[allthingsd.com] AMD getting ready for another round of Layoffs

[Zstream]

ARM will purchase AMD which will = New Company

Anyone want to take a wager?

 

[ShintaiDK]

ARM will purchase AMD which will = New Company

Anyone want to take a wager?

Why would the little ARM buy AMD?
http://en.wikipedia.org/wiki/ARM_Holdings

Its simply not their business model. And AMD got nothing they want. Its much cheaper to pick whatever you like from the dead. Not to mention ARM doesnt exactly float in money. And buying AMD means no x86 anyway. meaning layoffs thats just extra expenses. Plus all the IP that relates to x86 that cant be used. So what would they need AMD for again?

 

[Keysplayr]

Why would the little ARM buy AMD?
http://en.wikipedia.org/wiki/ARM_Holdings

Its simply not their business model. And AMD got nothing they want. Its much cheaper to pick whatever you like from the dead. Not to mention ARM doesnt exactly float in money. And buying AMD means no x86 anyway. meaning layoffs thats just extra expenses. Plus all the IP that relates to x86 that cant be used. So what would they need AMD for again?

"Little" ARM has a market cap of almost 13 Billion.
http://www.nasdaq.com/symbol/armh

 

[ViRGE]

"Little" ARM has a market cap of almost 13 Billion.
http://www.nasdaq.com/symbol/armh

However ARM has revenue under 500mil/year, and a P/E ratio of 56. From a financial standpoint ARM buying AMD would be equal to if not more insane than AMD buying ATI. It would take a far larger company to properly digest AMD.

 

[NIGELG]

There is nothing AMD got that Intel not already got access to and want.

GPU'S....Intel GPU'S are crap and Larabbee was crap that never saw the light of day.

Intel needs to be shown how to make GPU'S.

You presume to speak for Intel?How could you know what they want?

 

[ShintaiDK]

"Little" ARM has a market cap of almost 13 Billion.
http://www.nasdaq.com/symbol/armh

Thats your only argument?

Its a little company with a revenue of around 500mio(150mio profit) and 2000 employees. And they dont manufactor a single chip themselves.

They dont exactly got alot of cash to buy AMD for either:
http://www.google.com/finance?q=NASDAQ%3AARMH&fstype=ii&ei=xrx6UIjfF8WYwQOC8QE

The company grew ~75% the last 5 years.

So, why do they want AMD again?

 

[ShintaiDK]

GPU'S....Intel GPU'S are crap and Larabbee was crap that never saw the light of day.

Intel needs to be shown how to make GPU'S.

You presume to speak for Intel?How could you know what they want?

If I couldnt know, then how do you know?

Larabee was an experiment to get everything in a single system. Basicly what the GPU+CPU will never accomplish.

Intel GPUs aint crap. You just mix up what target groups they got. They work perfect for the target audience. And AMDs lack of ability to dominate with APUs shows that.

AMDs GPUs in the APUs are horrible too when compared to my GTX680. But APUs are still good enough for 70-80% of the worlds PC users, just like Intels.

Also buying a company is more than just buying it. It needs to fit into the corporate structure. Something AMD got a hard time with ATI. Intel buying nVidia for example would be directly insane due to one roughly be suits and rulebook people and the other something equal to hippies.

 

[Keysplayr]

Thats your only argument?

Its a little company with a revenue of around 500mio(150mio profit) and 2000 employees. And they dont manufactor a single chip themselves.

They dont exactly got alot of cash to buy AMD for either:
http://www.google.com/finance?q=NASDAQ%3AARMH&fstype=ii&ei=xrx6UIjfF8WYwQOC8QE

The company grew ~75% the last 5 years.

So, why do they want AMD again?

What argument? I'm saying they're not little.
And where in the universe did you see me saying ARM "wanted" AMD?
OP said that. Not I.

 

[NIGELG]

If I couldnt know, then how do you know?

Larabee was an experiment to get everything in a single system. Basicly what the GPU+CPU will never accomplish.

Intel GPUs aint crap. You just mix up what target groups they got. They work perfect for the target audience. And AMDs lack of ability to dominate with APUs shows that.

AMDs GPUs in the APUs are horrible too when compared to my GTX680. But APUs are still good enough for 70-80% of the worlds PC users, just like Intels.

Also buying a company is more than just buying it. It needs to fit into the corporate structure. Something AMD got a hard time with ATI. Intel buying nVidia for example would be directly insane due to one roughly be suits and rulebook people and the other something equal to hippies.

