[AMD down 16.79%] ARM Boom Leaves AMD Out In The Cold!

[_nobody_]

On Friday, October 19th, 2012. AMD shares plunged 16.79%. It is NOT that INTEL is killing AMD, it is rather ARM.

http://finance.yahoo.com/q?s=AMD&ql=1

ARM started to dominate Compute Continuum market (tablets, smart phones). The key advantage: Open Source. And Samsung ARM11 (ARMv6), Nvidia Tegra designs (Cortex A9, ARMv7). Tegra2, and especially Tegra3. There are hundreds of thousands of Linux and Android developers out, porting/developing on ARM designs, literally speaking. With all free SW and tools, around the cyber corner.

The Doom and Gloom of MicroSoft based WIN silicon came to execution. INTEL is caught in crossfire of being ignorant... INTEL top management woke up, but NOT in Time to realize how big this threat is. AMD seems continues sleeping on their "invincible" Graphic designs!

I am an INTEL fan, so... For INTEL, it is time to mobilize all forces. INTEL should NOT miss this train. Should jump over AMD, and buy the competitor. To secure INTEL's back. Then, in adjacent thread, it should open all ATOM IPs, and make ATOM lookalike ARM. GPL and SW wise. The SW base is what is missing here. ATOM (Valley View) suppose to be better then Tegra3, in performance comparison.

http://www.anandtech.com/show/6203/...soc-emerge-22nm-atom-with-ivy-bridge-graphics

INTEL is heavily outnumbered by SW ARM developers. This is INTEL Achilles Hill. This is what should be taken care of. This is where main effort should focus. Spread the ATOM silicon, and Open Source knowledge of it!

All written here (as INTEL future actions and trends) IMHO!

_nobody_

 

[jackschmittusa]

AMD seems to be on a slide to the bottom. It will be a sad day when they are gone.

 

[cybrsage]

Intel cannot let them go away - it will mean they turn into a monopoly.

EDIT: I know it is not Intel's fault - but just looking at what our congressmen and legal "lets fine US companies" teams will see.

 

[Nintendesert]

Not sure I'm reading the OP correctly, but why would they buy AMD? Nothing to be gained there but debt. If Intel needed to buy someone to advance their ability to counter ARM it would be TI's chip division.

 

[thecoolnessrune]

Intel and AMD are American companies under American jurisdiction. I do not see a snowball's chance in hell that the SEC and other governing bodies would ever rubber stamp an AMD acquisition by Intel.

 

[Agent11]

The easy solution would be for intel or AMD to buy ARM. I read a bit ago that apple was thinking about it.

 

[dmcowen674]

AMD will be fine, Qualcomm is going to buy them out so there will continue to be high quality alternative CPU's to Intel and GPU's to Nvidia.

 

[Londo_Jowo]

I seriously doubt Qualcomm will acquire AMD as it wouldn't gain them a chip fabrication factory. The last I heard Samsung was interested in buying AMD.

 

[_nobody_]

http://www.eetimes.com/electronics-...e-web/4391415/AMD--who-will-buy-?pageNumber=0

Interesting article. Found it on the net, triggered by all these thoughts posted here.

Apple to buy ARM Foundation? What will be the price of ARM at this point in time? :whiste:

_nobody_

 

[theevilsharpie]

Apple to buy ARM Foundation? What will be the price of ARM at this point in time? :whiste:

Not gonna happen.

 

[Cerb]

The Doom and Gloom of MicroSoft based WIN silicon came to execution.

Way more runs on x86 than what Microsoft puts out, you know. "Wintel" died a quiet death, and was buried out back by the shed, OK? Intel has absolutely outstanding FOSS support, and has had for years. They've also been quite open when it comes to documenting their hardware.

The upcoming "Windows-only" Atoms are a very unique and unusual thing for Intel, these days.

INTEL is caught in crossfire of being ignorant... INTEL top management woke up, but NOT in Time to realize how big this threat is.

Bah! Intel knows very well how big of a threat ARM is. What they don't know is exactly how to compete, yet, since the ARM ecosystem functions in an entirely different way than theirs.

AMD has nothing they might want that they don't already have. AMD's problem is AMD's own, and they are threatening to pull a VIA, which will really suck for all of us.

INTEL is heavily outnumbered by SW ARM developers. This is INTEL Achilles Hill. This is what should be taken care of. This is where main effort should focus. Spread the ATOM silicon, and Open Source knowledge of it!

