came into public shortly after Moorestown was released.

Reply Score: 1

RE: Rumor started by Intel
by vivainio on Thu 22nd Apr 2010 18:10 UTC in reply to "Rumor started by Intel"
vivainio Member since:
2008-12-26

I bet Intel started that rumor to spread FUD about ARM in order to increase interest in Moorestown Atom CPUs and future iterations.


I thought of the same thing as well. Nothing would please intel more than seeing their #1 competition in mobile market suddenly become irrelevant.

They'd absolutely make a killing.

Reply Score: 2

no way
by puenktchen on Thu 22nd Apr 2010 14:08 UTC
puenktchen
Member since:
2007-07-27

arm more or less has a monopoly on core chip design in the mobile space. there is no way buying arm to kick the competition out of business would and should be legal.

buying a strong minority share on the other hand would make sense.

Reply Score: 2

RE: no way
by fatjoe on Thu 22nd Apr 2010 14:52 UTC in reply to "no way"
fatjoe Member since:
2010-01-12

ARM licenses their IP to anyone, which includes Apple and Nokia. And they don't compete with their own costumers.


Now, if ARM is controlled by Apple, Apple will have an unfair advantage against Nokia and the rest. The EU would of course never allow that.

Reply Score: 1

Comment by Lazarus
by Lazarus on Thu 22nd Apr 2010 14:54 UTC
Lazarus
Member since:
2005-08-10

Bad mojo.

Reply Score: 2

Apple is NOT ARM's biggest customer
by fatjoe on Thu 22nd Apr 2010 15:14 UTC
fatjoe
Member since:
2010-01-12

The article is incorrect.

I think the largest ARM customer is either TI, Qualcomm, Samsung or Nokia. Given that Nokia ships far more handsets than anyone else [we are talking all types of phones, not only smartphones], I would say ARMs biggest customer is Nokia or whoever makes their SoC.

If they cant even get their facts right, it is fair to assume that the speculative part of the article is pure BS.

Edited 2010-04-22 15:15 UTC

Reply Score: 1

ariarinen Member since:
2009-02-07

The article is incorrect.

I think the largest ARM customer is either TI, Qualcomm, Samsung or Nokia. Given that Nokia ships far more handsets than anyone else [we are talking all types of phones, not only smartphones], I would say ARMs biggest customer is Nokia or whoever makes their SoC.

If they cant even get their facts right, it is fair to assume that the speculative part of the article is pure BS.
Well all of those are the largest ARM customers in one way or other.

Nokia has a few SoC suppliers like TI, Broadcom, ST Ericsson and Infineon ETC.
------

And current ARM licenses would never let a single company control ARM, they are too dependent and they would counter or overbid.

Reply Score: 1

Loongson
by 3rdalbum on Thu 22nd Apr 2010 15:22 UTC
3rdalbum
Member since:
2008-05-26

Assuming that such a buyout would be allowed (which I doubt as it allows Apple to cut off its competitors access to parts), it would be a godsend for the Chinese and their Loongson MIPS-compatible CPUs; you can already run Windows CE and Android on them so it wouldn't take too much to make a smartphone based around it.

Reply Score: 2

If this is true...
by Tuishimi on Thu 22nd Apr 2010 16:01 UTC
Tuishimi
Member since:
2005-07-06

...Then I wonder if Intel, Apple, Microsoft and even Google and Oracle execs play poker once/week and sit there scheming about how they can mess with the rest of us...

"Listen, Stevie... I had an idea. What if you bought ARM? Their stock would go up, our stock [Intel] would go up and we'd stick it to those cell-phone companies who think they are as cool as we are. It's an all-around win! Plus, I need a new yacht for Christmas. My old 150 footer is getting cramped."

Reply Score: 3

best thing if apple acquires
by bnolsen on Thu 22nd Apr 2010 17:15 UTC
bnolsen
Member since:
2006-01-06

It might make room for a 3rd player to step up to the table. Best of all worlds the third player would be a new innovative unencumbered cpu that could scale very nicely (and be cheap of course).

Here's to dreaming...

Reply Score: 2

I don't understand you all
by daveak on Thu 22nd Apr 2010 18:16 UTC
daveak
Member since:
2008-12-29

Why do all the comments say that if Apple went ahead and bought ARM that they would cripple ARM by restricting usage of their processor designs.