I didn't say I know what Intel wants.It's you who presumes to know what Intel wants.

I'm merely saying that AMD GPU's are a POSSIBLE reason for them to want something AMD has.....

 

[Keysplayr]

I didn't say I know what Intel wants.It's you who presumes to know what Intel wants.

I'm merely saying that AMD GPU's are a POSSIBLE reason for them to want something AMD has.....

Intel has been improving their HD GPUs fairly quickly over the last few gens.
I don't think AMD GPUs are as enticing to them as you may think. Although ANYTHING is possible.

 

[NIGELG]

Intel has been improving their HD GPUs fairly quickly over the last few gens.
I don't think AMD GPUs are as enticing to them as you may think. Although ANYTHING is possible.

For anything but low resolution or retro gaming Intel can't compete with anything AMD has in their GPU line up.

If they want to be a force in PC gaming it might be a good idea to take over the GPU division only.

Unless you think most PC gamers are Farmville type players only.

 

[Keysplayr]

For anything but low resolution or retro gaming Intel can't compete with anything AMD has in their GPU line up.

If they want to be a force in PC gaming it might be a good idea to take over the GPU division only.

Unless you think most PC gamers are Farmville type players only.

They don't have to be a force in PC gaming. They are the biggest force in integrated graphics by far. If you only want Intel to buy the "ATI" division from AMD just for the sake of "saving ATI", well, that's not likely. They don't need to.
http://www.techpowerup.com/162605/H...kage-Cache-to-Boost-Graphics-Performance.html

http://hothardware.com/Reviews/IDF-Day-1-Haswell-Live-Graphics-Demo/

 

[NIGELG]

They don't have to be a force in PC gaming. They are the biggest force in integrated graphics by far. If you only want Intel to buy the "ATI" division from AMD just for the sake of "saving ATI", well, that's not likely. They don't need to.
http://www.techpowerup.com/162605/H...kage-Cache-to-Boost-Graphics-Performance.html

They don't have to...but they can.Nvidia needs competition unless you would rather have Nvidia alone making high end graphics cards.

Can Haswell play the latest games at 1080 p??Can they put out anything integrated to compete with a 7970??

Are you saying all that AMD GPU tech is useless and unwanted??

 

[Keysplayr]

They don't have to...but they can.Nvidia needs competition unless you would rather have Nvidia alone making high end graphics cards.

Can Haswell play the latest games at 1080 p??Can they put out anything integrated to compete with a 7970??

Are you saying all that AMD GPU tech is useless and unwanted??

NIGEL, if everybody wanted AMD's tech, they wouldn't be in the predicament they are currently in.

 

[railven]

NIGEL, if everybody wanted AMD's tech, they wouldn't be in the predicament they are currently in. Wouldn't you agree?

I don't know if Haswell can play the latest games at 1080p. Maybe it can, maybe it can't. The point is, at the very least they are twice as fast as their 3rd gen series of HD. That is pretty huge increase in such a short amount of time. And, I feel it's only a short matter of time before Intel equals the GPU performance of what AMD is currently offering in their APU's.

I think you are missing what he is saying, while I agree with what you're saying (Intel doesn't have to.)

As an old ATI user, I would also like if Intel managed to wrestle the GPU portfolio from AMD (not just to make better iGPUs) but to create an Intel/ATI branded high end GPU that can compete with discrete options (again not just iGPUs which is the point you keep missing.)

Intel with proper GPU tech can be a damn good competitor.

 

[NIGELG]

NIGEL, if everybody wanted AMD's tech, they wouldn't be in the predicament they are currently in.

GPU tech,Sir.GPU tech.

As hard as this is for you to grasp their GPU'S are very desirable to 40% of us.

Why would one ignore such a market?

 

[railven]

GPU tech,Sir.GPU tech.

As hard as this is for you to grasp their GPU'S are very desirable to 40% of us.

Why would one ignore such a market?

Unfortunately, that 40% of a market is what? ~10-15% of the whole market?

I can understand why people argue Intel doesn't have to do it, or even think about doing it.

 

[NIGELG]

Unfortunately, that 40% of a market is what? ~10-15% of the whole market?

I can understand why people argue Intel doesn't have to do it, or even think about doing it.

Having Nvidia cards and Intel CPU'S as the only choices is not something I look forward to.