No. Most developers that couldn't care less whether what they write is going to run on ARM, MIPS, PPC, or x86. There are at least as many that know x86 like the back of their hand, too. X86 has a long embedded history, and a short but very strong workstation and server one. Along with those, there are those that specifically want to make sure it doesn't matter what it will run on. This is a complete non-issue, largely made up by press monkeys vying for ad revenue.

ARM's strength is in having unique hardware designs done by others using their chips, with fast time to market. When demands change, ARM vendors only need 1-2 years to adapt. This occurs with low margins all-around, since most companies making ARM hardware are integrating premade components that they license. Intel needs to find a way to fit in, while keeping their historically very high margins, and vertical integration that allows for that. That's their problem. It's got nothing to do with Android, nothing to do with Windows, and nothing to do with AMD.

 

[Geosurface]

I only ever had one AMD and the rest I've had have been Intel, well wait... I had a Packard Bell 386 in 1989, would that have had an Intel chip?

Anyway, I'm sad to see AMD go even though I didn't care for them personally.

It's crazy to think what would happen if Intel somehow went too, I mean... our civilization kind of requires that we have them, ha.

 

[unokitty]

AMD is Imploding because Management Doesn't Understand Semiconductors

While I would like to see AMD stay as competition for Intel, my perception is that the competition for the future is more likely to be ARM vs Intel X86. For example:

Is Qualcomm the New AMD? or the next Intel?

Uno

 

[Agent11]

Damn, I had no idea things were so bad at AMD. I Wasn't paying close enough attention I suppose.

 

[OBLAMA2009]

I seriously doubt Qualcomm will acquire AMD as it wouldn't gain them a chip fabrication factory. The last I heard Samsung was interested in buying AMD.

i dont think it will happen but i wish apple would buy amd.

 

[NoStateofMind]

Not sure I'm reading the OP correctly, but why would they buy AMD? Nothing to be gained there but debt. If Intel needed to buy someone to advance their ability to counter ARM it would be TI's chip division.

GPU architecture. Even you would agree after their previous attempts. Larrabee anyone?

^^thats just ONE reason.

 

[ericlp]

AMD will never go under ... If anything ATI will hold them afloat for some time to come. My opinion. After all they are making some of the best graphics hardware available today.

 

[_nobody_]

The upcoming "Windows-only" Atoms are a very unique and unusual thing for Intel, these days.

Yes, I agree with this statement. But still, Win is written (as OS) x86-centric. So, as Valey View (VV) comes to play: http://www.anandtech.com/show/6203/...soc-emerge-22nm-atom-with-ivy-bridge-graphics

No ARM silicon can match VV performance. VV suppose to be the next ATOM vehicle carrying tablets, as my best assumption is. The problem is that VV is comming (Silvermont core, according to this article) too late: Q4 2013.

VV must come much faster: Q1 2013, at most Q2 2013.

Intel knows very well how big of a threat ARM is. What they don't know is exactly how to compete, yet, since the ARM ecosystem functions in an entirely different way than theirs.

If you could not win them, join them (strategy wise)! And INTEL does NOT join them.

No. Most developers that couldn't care less whether what they write is going to run on ARM, MIPS, PPC, or x86. There are at least as many that know x86 like the back of their hand, too. X86 has a long embedded history, and a short but very strong workstation and server one. Along with those, there are those that specifically want to make sure it doesn't matter what it will run on. This is a complete non-issue, largely made up by press monkeys vying for ad revenue.

I agree with you in some business segments. Do NOT forget, server space is NOT embedded space. INTEL put lot of effort for the Open Source community there (Linux upgrading x86 architecture and keeping it up to date), since most servers in The World run on Linux. Do not know about the WSs, could not comment on this.

We are talking here complete opposite Compute Continuum End! About smart phones and tablets! Only in 2012 there are approximately 120 million tablets sold. 60+ million less laptops sold, which is killing INTEL, and AMD.

ARM's strength is in having unique hardware designs done by others using their chips, with fast time to market. When demands change, ARM vendors only need 1-2 years to adapt. This occurs with low margins all-around, since most companies making ARM hardware are integrating premade components that they license. Intel needs to find a way to fit in, while keeping their historically very high margins, and vertical integration that allows for that. That's their problem. It's got nothing to do with Android, nothing to do with Windows, and nothing to do with AMD.