Apple used to own 40% of ARM as did Acorn and VLSI (? not sure on the last one, problem got letters in the wrong order or the wrong ones) the other 20%. They did not restrict usage then why would they do so now and cut off what makes ARM such a well performing company/stock?

Do they restrict or have their been reports of attempts to restrict usage of PowerVR GPUs at all given the large holding they have there? No.

Edited 2010-04-22 18:17 UTC

Reply Score: 1

RE: I don't understand you all
by alcibiades on Thu 22nd Apr 2010 18:46 UTC in reply to "I don't understand you all"
alcibiades Member since:
2005-10-12

Why do all the comments say that if Apple went ahead and bought ARM that they would cripple ARM by restricting usage of their processor designs......


I don't think that is the point. What this comes from is people wondering what on earth Apple would do to extract value from its purchase. At the moment existing shareholders of ARM are getting a certain return based on the market price of the shares. So Apple comes along and pays probably double to acquire the company. It must have some economic rationale for this. What could it be? There must be something Apple can do with ARM that it can only do if it owns it. What could this be?

Its at this point that people start speculating. Maybe Apple could stop anyone else buying ARM chips. Well, yes, maybe. You would reduce the value of your investment by a huge proportion. Maybe 75%+? Could it be worth it? Probably not. So its not that.

OK, you could ask for early releases on some chips. Is buying it the only way to get that? Probably not. Will it be worth it? Probably not. And anyway, to get it, you will have to annoy much larger customers of ARM than yourself. So its not that.

OK, what could it be? You are looking for something that can only be done with ARM if you own it, and it must also be something which buying it is the optimal and perhaps the only way to get it done. And you are looking also for something that is worth paying a premium over the market price as it was before the offer.

Maybe there is some direction of development which has so far proved invisible to the ARM management, but which Apple has discovered, and it will reveal it once it has done the purchase? Have we gone mad?

I don't believe it exists. Veterans of this stuff however will recognize that once the investment bankers are inside a company, social realities take over, and they have nothing to do with sensible business reasoning. The investment bankers start to build a coalition of those who expect to profit personally from the acquisition. All others get frozen out. This builds consensus, so called - that is, an alliance of those who have internal doubts about why we are doing this thing, but who say to themselves that its going to happen anyway, so lets do the best we can with it and position ourselves accordingly.

The investment bankers meanwhile are churning out spreadsheets and models showing how good a deal it will be. They are full of macros, very complicated, no-one understands them, they change all the time, they all show positive outcomes, and after all, these are very smart, well paid people, in lovely well fitting suits, hand made shirts and shoes. How can they be wrong?

This is why 90% of acquisitions fail. It is not that they are badly managed once done. Its that they never made any sense in the first place, but socially, it was impossible to stop the train.

Edited 2010-04-22 18:47 UTC

Reply Score: 4

In the real world
by Morty on Thu 22nd Apr 2010 18:54 UTC
Morty
Member since:
2005-07-06

I wish people could get the glamor of smartphones and Apple out of their eyes, and recognize what ARM is about.

The biggest competitors to ARM is not Atom, PowerPC, MIPS, Sparc or the VIAs. Their main competitors are from AVRs, PICs, ColdFires, Blackfins and S12s.

For every SOC or smarthphone there are hundreds of micro-controllers deployed, ARM are used in a huge and growing part of that market. The falling price on ARM based micro-controllers have more or less destroyed the market for 16bit controllers, and in some cases even become a valid alternative to some 8bit solutions.

Take any reasonably new laptop it's more likely than not, that it contains 2-3 ARMs or more.

Reply Score: 2

sums it up best here
by poundsmack on Thu 22nd Apr 2010 19:20 UTC
poundsmack
Member since:
2005-07-13

"...there are several reasons why Apple might not actually end up owning ARM.

For one thing, there are too many other influential companies – folks like Google (GOOG), LG, Marvell (MRVL), Nokia (NOK) and Samsung – who depend on ARM's technology to run their businesses, and they are sure to start a bidding war if they believe ARM is in play. Those companies might not have as much cash as Apple, but they can raise a stink.

And then there are the regulators. With so many companies threatened by the prospect of Apple buying ARM, some regulator somewhere would certainly argue that a sale would be bad for competition in the global mobile market. If they allowed Apple to buy ARM at all, I imagine they would demand assurances that ARM technology would continue to be available to Apple's competitors under decent licensing terms.