 

[railven]

Having Nvidia cards and Intel CPU'S as the only choices is not something I look forward to.

Agree, I also prefer options, however, free enterpise market dictates the rules. We can't ignore sale trends, out side of AMD's own incompetence, the markets seemed to favor Intel + nVidia.

We reep what we sow.

 

[jones377]

I don't see a reason for any company to buy AMD outright and then continue to compete with Intel in the X86 market. It's losing proposition. However, should AMD go into bankruptcy, I could see them getting bought out.

Best case is that AMD would be bought out and continue their current products with infused capital strength. But this is probably just a pipe dream and not very likely. Possible candidate could be ATIC or a similar investment company.

Worst case scenario but also the most likely IMHO would be AMD getting chopped into pieces and then sold to various companies. I'm sure there are a number of companies that would love to get their hands on the engineers and patents that AMD currently holds.

Samsung could pickup the CPU division and set them to work designing their own custom ARM core. EDIT: Even ARM itself might be interested as it could accelerate their efforts to design a more powerful ARM CPU intended for servers and laptops.

The GPU division and the Radeon line might not be very interesting because discrete GPUs are a declining market. Without a CPU to integrate into it doesn't have enough viable long term prospects. Buying the division in Canada and then repurpose them for designing low power GPUs intended for SOCs is a possibility but this market is crowded already as it is. Apple could be a candidate here however. Samsung a lesser possibility.

Intel would likely be blocked by the government from buying any patents from AMD. They would surely be interested in getting both hardware and software engineers (GPU drivers) through normal hirings as Intel and AMD are operating in several common areas.

AMD's x86 patents could be picked up by Nvidia, although I doubt they would be able to negotiate a good enough deal with Intel that would let them produce their own x86 CPU and still make a profit. Plus Nvidia probably closed the door on x86 for good by now. It's more likely getting bought by a patent troll company and then used unsuccessfully to extort Intel for license fees.

A number of companies (mostly the SOC vendors) would fight over leftover engineers and patents to strengthen their own portfolios.

It's not a pretty future for AMD but a real possibility within 5 years or so unless AMD can really turn things around ASAP.

 

[Idontcare]

GPU tech,Sir.GPU tech.

As hard as this is for you to grasp their GPU'S are very desirable to 40% of us.

Why would one ignore such a market?

Let's think about this from a forest-for-the-trees perspective.

Intel's revenue in 2011 was $54B.

AMD's GPU revenue in 2011 was $1.56B (which was down from $1.66B in 2010).

So from Intel's perspective, by not having AMD's GPU IP and the ability to make GPU's with that IP as AMD does, that lack of capability on Intel's part is at most causing them to lose access to $1.56B in annual revenue.

AMD's entire graphics division pulls in less than 3% of Intel's annual revenue.

Let that sink in for a moment. Intel's decision makers are not going to stop and give 5 minutes of their time to even contemplate doing something that at most stands to impact their revenue by a mere 3%. Those guys are wrestling with ridiculously bigger fish to fry.

They need to find growth engines for their $54B company that will take it to $60B, then $65B, then $75B, etc. Sustained growth that exceeds global inflationary values. They are not going to find that by going after a market that at best represents a mere 3% growth in revenue.

They don't need it to hit their current revenue targets, as their revenue numbers would indicate given that they have those revenue numbers without the benefit of AMD's GPU IP, and even if they had it they still aren't going to be anywhere close to meeting the annual growth numbers they need to deliver so as to justify them having their jobs in the first place.

Big picture. Yes GPU's mean the world to us, but we are a decimal place in the big picture of where revenue is to be found in the global semiconductor TAM.

 

[railven]

Let's think about this from a forest-for-the-trees perspective.

Intel's revenue in 2011 was $54B.

AMD's GPU revenue in 2011 was $1.56B (which was down from $1.66B in 2010).

So from Intel's perspective, by not having AMD's GPU IP and the ability to make GPU's with that IP as AMD does, that lack of capability on Intel's part is at most causing them to lose access to $1.56B in annual revenue.

AMD's entire graphics division pulls in less than 3% of Intel's annual revenue.

Let that sink in for a moment. Intel's decision makers are not going to stop and give 5 minutes of their time to even contemplate doing something that at most stands to impact their revenue by a mere 3%. Those guys are wrestling with ridiculously bigger fish to fry.