Here we have exactly what I was talking about. We have hundreds of thousands ARM HW developers, you have corrected me there. There are many companies making their own ARM creations for the smart phone and tablet markets. Qualcomm with Snapdragon. Samsung is using these in their most successful Galaxy phones, while Samsung entirely sells its own ARM silicon to Apple (Apple has Samsuns as entirely one provider of silicon) for iPhones and iPads!! Do not forget TI with their OMAP program which already runs for 10 years. Nvidia with TEGRA line of products! :whiste:

How INTEL can compete with ARM, if INTEL puts out VV, and it expects to wrap up in Q4, 2013!? And it puts closed IP as pre-bootloader for entire line of ATOM?

Do NOT forget, smart phones and tablets do NOT run BIOS. They run u-boot/coreboot bootloaders, all Source Code driven for ARM.

From this point, the desperate move will be an excellent one: Samsung or Apple to buy AMD and adopt AMD's x86 (for smart phones, tablets) the same way as ARM! And yes, they know how to do it!

_nobody_

 

[SparkyJJO]

i dont think it will happen but i wish apple would buy amd.

HECK no.

 

[mchammer187]

The easy solution would be for intel or AMD to buy ARM. I read a bit ago that apple was thinking about it.

ARM is 10X the size of AMD by market cap

 

[rchiu]

Yes, I agree with this statement. But still, Win is written (as OS) x86-centric. So, as Valey View (VV) comes to play: http://www.anandtech.com/show/6203/...soc-emerge-22nm-atom-with-ivy-bridge-graphics

No ARM silicon can match VV performance. VV suppose to be the next ATOM vehicle carrying tablets, as my best assumption is. The problem is that VV is comming (Silvermont core, according to this article) too late: Q4 2013.

VV must come much faster: Q1 2013, at most Q2 2013.



If you could not win them, join them (strategy wise)! And INTEL does NOT join them.



I agree with you in some business segments. Do NOT forget, server space is NOT embedded space. INTEL put lot of effort for the Open Source community there (Linux upgrading x86 architecture and keeping it up to date), since most servers in The World run on Linux. Do not know about the WSs, could not comment on this.

We are talking here complete opposite Compute Continuum End! About smart phones and tablets! Only in 2012 there are approximately 120 million tablets sold. 60+ million less laptops sold, which is killing INTEL, and AMD.



Here we have exactly what I was talking about. We have hundreds of thousands ARM HW developers, you have corrected me there. There are many companies making their own ARM creations for the smart phone and tablet markets. Qualcomm with Snapdragon. Samsung is using these in their most successful Galaxy phones, while Samsung entirely sells its own ARM silicon to Apple (Apple has Samsuns as entirely one provider of silicon) for iPhones and iPads!! Do not forget TI with their OMAP program which already runs for 10 years. Nvidia with TEGRA line of products! :whiste:

How INTEL can compete with ARM, if INTEL puts out VV, and it expects to wrap up in Q4, 2013!? And it puts closed IP as pre-bootloader for entire line of ATOM?

Do NOT forget, smart phones and tablets do NOT run BIOS. They run u-boot/coreboot bootloaders, all Source Code driven for ARM.

From this point, the desperate move will be an excellent one: Samsung or Apple to buy AMD and adopt AMD's x86 (for smart phones, tablets) the same way as ARM! And yes, they know how to do it!

_nobody_

I know ARM is all the rage in the chip industry now. But in my opinion, AMD is too late in the game and lack the R&D resource to compete. Instead of diverting their attention to a new market they will for sure getting their butt kicked, might as well focus on their core market, even if their core market is not growing too much.

Yes laptop sales has been replaced by tablet sales. But with win8 coming up, x86 may start to gain some ground in the tablet space. In addition, AMD should focus on corporate PC and server. For corporate PC, utilize their integrate graphic platform to offer low cost, low energy, reasonably fast integrated graphic chip for increasing media rich corporate desktop requirement. For server, yes AMD is lagging badly. But if they focus on what customers want - lower energy operation, better multi-core capability for virtualization, they may win some points even if their lose in terms of pure clock speed.

Lastly, with AMD's integrated graphic chip, console is definitely an area they should focus in winning.

Bottomline, instead of blindly following the analyst and hot industry trend, AMD should really look at their strength, and which market they have the best chance to win, meaning continue to focus in the x86 space and utilize their strength with ATI technology to develop lower cost, lower energy, better performing chips (in terms of cpu+integrated graphic one chip solution). You pick the battle you can win, instead of going for glory and know you will get massacred.

 

[Cerb]

Yes, I agree with this statement. But still, Win is written (as OS) x86-centric. So, as Valey View (VV) comes to play: http://www.anandtech.com/show/6203/...soc-emerge-22nm-atom-with-ivy-bridge-graphics

No, Windows' desktop and applications are x86-centric. It runs on ARM fine. It runs on Power fine. It has run on Alpha, MIPS, and SH fine, too. Also, MS has been working to free the user-land bits of their x86 ties, and will have completely done so by the next version of Windows. While likely legal (since both companies know they're being watched), there is surely a less than savory deal behind the "Windows only" Atoms.

No ARM silicon can match VV performance. VV suppose to be the next ATOM vehicle carrying tablets, as my best assumption is. The problem is that VV is comming (Silvermont core, according to this article) too late: Q4 2013.

No need. If ARM keeps making cores significantly faster than their previous ones, they'll be fine. X86 needs high performance more than ARM does, because without it, it will not remain different enough.

VV must come much faster: Q1 2013, at most Q2 2013.

It will come when it is good and ready (cheap enough per chip).

If you could not win them, join them (strategy wise)! And INTEL does NOT join them.

So, if joining doesn't work, but people don't care, then what? One of Intel's weaknesses is that most people really don't care about the performance of their portable devices, at this point. They can undercut them, but that would mean less profit.

I agree with you in some business segments. Do NOT forget, server space is NOT embedded space. INTEL put lot of effort for the Open Source community there (Linux upgrading x86 architecture and keeping it up to date), since most servers in The World run on Linux. Do not know about the WSs, could not comment on this.

There's no great cabal of ARM developers, or x86 developers, or any other kind. There is, at most, small pockets of those people. As for servers, ARM is non-existent, there, and even if the soon-to-come ARMV8 CPUs are pretty good, will only gain small niches. It's not that server == embedded, but that Intel's general openness, wrt to hardware docs, has a long history, where their server business, in its entirety, does not.

We are talking here complete opposite Compute Continuum End! About smart phones and tablets! Only in 2012 there are approximately 120 million tablets sold. 60+ million less laptops sold, which is killing INTEL, and AMD.

Really? I could have sworn Ruiz and friends did most of that to AMD, Meyer barely got them in the black, and now Read isn't helping, in the short term. Intel is doing quite well, ATM.

AMD needs higher ASPs, and Intel wants higher ASPs, but if it doesn't have an Apple logo, people would prefer to give up features and performance instead of paying higher ASPs. That means keeping computers longer, and buying cheaper new ones.

If people not buying new items because their old ones are still good enough is enough to sink a company, and they'd rather buy the cheaper ones, today, that company deserves to be sunk. People driving used cars, and mostly buying smaller cars, hasn't sunk the car companies. They wanted to blame those kinds of forces, but it mostly mismanagement, and wanting to pretend the market was doing things that it wasn't.

IoW, you're reading tech and financial press too much.

Here we have exactly what I was talking about. We have hundreds of thousands ARM HW developers, you have corrected me there. There are many companies making their own ARM creations for the smart phone and tablet markets. Qualcomm with Snapdragon. Samsung is using these in their most successful Galaxy phones, while Samsung entirely sells its own ARM silicon to Apple (Apple has Samsuns as entirely one provider of silicon) for iPhones and iPads!! Do not forget TI with their OMAP program which already runs for 10 years. Nvidia with TEGRA line of products! :whiste:

OMAP is going away, and really a little later than it should have. Samsung is also not selling its ARM silicon to Apple. It is selling chips to Apple that are Apple's own designs.

How INTEL can compete with ARM, if INTEL puts out VV, and it expects to wrap up in Q4, 2013!? And it puts closed IP as pre-bootloader for entire line of ATOM?

What quarter it comes out in really only matters if OEMs are waiting for chips, thinking it would be earlier. Their competition with ARM is not short-term, like that. They need to keep x86 relevant, with high priced parts, in an era where people are gravitating towards lower cost, and caring less about x86's software inertia each year. It's a 5-10 year problem for Intel, not a quarterly one. They have time to figure something out, or brute-force their way in until they can get x86 inertia in portable devices.

Do NOT forget, smart phones and tablets do NOT run BIOS. They run u-boot/coreboot bootloaders, all Source Code driven for ARM.

