AMD shut out of the debt market

[Centauri]

..
http://forums.anandtech.com/showthread.php?t=2283777&highlight=

the stock action is in the short position ,

but eventually AMD will die/merge and assets reassigned/sold ...

If you ever make a book of your forum posts, count me in for none.

 

[erunion]

x86 license is non transferable. And its value is between minimal to nonexistant. Other companies also still got x86 licenses. Its simply not worth making the CPUs in terms of competition.

Then there is all the debt and liabilities in AMD on top.

True, x86 license is not in demand. A few years ago Nvidia would have tried to get it. And maybe they are still interested, but I think they have settled plans for ARM now. (About non transferable, Intel has had their terms forcibly altered by the courts more than a few times.)

What blows my mind is that not long ago AMD paid $5 Billion for ATI and now their total liquidation value is under $700 million. Talk about destroying value.

 

[pablo87]

I been thinking about this and that agency doesn't blame the WSA and the deal with Global Foundries directly as the source of AMD's problems but reading between the lines...I wonder if Mubadala is a customer of theirs...

AMD is trapped by the WSA, the only way out is via Global Foundries. AMD is West Berlin pre 1971, only the Airport is owned by the USSR.:twisted:

 

[Idontcare]

I been thinking about this and that agency doesn't blame the WSA and the deal with Global Foundries directly as the source of AMD's problems but reading between the lines...I wonder if Mubadala is a customer of theirs...

AMD is trapped by the WSA, the only way out is via Global Foundries. AMD is West Berlin pre 1971, only the Airport is owned by the USSR.:twisted:

Its a little worse than that, they can use any airport they like so long as it is the on owned by the USSR...and regardless whether or not they actually use the airport they have to pay for 15 flights a day. Destination? Nowhere. Time of arrival? Soon, very soon.

 

[mrmt]

I been thinking about this and that agency doesn't blame the WSA and the deal with Global Foundries directly as the source of AMD's problems but reading between the lines...I wonder if Mubadala is a customer of theirs...

Mubadala is rated by the big three, which means they should have some kind of relationship.

As for the rating companies pointing for the WSA... Fitch actually points to WSA payments as one of the risk factors for AMD, but the reasons for AMD demise are a bit worse than the WSA: They don't have competitive products.

 

[Centauri]

I think there's this weird perception that AMD doesn't have competitive products if they aren't offering products that compete with Intel's best. Now don't get me wrong because I'm an AMD fanboy, I'd love AMD to have the upper hand or at least be playing an even game, but AMD's fairly rushed efforts to put up a fight in the low-mid range/mobile markets have been surprisingly fruitful.

If AMD winds up deciding that they're done competing with Intel anywhere near or above $200, then so be it. Let them concentrate their still-large engineering capabilities (it's easy to forget that AMD has resources that are the envy of pretty much everybody except Intel and maybe Qualcomm) entirely on APUs and the forthcoming ARM strategies.

AMD may yet come out of this a lean, profitable company. But they need to position themselves as such before it can happen, because over the past several years they've spread themselves way too thin and butted heads with far too many competitive offerings from far too many companies.

AMD's future is probably somewhere along the lines of 'dominant niche'. Their main problem in getting there though is that they didn't have the foresight to change course before putting themselves in a position where they needed to mortgage themselves to the hilt.

 

[Phynaz]

I think there's this weird perception that AMD doesn't have competitive products if they aren't offering products that compete with Intel's best.

The perception (and it's not weird) that AMD doesn't have competitive products is based upon AMD's sales. If their products were competitive they would be selling a lot more. How people can argue against this very simple fact is lost to me.

 

[piesquared]

The perception (and it's not weird) that AMD doesn't have competitive products is based upon AMD's sales. If their products were competitive they would be selling a lot more. How people can argue against this very simple fact is lost to me.

Because it's not a simple fact. P4 vs A64 for example. Or Radeon vs intel igp.

 

[mrmt]

Because it's not a simple fact. P4 vs A64 for example. Or Radeon vs intel igp.

It is a simple fact today. The P4 is not sold anymore, much less the A64.

As for Radeon vs IGPU, you are right, Radeons are a superior alternative to Intel iGPU, the problem is that AMD must spend a lot of die area on the GPU without a corresponding increase on the ASP of their processors. Semicon companies die or live by the amount of money they can extract from a wafer, and the APU concept isn't a winner here.

You may argue that AMD would be worse if they didn't pursue the APU concept. From a marketing standpoint, this would be right. AMD wouldn't be able to claim superiority in anything. But marketing alone does not pay the bills, and maybe AMD would be in a better position if they had spend their money in a good CPU architecture.

 

[frozentundra123456]

The perception (and it's not weird) that AMD doesn't have competitive products is based upon AMD's sales. If their products were competitive they would be selling a lot more. How people can argue against this very simple fact is lost to me.