They need to find growth engines for their $54B company that will take it to $60B, then $65B, then $75B, etc. Sustained growth that exceeds global inflationary values. They are not going to find that by going after a market that at best represents a mere 3% growth in revenue.

They don't need it to hit their current revenue targets, as their revenue numbers would indicate given that they have those revenue numbers without the benefit of AMD's GPU IP, and even if they had it they still aren't going to be anywhere close to meeting the annual growth numbers they need to deliver so as to justify them having their jobs in the first place.

Big picture. Yes GPU's mean the world to us, but we are a decimal place in the big picture of where revenue is to be found in the global semiconductor TAM.

Great post, but to counter this, what is nVidia's GPU revenues? AMD is a floundering mess and their technology is not earning any where near what it could possibly earn (due to of course AMD being run by a bunch of morons, it seems.)

Change management and there is a possibility that that revenue can grow, perhaps not match nVidia, but as you said - growth. A 3% revenue boost with Intel at the helm can grow to 5%, 10%? It's no longer small fries.

Because of the abysmal situation AMD is in, this could be a great chance for growth in a sector that Intel has no presence (Discrete GPU) at a cheap investment (not the $5B ATI cost in it's hayday, hell if they can pull it off for $500-750K, why not?)

 

[Idontcare]

Great post, but to counter this, what is nVidia's GPU revenues? AMD is a floundering mess and their technology is not earning any where near what it could possibly earn (due to of course AMD being run by a bunch of morons, it seems.)

Change management and there is a possibility that that revenue can grow, perhaps not match nVidia, but as you said - growth. A 3% revenue boost with Intel at the helm can grow to 5%, 10%? It's no longer small fries.

Because of the abysmal situation AMD is in, this could be a great chance for growth in a sector that Intel has no presence (Discrete GPU) at a cheap investment (not the $5B ATI cost in it's hayday, hell if they can pull it off for $500-750K, why not?)

The growth in TAM for discrete GPU is not in gaming or graphics, AMD's revenue there declined 2010 -> 2011.

The growth in TAM for GPU's is all in HPC (high performance computing) which is why you see Nvidia's all-out push for CUDA and GPGPU, likewise that is why you see Intel repurposing their failed discrete GPU effort (Larrabee) as a purely HPC-targeted product (XEON Phi)...because that IS where the revenue growth in GPU is going to come in the forthcoming decade.

This is why it is safe to conclude Intel probably is not interested in acquiring AMD's GPU IP - for what Intel is after, high margin revenue, they don't need to get mired in the low-margin market that is discrete GPU.

 

[Ben90]

AMD's CEO position was apparently so undesirable that they took forever trying to find a permanent CEO and came up with ...Rory Read?!?!

Intel - Paul Otellini, Bachelors Economics, USF; MBA UC Berkeley (Haas school of business)

NVidia - Jen-Hsun Huang, BS (EE) Oregon State University, MS (EE) Stanford University

AMD - Rory Read, bachelor's degree in Information Systems from Hartwick College

One of these guys doesn't belong.

Education background is probably the least useful statistic to measure a CEO's performance. Realworld experience managing from all levels is much much more important, even if they didn't complete highschool.

 

[railven]

The growth in TAM for discrete GPU is not in gaming or graphics, AMD's revenue there declined 2010 -> 2011.

The growth in TAM for GPU's is all in HPC (high performance computing) which is why you see Nvidia's all-out push for CUDA and GPGPU, likewise that is why you see Intel repurposing their failed discrete GPU effort (Larrabee) as a purely HPC-targeted product (XEON Phi)...because that IS where the revenue growth in GPU is going to come in the forthcoming decade.

This is why it is safe to conclude Intel probably is not interested in acquiring AMD's GPU IP - for what Intel is after, high margin revenue, they don't need to get mired in the low-margin market that is discrete GPU.

I understand that, and my suggestion was more of a "bolted on" kind of win. They could acquire AMD's GPU technology on the cheap (which *should* include their GNC compute technology) and perhaps with their better management team create a counter to the CUDA Ecosystem that creates nVidia's giant revenues. Intel is trying to compete with nVidia in this sector already, while they do that they get a little push into discrete GPUs as well.

While there is no part of me that thinks Intel would ever go for it, I'm thinking out loud of "what ifs." If the patents were put on auction (or private sale) for a fraction of their original price - I think every corporation would love for a chance at them, Intel would be no different.