Nothing but IBM PC compatible computers and emulators use BIOS. That really has nothing to do with any of this. An x86 tablet, with no need for IBM PC compatibility, might also not have any BIOS support (it's not like it will get random cards plugged into it that add boot-time features through BIOS).

From this point, the desperate move will be an excellent one: Samsung or Apple to buy AMD and adopt AMD's x86 (for smart phones, tablets) the same way as ARM! And yes, they know how to do it!

_nobody_

Only if their BOD gets really drunk, one night. Aside from maybe the Radeon patent portfolio and engineers, nobody wants anything AMD has but AMD itself. If they want faster CPUs, they could afford to design faster ARM CPUs. X86 is a can of worms that nobody in their right mind wants to gain access to.

 

[Genx87]

AMD will be fine, Qualcomm is going to buy them out so there will continue to be high quality alternative CPU's to Intel and GPU's to Nvidia.

Qualcomm already bought the most important part of AMD in 2009 for a pathetic 65 million. It was their mobile divison which included snapdragon. What else does AMD have now besides a mountain of debt and little to no growth potential? May as well wait for the bankruptcy auction to acquire their IP without the debt.

 

[_nobody_]

So, if joining doesn't work, but people don't care, then what? One of Intel's weaknesses is that most people really don't care about the performance of their portable devices, at this point. They can undercut them, but that would mean less profit.

[snap]

AMD needs higher ASPs, and Intel wants higher ASPs, but if it doesn't have an Apple logo, people would prefer to give up features and performance instead of paying higher ASPs. That means keeping computers longer, and buying cheaper new ones.

If INTEL wants foot into the tablet/smart phone market, INTEL has to undercut price of the ATOM silicon and get to the low margins. Even if this does mean (and this does mean) that there is NO profit. Not for one Q, for several adjacent ones!

If people not buying new items because their old ones are still good enough is enough to sink a company, and they'd rather buy the cheaper ones, today, that company deserves to be sunk. People driving used cars, and mostly buying smaller cars, hasn't sunk the car companies. They wanted to blame those kinds of forces, but it mostly mismanagement, and wanting to pretend the market was doing things that it wasn't.

The simple single word for this: hyperproduction! This is the huge part of the problem Western markets are facing, Present Time. Emerging markets are the targets, Samsung already dominates Africa.

IoW, you're reading tech and financial press too much.

Justification below.

OMAP is going away, and really a little later than it should have.

Amazon buying TI OMAP division:
http://seekingalpha.com/article/928401-amazon-buying-texas-instruments-omap-division-strategic-or-nonsensical

Samsung is also not selling its ARM silicon to Apple. It is selling chips to Apple that are Apple's own designs.

Thank you for correcting me. Apple is fabless, using its own designs, and the sole manufacturer for Apple is Samsung. This will change in The Future.

What quarter it comes out in really only matters if OEMs are waiting for chips, thinking it would be earlier. Their competition with ARM is not short-term, like that. They need to keep x86 relevant, with high priced parts, in an era where people are gravitating towards lower cost, and caring less about x86's software inertia each year. It's a 5-10 year problem for Intel, not a quarterly one. They have time to figure something out, or brute-force their way in until they can get x86 inertia in portable devices.

INTEL, as huge company with lot of monies, can afford to loose monies in order to brute-force their own ATOM line of products until it gets some inertia. Example: Motorola RAZR, based upon ATOM design! It now gets some traction!

Only if their BOD gets really drunk, one night. Aside from maybe the Radeon patent portfolio and engineers, nobody wants anything AMD has but AMD itself. If they want faster CPUs, they could afford to design faster ARM CPUs. X86 is a can of worms that nobody in their right mind wants to gain access to.

Yup, seems that you get this one right! These are Radeon IP properties which are worth some significant monies. http://en.wikipedia.org/wiki/Radeon

But dunno forget: AMD is in unique position in the industry trying to bring both High Performance CPU and GPUs, all together. It is worth some cash in hand, to preserve these teams intact.

_nobody_

 

[_nobody_]

AMD Plans to Use ARM Designs in Chips for Servers in 2014
http://www.bloomberg.com/news/2012-...g-cash-crunch-amid-quest-for-new-markets.html

Interesting information. Seems that AMD found itself in unchartered territories, and AMD is trying to pull rabit out of the hat, playing on server market, and betting on ARM support for survival.

Not a bad strategy, since Read is preparing to lift AMD share price, and to sell AMD to strong positioned ARM based company (such as Apple, Google, or even Samsung). :whiste:

_nobody_