Their products are more competitive since Piledriver, but they are competitive in only small niches. Here is my take: Intel has the low, low end with Celeron and Pentium, AMD is competitive in the low/mid range with A10 and FX6300 vs i3, while they obviously trail badly in the high end except in certain very limited productivity scenarios.

The problem I see is that they still use more power for similar or lesser performance (CPU wise) and their image took a terrible hit with Bulldozer. And it is not bad enough that bulldozer was not technically what it should have been, AMDs credibility also took a serious hit from all the hype leading up to the launch.

 

[OCGuy]

Only people who have a emotional interest in AMD due to their products have not seen what is coming.

Every couple months now there are rumours of a buy-out, merger, etc, and only on tech boards do some people ignore them.

The old saying "Where there is smoke, there is fire" is applicable to the downfall of AMD as we have known it.

 

[Centauri]

I'd be fine with a buyout or a merger, and I don't know of many AMD heads who wouldn't be. The contention is over the fact that a lot of enthusiasts seem to be pining for a liquidation, as if it would benefit anybody.

 

[erunion]

No, we're just discussing a major event in the semiconductor industry.

Being in denial doesn't make it any less real.

 

[tomatom]

If you ever make a book of your forum posts, count me in for none.

" How is this a Hot Deal? "

last time it was this low gutsy buyers made 4-5 times their $s

I tried to tell all of you AMD FanBoys < myself included > ,

that $1.85 was a screaming BUY < went straight up to $3 >

... the AMD Reality Show is better than any " Cable-TV Reality Show " .

Although we appreciate everyone's enthusiasm, I'd like to remind you guys that this is a tech forum, not an investment forum. AMD's business situation is an important part of the company's fate, but let's not lose focus here.

-ViRGE

 

[pablo87]

Mubadala is rated by the big three, which means they should have some kind of relationship.

As for the rating companies pointing for the WSA... Fitch actually points to WSA payments as one of the risk factors for AMD, but the reasons for AMD demise are a bit worse than the WSA: They don't have competitive products.

You are right, they don't have competitive products for now. 2013 looks more promising, 2014 even more so, though the WSA will continue to rear its head.

Despite this though, excluding the costs directly attributable to the WSA, and the $100MM excess Llano write-off which is more gloflo shenanigans, AMD didn't lose money.

And if we consider the roadmap altering behaviour caused by the WSA*, AMD would have made money, at least $500MM by my count.



* once Kabini ships, Richland has no hope in mobile - consumes more power and has an extra part, even has less real cores - and it costs at least 3x to make. Why do it? WSA.

* PD in desktop did not need L3 - its only there for servers - die size would be 200mm^2 at most (based on trinity and 30mm^2 per module), and consolidate sockets (AM3/FM2). Why do it? WSA.

* Kabini is strangely much larger than Brazos, my pet theory is they designed it to fit both TSMC and Gloflo which meant a less dense design. Why do it? WSA.

* Kabini is your low cost vanilla, volume part, it should be 2/3 of the overall volume, why won't it be? WSA. see Brazos po's to TSMC stuck at Seifert's mailstop.

 

[Idontcare]

And if we consider the roadmap altering behaviour caused by the WSA*, AMD would have made money, at least $500MM by my count.

This kind of hypothetical supposition is difficult to wade into because the contract pricing for prior business quarters benefited from the take-or-pay (ToP) backdrop that was entrenched in the "risk management" side of the capacity-build equation on the side of GloFo's management.

Were the WSA not in place, and were it not for the take-or-pay clause, AMD would have surely paid a far steeper price per wafer (and likely never swung the cherry deal of paying only for good die in the early 32nm ramp days), and they would have been far more susceptible to the capacity expansion plans of GloFo's management.

In other words, yes to be sure Q4's numbers would have been markedly improved if the WSA and ToP contract did not exist, but all the numbers for the prior quarters would have been commensurately worse if it weren't for those same contracts.

 

[piesquared]

It is a simple fact today. The P4 is not sold anymore, much less the A64.

As for Radeon vs IGPU, you are right, Radeons are a superior alternative to Intel iGPU, the problem is that AMD must spend a lot of die area on the GPU without a corresponding increase on the ASP of their processors. Semicon companies die or live by the amount of money they can extract from a wafer, and the APU concept isn't a winner here.

You may argue that AMD would be worse if they didn't pursue the APU concept. From a marketing standpoint, this would be right. AMD wouldn't be able to claim superiority in anything. But marketing alone does not pay the bills, and maybe AMD would be in a better position if they had spend their money in a good CPU architecture.

No, it's not a simple fact today or ever. And your post about die size has nothing to do with my response to Phynaz about sales and competitive products.

 

[Idontcare]

No, it's not a simple fact today or ever. And your post about die size has nothing to do with my response to Phynaz about sales and competitive products.