Bottom line: It's easy to see why Apple might want ARM, harder to imagine a deal going through. Can you imagine Steve Jobs owning the intellectual property behind the world's cell phone chips, and happily licensing it to competitors who are building phones with Google and Microsoft (MSFT)? Me either."

http://tech.fortune.cnn.com/2010/04/22/why-apple-likes-arm/?source=...

Reply Score: 2

ARM is overrated
by kjamc1982 on Thu 22nd Apr 2010 22:41 UTC
kjamc1982
Member since:
2007-05-09

I think Texas Instruments with the OMAP 4 platform is the truly next generation of mobile devices.

http://focus.ti.com/general/docs/wtbu/wtbuproductcontent.tsp?templa...

Reply Score: 1

RE: ARM is overrated
by bhtooefr on Fri 23rd Apr 2010 00:17 UTC in reply to "ARM is overrated"
bhtooefr Member since:
2009-02-19

*sigh*

Time to teach you how ARM works, and how SoCs work.

OMAP4 is a SoC, or System on Chip. A SoC contains one or more CPUs, all the I/O hardware, memory controller, various coprocessors that are needed, and sometimes (depending on purpose) a graphics controller. It's basically a complete chipset, just on one chip. (BTW, the difference between an SoC and a microcontroller is that SoCs tend to be more powerful, and have RAM and ROM off-chip, whereas microcontrollers have their RAM and ROM on the same die as the CPU and I/O hardware.

The CPU in the OMAP4 SoC is a dual-core ARM Cortex-A9 MPCore.

ARM doesn't make CPUs, they just design them, and license them to various companies at several levels - a company can license a design and use it as is, they can license a design and modify it, or they can license the architecture itself and create a new design based on that architecture.

In this case, TI licensed their Cortex-A9 MPCore for the OMAP4 SoC's CPU.

Reply Score: 2

RE: ARM is overrated
by Morty on Fri 23rd Apr 2010 00:42 UTC in reply to "ARM is overrated"
Morty Member since:
2005-07-06

I think Texas Instruments with the OMAP 4 platform is the truly next generation of mobile devices.

No disagreement the OPMAP 4 is a great processor platform, but so are also the NVIDIA Tegra 2. And more to the point, both are dual-core ARM Cortex-A9 based processors.

And I have no doubt that Marvell, Qualcomm, ST and Freescale have similar Cortex-A9 designs in the works. That is one of the things that make ARM so great. Another great thing is the scalability, from the high performance Cortex-A9 processors to the simpler cheap 20-100Mhz micro-controllers from about $1.

Reply Score: 2

Terrible for ARM and the mobile industry!
by Ravyne on Fri 23rd Apr 2010 01:39 UTC
Ravyne
Member since:
2006-01-08

I very much hope that this is merely a rumor.

As has been said, the only reason I would see Apple buying ARM outright is to keep everything for themselves. While owning all the great IP ARM has would make any mobile-oriented company happy, it would utterly destroy the current mobile-technology eco-system, which is 95% or more ARM-based. Even if they kept suppling non-competitors (ARM-based hard-disk controllers, automotive, embedded) and liscensed competitors on 'reasonable' terms, said competitors would still be contributing to Apple's bottom-line and essentially funding their R&D -- Competitors would look to jump ship, but find no truly-competetive alternative. The rest of the industry would loose years bringing up another competetive chip, and still probably loose money to Apple in liscensed ARM IP.

For ARM, I don't think this deal would be any good either -- Sure, Apple probably has enough money to buy them at a reasonable value now, but ARM is an increadibly prolific company with an incredibly bright future. I've said many times that ARM is the most viable competitor to intel and x86 that has ever existed, and I belive ARM will, with newer, more performance-oriented products, begin to encroach more and more into traditionally x86-based markets -- first with netbooks, nettops and STBs this year, thin-and-light laptops within 3-5 years, and will challenge in the server-deskop within 10-15 years. ARM is onto something, and while it may not be on the same magnitude, selling out now would be similar to the poor fool who sold his software to Bill Gates back in the day for a measly few thousand bucks.

Reply Score: 1

And ?
by yopmaster on Fri 23rd Apr 2010 08:35 UTC
yopmaster
Member since:
2009-10-28

No ARM in left town ? Then other makers will use another RISC low-power processor and ARM will be left behind. ARM is not successful because it's better, but only because they were at the right place at the right moment. Killing ARM will be a good news for MIPS though.
And by the way, can a company cut licenses it already sold to other (very big) companies ? Seriously...

Reply Score: 1