Going by the financials, it seems fair to say that the products are not competitive in cost-structure and ASP.

The whole purpose of AMD spinning off their fabs was to lower their cost-structure. And yet they are less competitive than Intel with its higher capex footprint and R&D expenses for node development.

And you can't claim this is just an AMD vs Intel size issue either because you see AMD struggling in the same financial metrics when comparing them to Nvidia as well.

The numbers tell the reality like it is, all the talk aside, if they had competitive products then their marketshare growth and revenue would show it. Instead it is their competitors who have marketshare growth and revenue to show for their efforts in developing competitive products.

 

[piesquared]

Going by the financials, it seems fair to say that the products are not competitive in cost-structure and ASP.

The whole purpose of AMD spinning off their fabs was to lower their cost-structure. And yet they are less competitive than Intel with its higher capex footprint and R&D expenses for node development.

And you can't claim this is just an AMD vs Intel size issue either because you see AMD struggling in the same financial metrics when comparing them to Nvidia as well.

The numbers tell the reality like it is, all the talk aside, if they had competitive products then their marketshare growth and revenue would show it. Instead it is their competitors who have marketshare growth and revenue to show for their efforts in developing competitive products.

Interesting tactic, confusing correlation with causation. IDC, the fact and point of the matter is that AMD DOES have competitive products. And in fact do have the highest performing, most efficient and advanced GPU's on the market. For over a year. Their APU's, once the marketing spin is swept aside, would also be the most desirable products in their class to consumers. This demographic doesn't care about a few seconds here and there in some single threaded benchmarks. But the difference in graphics performance in these products is a very tangible and perceptual experience. It's something they can feel, and it affects their experience. So market share is not necessarily an indication of competitive products or AMD would have a lot more of it. I don't think there is any doubt that there are other forces at work. I have a few opinions on that as well.

But, what's coming around the corner will likely make it all moot. HSA has the backing of a very large cross section of the industry and with it comes mindshare and automatic marketing power.

 

[CHADBOGA]

Their APU's, once the marketing spin is swept aside, would also be the most desirable products in their class to consumers.

And their problem is that this is a tiny class of consumers.

This demographic doesn't care about a few seconds here and there in some single threaded benchmarks.

It would be all single threaded benchmarks wouldn't it?

But the difference in graphics performance in these products is a very tangible and perceptual experience. It's something they can feel, and it affects their experience.

For that small group of consumers who care enough about graphics performance that is above Intel's IGP's, and performance that is below discrete cards.

That is not a very big grouping.

So market share is not necessarily an indication of competitive products or AMD would have a lot more of it. I don't think there is any doubt that there are other forces at work. I have a few opinions on that as well.

Those other forces at work would no doubt include users valuing energy usage and being rather ordinary desktop users, for whom, single core performance actually matters.

But, what's coming around the corner will likely make it all moot. HSA has the backing of a very large cross section of the industry and with it comes mindshare and automatic marketing power.

Will HSA even matter enough to be called a niche? I doubt it.

 

[mrmt]

Interesting tactic, confusing correlation with causation. IDC, the fact and point of the matter is that AMD DOES have competitive products. And in fact do have the highest performing, most efficient and advanced GPU's on the market. For over a year.

When I say they are not competitive, by no means I say that the performance of AMD products isn't comparable to the performance of their competitors.

But when you look from a holistic perspective, meaning pricing, cost, performance, support structure, features, whatever you can throw here, it is obvious they are behind the competition. The lack of value perceived by the consumers is translated in low prices and shrinking sales we have been seeing. The lack of sales *is* a serious indicative of lack of competitiveness.

Take for example, GPUs. AMD strategy of delivering more efficiency than Nvidia for the last two generations made their share jump in the GPU market, it was only when they changed this proposition and went for a GPGPU monster that they share shrank.

Those other forces you deem guilty for AMD woes of lately are some of them creation by AMD itself. The financial markets didn't force AMD to launch Bulldozer, but the financial market is punishing AMD for not generating enough revenues. Also the reason on why nobody wants to put money in AMD, except for Mubadala, is because nobody really sees value in AMD product portfolio, a portfolio determined and researched by AMD and AMD only.

 

[mrmt]

You are right, they don't have competitive products for now. 2013 looks more promising, 2014 even more so, though the WSA will continue to rear its head.

What competitive products do you see for 2014?

I see 2014 as a serious year for them, 14nm Atom and Core from Intel, custom ARM64 cores from the ARM crowd, and AMD stuck with a subpar foundry with no competitive advantage on designs.

Despite this though, excluding the costs directly attributable to the WSA, and the $100MM excess Llano write-off which is more gloflo shenanigans, AMD didn't lose money.

And if we consider the roadmap altering behaviour caused by the WSA*, AMD would have made money, at least $500MM by my count.

Yes, they would be far better than without the WSA. By how much IDK. Remember that in 2011 AMD did pay only for good dies, WSA courtesy too.

* PD in desktop did not need L3 - its only there for servers - die size would be 200mm^2 at most (based on trinity and 30mm^2 per module), and consolidate sockets (AM3/FM2). Why do it? WSA.

This here. It's not WSA fault.

The server SKU in desktops makes sense for a two reasons:

- No extra validation effort

- Die salvaging

The reason for that is that AMD assumed that they could sell a X number of server chips at Y dollars, and in their analysis this alternative would be more profitable than designing a new die and selling cheaper and then having to eat all the defective server parts.

This backfired and they are bleeding market share at an alarming rate. The WSA may have been a constraint that led to the analysis and posterior decision, but it was far from being a condition imposed on them.

For the other reasons you listed the WSA should be the culprit.

 

[piesquared]

When I say they are not competitive, by no means I say that the performance of AMD products isn't comparable to the performance of their competitors.

Yes, AMD do have and has had competitive products for many years, so we've established that is not an issue.

But when you look from a holistic perspective, meaning pricing, cost, performance, support structure, features, whatever you can throw here, it is obvious they are behind the competition. The lack of value perceived by the consumers is translated in low prices and shrinking sales we have been seeing. The lack of sales *is* a serious indicative of lack of competitiveness.

From the holistic perspective examples you listed:

Pricing- AMD has been leading in pricing almost their entire existence. Scratch that, their entire existence.

Cost- Hard to comment without knowing the cost of all 3 competitors. What's clear is that AMD can operate at much lower margins than the other 2.

Performance- We've already established that AMD is not only competitive, but leading in many categories.

Support structure- What are you referring to here?

Features- AMD has FMA3, FMA4, BMI in their latest CPU's, 1 year ahead of the competition. For GPU's, they've consistently lead NVidia in features for many years. intel isn't even on the same planet let alone ballpark, in regards to GPU features, performance, efficiency, drivers, dev relations and a plethora of other factors. What could you be referring to here when you say they are behind the competition on features?

Whatever you can throw in there- I'm thinking about a salesperson's bias and commission, ie retail edge. Online viral marketing. etc

As for lower pricing, if lower prices on competitive, performance and efficiency leading products doesn't translate to higher sales then there have to be other dynamics in play. We've seen one such example during the P4 and antitrust days. Perhaps this is still ongoing behind the scenes.

TL;DR- Lower sales *is not* necessarily indicative of competiveness.

Take for example, GPUs. AMD strategy of delivering more efficiency than Nvidia for the last two generations made their share jump in the GPU market, it was only when they changed this proposition and went for a GPGPU monster that they share shrank.

Considering GPGPU and heterogeneous computing is the future of computing, I wouldn't consider this a liability in any way. It does however show that AMD is once again ahead of the curve in features, and the market is headed in this direction with the overwhelming support of HSA.

AMD's GPU market share fell when intel began supporting nv enthusiastically in their products. This is indicative of a monopoly controlling the market.

Those other forces you deem guilty for AMD woes of lately are some of them creation by AMD itself. The financial markets didn't force AMD to launch Bulldozer, but the financial market is punishing AMD for not generating enough revenues. Also the reason on why nobody wants to put money in AMD, except for Mubadala, is because nobody really sees value in AMD product portfolio, a portfolio determined and researched by AMD and AMD only.

The financial markets are puppets of the U.S. administration thanks to bailouts on the taxpayers backs. The big MM's and financial institutions manipulate the markets constantly. Anyone watching could see that plainly. And lest we forget that Otellini is on Obama's technical advisory squad. From there, it's just a matter of connecting the dots.

 

[mrmt]

Whatever you can throw in there- I'm thinking about a salesperson's bias and commission, ie retail edge. Online viral marketing. etc

As for lower pricing, if lower prices on competitive, performance and efficiency leading products doesn't translate to higher sales then there have to be other dynamics in play. We've seen one such example during the P4 and antitrust days. Perhaps this is still ongoing behind the scenes.

What you are essentially saying is that AMD didn't commit any mistakes, that selling 315mm^2 chips against 100mm^2 is fine, that the delays in the road maps didn't matter, that they have the right products on the right time, and despite that, they are still in such a dire situation because of Intel and Nvidia enforcing monopolistic practices.

If this is your analysis, there is no point in discussing anything, write off AMD and wait for liquidation. No matter how good AMD products are, they won't be able to sell and will rather sooner than later run out of cash.

 

[CHADBOGA]

The financial markets are puppets of the U.S. administration thanks to bailouts on the taxpayers backs. The big MM's and financial institutions manipulate the markets constantly. Anyone watching could see that plainly. And lest we forget that Otellini is on Obama's technical advisory squad. From there, it's just a matter of connecting the dots.

And adjusting one's Tinfoil Hat. :awe: