AMD cutting 15% of workforce, admits PC market changes happening faster than thought

[caveman-jim]

In today's Q3 results call, AMD announced they will be cutting 15% of their workforce in a restructuring move to account for their failure to anticipate the rate of change of the PC market:

"The PC industry is going through a period of very significant change that is impacting both the ecosystem and AMD," said Rory Read, AMD president and CEO.

"It is clear that the trends we knew would re-shape the industry are happening at a much faster pace than we anticipated. As a result, we must accelerate our strategic initiatives to position AMD to take advantage of these shifts and put in place a lower cost business model. Our restructuring efforts are designed to simplify our product development cycles, reduce our breakeven point and enable us to fund differentiated product roadmaps and strategic breakaway opportunities."

Source - AMD Press Release

AMD are not the only company who appear to have been caught out by the shifting market, as Microsoft and Intel both reported decreases of significant amounts.

It has been widely reported that the workforce layoffs are going to be targetted at AMD's Graphics Product's division in Markham, Ontario. If true, this might point to decreased R&D investment and a change in future product focus - no more big GPU's? Or perhaps, less investment in niche markets like high performance desktop.

[Crisler]

First it is good to here that the layoffs are not as deep as originally projected. If this move is mostly from the GPU side it actually makes sense. The high end GPUs have the lowest cost return of the various chips AMD makes.

I think what is needed now is a clear target and direction for AMD as a whole. The time has come for the seperation of the GPU and CPU divisions to truly end. My hope is we will see a clear, unified target and direction from the company.

[Lurk]

from a very egotistical standpoint, I'm asking myself how much this is going to affect us AMD consumers and specifically if this will mean decreased quality in terms of software support, my greatest concern atm.

[caveman-jim]

From the CFO Commentary (PDF link)

Restructuring Plan and Operational Efficiency Initiatives
• Implemented to improve cost structure and to enhance the company’s competitiveness in
core growth areas, including low-power and ambidextrous architecture.
• Fourth quarter actions expected to create a more efficient cost structure and result in
operational savings of approximately $20 million in Q4 2012 and $190 million in 2013.
• Cost reductions driven primarily by a reduction of AMD’s global workforce of approximately
15% and site consolidations in Q4 2012 as AMD realizes design methodology efficiencies
and simplification of processes.
• Estimated restructuring expense of approximately $80 million will be recorded in Q4 2012.
• AMD is also putting in place a business model to break even at an operating income level
of $1.3 billion of quarterly revenue. The company is targeting to achieve this by the end
of the third quarter of 2013.
• By taking these actions, AMD expects to position the company to compete more effectively
and drive enhanced returns for shareholder.

125 designs using Windows 8 are coming with AMD processors (APU's and CPU).

Computing Solutions segment revenue decreased 11 percent sequentially and 28 percent year-over-year. The sequential decrease was driven primarily by a weaker consumer buying environment impacting sales to Original Equipment Manufacturers (OEMs) as well as lower ASPs across all geographies.

Graphics segment revenue decreased seven percent sequentially and 15 percent year-over-year. Graphics processor unit (GPU) revenue decreased 14 percent sequentially due to lower unit shipments to OEMs partially offset by higher channel sales.

It appears the CPU business decreased from having no demand from OEM's for consumer CPU's. This correlates to the A-series APU's based on Llano write down of $100M for unsold inventory, which is a turn around from last year where Llano sold out - demand was higher than production capacity.

Graphics sales were higher in the retail channel but down in the OEM channel, a consequence of APU's and possibly the AMD rebranding for the low and mid-range OEM only Radeon HD 7000 series.

[Crisler]

Tom's had an article on this, the section I am posting is what caught my eye.

Reads idea is to be less vulnerable to those problems, especially because he does not believe that the PC market will return to normal growth for "several quarters". As a result, he is not only reducing AMD's cost, he is also moving away from the PC market as a whole. "40 to 50 percent" of AMD's future business will not be focused on PCs. Instead, he will aim half of AMD's business three areas: At servers, which will leverage AMD's own CPUs, "third-party" CPUs, and will count on SeaMicro's server fabric to provide custom solutions. Another area will be "semi-custom" APUs for the gaming, industrial and communications market and AMD will be aiming its APUs at ultramobile devices. Despite the reduced headcount, Read believes this is possible by reusing its technology across more platforms and by simplifying product development cycles.

A move like this, if this is true, pretty much pulls AMD out of the PC market. That is a serious hit to the PC division in terms of resources, especially for development.

[Higgy10]

Sad to me personally since most of the cuts appear to be coming from there main cdn facility(the old ati hq) ATI as it was known will be gone and buried after this round of cuts.

[Omega53]

Sad to me personally since most of the cuts appear to be coming from there main cdn facility(the old ati hq) ATI as it was known will be gone and buried after this round of cuts.

Intel and nVidia would never be this heartless. How a company could buy another company then dismantle it into extinction is sad.


*Pops in my old 3dfx voodoo 2 for memories*

[acroig]

Intel and nVidia would never be this heartless. How a company could buy another company then dismantle it into extinction is sad.


*Pops in my old 3dfx voodoo 2 for memories*

[caveman-jim]

Tom's had an article on this, the section I am posting is what caught my eye.



A move like this, if this is true, pretty much pulls AMD out of the PC market. That is a serious hit to the PC division in terms of resources, especially for development.

Rory also said the he belives the PC market will be around for at least another 10 years, and they're still going to sell into it.

[AllexxisF1]

I know nothing about running a global enterprise of the likes of AMD.

But it seems to me that if you want to win in the chip market you make the fastest chip.

The fast is what sells, and if you make it efficient, even better. All I have ever heard from AMD post the early Opteron days is efficiency this and efficiency that, and where has it gotten the company.

Speed is sex and at the end of it all its what makes you deliver of e-penis's. It is what sells.

If AMD is not in the game to make the fastest most incredible chips out there for servers, enthusiast and graphics. Then pack the whole mess up and just go home. Just leave.

Again this is coming from someone who knows absolutely nothing about this sort of thing.

[caveman-jim]

For high performance desktops, that's the way to win. The problem is that market is shrinking and was at it's biggest 10% of the desktop market. $1000 procs don't work when only 3% of the high performance desktop market buy desktops that cost over $1500.

Look at Intel, the fastest Sandy Bridge and Ivy Bridge (S1155) chips are under $350. The enthusiast market uses workstation/server rebadges, S2011 Core i-series. Who buys the $1100 3960X? You can count the number of systems on one hand on Rage3D. So does performance = sex = sales? Nope. Way more guys have dual and triple GPU's than top end fastest processors.

[Crisler]

I am going to disagree, you do not need the fastest chip you need the best marketing to win.

AMD has for some time had a chip that would cause serious trouble for Intel in the OEM world but they failed to market it. Instead they put money into overclocking events and various other endeavors to make sure the enthusiast world heard about them. They missed the boat and Intel capitalized by making Intel a household name. AMD has a big issue in the fact that consumers, in general do not know who they are or have bad information about AMD. AMD allowed this, even today.

The e-peen and sexy speed sounds good to the enthusiast community but the simple truth is that crowd, us, account for around 1% of the purchases of the chips and components using them.

[shadow001]

For high performance desktops, that's the way to win. The problem is that market is shrinking and was at it's biggest 10% of the desktop market. $1000 procs don't work when only 3% of the high performance desktop market buy desktops that cost over $1500.

Look at Intel, the fastest Sandy Bridge and Ivy Bridge (S1155) chips are under $350. The enthusiast market uses workstation/server rebadges, S2011 Core i-series. Who buys the $1100 3960X? You can count the number of systems on one hand on Rage3D. So does performance = sex = sales? Nope. Way more guys have dual and triple GPU's than top end fastest processors.

True, but on the other hand even with a bad quarter, intel still made 3 billion net profit so it's not exactly that they're hurting in any shape or form....Heck it would take AMD about 3 years(or more) to make the kind of net profit Intel made in 3 months if they were making money every quarter...Just putting it into perspective.


Slowing down production on some of intel's fabs may mean an oportunity to start prepping them for the 14nm transition next year for instance.

[jlpktnst]

Well, there might be something to classic gamer PC market slowly dying out.

If nothing windows 8 is starting a radical movement of desktop to mobile like marketplace environment where everything is sold via that central marketplace under vendor's control.

There will always be hardcore gamers though, I guess its the margins thats the problem. Things have gotten quite far and I guess they have to stop somewhere eh? I don't have a problem if we have similar gpu chips than today in 10 years still. At least programmers will start optimizing and innovating again. I doubt they will totally stop the production, just r&d will slow down. Nvidia has done similar I think, moved a lot of resources to other markets. No point for them to r&d alone with dimnishing returns is it?

[caveman-jim]

AMD has for some time had a chip that would cause serious trouble for Intel in the OEM world but they failed to market it. Instead they put money into overclocking events and various other endeavors to make sure the enthusiast world heard about them. They missed the boat and Intel capitalized by making Intel a household name. AMD has a big issue in the fact that consumers, in general do not know who they are or have bad information about AMD. AMD allowed this, even today.

I don't understand how you don't see your internal consistencies here, unless I'm misunderstanding you.

Firstly, you say AMD failed to market their "chip that would cause serious trouble for Intel in the OEM world", I'm assuming you mean Trinity. From the vendors who are carrying Trinity APU products, it seems that they are aware of the product in mobile space and desktop space, but they are making more Intel products. Intel are doing this by their marketing funds for ultrabooks and other formfactors, providing development money which then precludes the use of competitor products in the similar form factor or market.

Intel has a $1Bn marketing budget, AMD doesn't even have a tenth of that. I don't think it's a matter of 'allowing' and more a matter of 'cant compete'. Without similar funding they can't break the OEM contracts with Intel by offering to pay off the penalties.

Now, AMD know this, because they've got a CEO from Lenovo - he knows first hand how this works, right up to the minute. They've got a CIO from Apple, another customer of Intel. These guys know the pressures and strategies first hand, so they can shape how to contrast against them.

AMD aren't giving up, they're reshaping what market they're thought of in and focusing on their strengths in markets that are growing, not shrinking.

[shadow001]

AMD aren't giving up, they're reshaping what market they're thought of in and focusing on their strengths in markets that are growing, not shrinking.

Tell that to the 10% of the workforce that lost their jobs last year and the 15% losing theirs by the end of this year...One thing is for certain is that is if the jobs are mostly in engineering affecting the CPU side of their business, it's one more nail in the coffin in terms of ever seeing a high performance CPU that can compete with what intel has at the time.


At least for the PC, it's becoming a one horse race basically....

[Redeemed]

Tell that to the 10% of the workforce that lost their jobs last year and the 15% losing theirs by the end of this year...One thing is for certain is that is if the jobs are mostly in engineering affecting the CPU side of their business, it's one more nail in the coffin in terms of ever seeing a high performance CPU that can compete with what intel has at the time.


At least for the PC, it's becoming a one horse race basically....

For the PC and gaming, this will likely not improve over the coming years.

[badsykes]

I am sticking with my 6950 for a couple of years...Windows 8 just nailed the PC gaming ... At least Indie games are getting much more traction and i have plenty of things to play in this gaming part...
I am curious what will happen to HIS , Sapphire .. they will disapear if ATI goes down?

[Crisler]

I don't understand how you don't see your internal consistencies here, unless I'm misunderstanding you.
.

Sorry you misunderstood, AMD is still pretty much unknown to the general public. We have had this discussion before about AMD needs to work on brand awareness with the consumer to get any real traction.

[caveman-jim]

more than 50% of AMD's revenue comes from 5 customers. That ain't end consumer. AMD sell very few products directly to the general public.

[shadow001]

more than 50% of AMD's revenue comes from 5 customers. That ain't end consumer. AMD sell very few products directly to the general public.

They've fallen quite a lot in the last few years, that is true.


I still remember the athlon 64 days and the athlon X2 series when all intel had was the pentium 4 wich sucked in a big way, and it still astounds me how AMD went from that situation to the current one where it's barely surviving.

From mid 2005 onwards, intel has been on a roll and nothing AMD has done since then has really challenged them in performance terms.

[cbsboyer]

They've fallen quite a lot in the last few years, that is true.


I still remember the athlon 64 days and the athlon X2 series when all intel had was the pentium 4 wich sucked in a big way, and it still astounds me how AMD went from that situation to the current one where it's barely surviving.

From mid 2005 onwards, intel has been on a roll and nothing AMD has done since then has really challenged them in performance terms.

The problem for AMD was that Intel has a massive R&D budget, so they can afford multiple concurrent development streams pursuing a particular future market and converge them on a winning design. AMD isn't really in a position to do that, and now that they have been in the position in the PC market where they have not had the "bang for the buck" advantage for quite some time, this revenue stream is suffering and they are being forced to retrench.

AMD's future is mobile and embedded computing simply because they can not afford to be competitive in mainstream PC components. If they can maintain and grow their distributed/data center/parallel computing presence they can be successful there, but the days of an AMD based enthusiast machine are over. They don't have the CPU lineup for it (and don't have the resources to do much about it) and the GPU ecosystem is very volatile where every product tends to be "do or die", which is an expensive way to run.

Or at least this is what I'm seeing from my desk as just some guy who's been around for a while.

[caveman-jim]

The problem for AMD was the anti-competitive practices put in place around the globe by Intel.

[cbsboyer]

Well, there is that too...

[Crisler]

more than 50% of AMD's revenue comes from 5 customers. That ain't end consumer. AMD sell very few products directly to the general public.

While this might be true in the books, this is NOT true in reality.

Walk down the street and hum the Intel sound from their ad campaign and 90% of people know that it is Intel. Ask them about Intel and you will get a general response about computers. Now go ask them about AMD and you will find that most have not heard of AMD and the ones that have really have no clue.

Average american walks in to buy a computer. They do not know squat about hardware, they just want to use their computer. They see in a local Bestbuy or Sam's Club this row of computers and they walk over. They see two systems, both at the price point that meet their budget. One says Intel on it the other says AMD. Eight out of Ten times they buy Intel. It is not because Intel is faster, it is because they KNOW that name and feel trust for it.

AMD through the Vision program and with Fusion chips have a WIN in the home consumer market from a pure hardware and sales pitch direction, by a huge margian I might add, and yet they lose. The reason is simple AMD has mythed itself into believing they do not sell to consumers.

They outsource their product and name recognition to partners and look what it has gotten them. I mean hell they do this for their Physics and that has worked out well. Their 3D is from a partner and wee see how that has taken off. They need to quit waiting for others and do the work themselves to really be successful.

AMD has a strong message but instead of blowing the trumpet and letting the news spring forth they wait for others to do it for them. It is almost like they are scared to go head to head with Intel.

I understand they do not have the money that Intel has for the budget but come on. They spend money for all these overclocking hypes that at the end of the day are meaningless. They post videos on YouTube that look are all cryptic and geeky but do nothing to get their message out except to the geeks that already know it. They place ads in magazines geeks use, why not use those ads in magazines to get brand awareness to the rest of the public?

There are people in AMD's own PR group that feel this way and when I asked them about it they are frustrated because they preach the same mantra and AMD still puts the money they have into the same useless ventures.

Sell to OEMs all you want but if the stuff on the shelf that is moving is NOT the AMD stuff then the OEMs will not buys as much. You have to get brand recognition to grow and AMD is alike a timid school girl standing in the corner.

[Nunz]

I am going to disagree, you do not need the fastest chip you need the best marketing to win.

AMD has for some time had a chip that would cause serious trouble for Intel

What chip is that again?

[AllexxisF1]

While this might be true in the books, this is NOT true in reality.

I agree at some level Crisler, however its how you market your item. I believe having the fastest chip is the ultimate in marketing. For example during the early Opteron and Athlon days when AMD had the better chip against Intel it owned the space...not because of grand marketing (although there were some commercials) but because they simply had the best chip.

The enthusiast knew which was better which leads to most of those blokes working in IT departments along with giving the normal advice to family and friends.

You need to be the winner in chip speed, consistently. Yes I know you only sell a few of those, blah, blah and blah. You need it to start changing perception and then catapult from that with a die hard marketing campaign. If AMD does not have the resources to do that, then they either partner up with someone else like IBM or close the doors and move on.

Personally what AMD needs to do is develop the fastest and efficient mobile processor for tablets. Period. I'm not talking a small jump ahead, I'm talking something from 3012. It needs to be revolutionary and it can be, outside the damn x86 instruction space.

[Crisler]

What chip is that again?

At the lower end the fusion based systems are a better choice than an i3, especially in the EOM market. The fusion base is more versatile and offers a better consumer level computing experience thanks to the onboard GPU ripping Intel apart.

[Dr. Zhivago]

I agree at some level Crisler, however its how you market your item. I believe having the fastest chip is the ultimate in marketing. For example during the early Opteron and Athlon days when AMD had the better chip against Intel it owned the space...not because of grand marketing (although there were some commercials) but because they simply had the best chip.

The enthusiast knew which was better which leads to most of those blokes working in IT departments along with giving the normal advice to family and friends.

You need to be the winner in chip speed, consistently. Yes I know you only sell a few of those, blah, blah and blah. You need it to start changing perception and then catapult from that with a die hard marketing campaign. If AMD does not have the resources to do that, then they either partner up with someone else like IBM or close the doors and move on.

Personally what AMD needs to do is develop the fastest and efficient mobile processor for tablets. Period. I'm not talking a small jump ahead, I'm talking something from 3012. It needs to be revolutionary and it can be, outside the damn x86 instruction space.

Even when AMD had the far better architecture for YEARS, it barely made a dent in the market and public mindshare due to the way Intel did and continues to do "Business". You're choosing to look at things through rose-colored glasses if you think AMD "owned the space."

AMD was the first to make it to 1GHz, but did the mainstream media report on that? No. When Intel released the P4 which had a higher GHz #, did the mainstream media report on that and claim that Intel had the "fastest CPU"? Of course they did. When AMD released the Athlon64 and Opteron which blew away the current Intel P4 and Xeon architecture, did the mainstream media report on ANY of that? No, of course not. When Apple moved to Intel did the mainstream media report on that? Yes, of course they did. See how that works?

The enthusiast market is a tiny, miniscule, microscopic fraction of the overall market. OEM's don't make any money on that stuff. Obviously, your view is skewed because you belong to a site like this and play in this space. The majority of the computing world buys pre-built $500 computers with Intel integrated graphics.

And have you talked to most IT people? They don't give a **** about performance. If they did, then most large company PC's and laptops wouldn't run like complete dogshit. They only care about "compatibility and security". They ALL favor Intel and nvidia. Most of them are clueless about anything else. At least that's my experience with all the IT folks I've ever dealt with in the Corporate world at any company of size.

Even though the AMD APU's are a much better choice for most people when talking about basic systems with integrated graphics, Intel will still make sure that they stay on top by offering their "incentives" to the OEM's to undercut AMD. It really doesn't matter what AMD does because unless Intel changes the way they do business and unless the rest of the world and most IT departments get a clue, nothing is going to change.

As to any IT folk who read my comments who might be offended, please don't be, unless you're one of the clueless.

[caveman-jim]

Crisler is right about Trinity, it is a great part, and you can look at the reasons out lined by Dr. Z as to why they're not a bigger success.

Alexxis is still wrong, he's not aware of the selling price points that AMD compete in across the board but only thinking about desktop performance chips, which is a smaller market. Desktop performance chips don't sell servers. Never have, never will. Win on Sunday, Sell on Monday hasn't applied for a long time now, and if you dig into what Dr. Z is saying, you'll find that the IT guys are going with what their vendor account managers tell them, and that comes from corporate as to which lines and aspects to push, which not to. Where does that come from? IHV marketing dollars, a point Crisler missed. Intel's $1Bn marketing budget is focused on OEM's and resellers, retailers and etailers. Everything they do is to benefit them. As an aside, that includes Intel marketing that's visible to consumers, because it's part of the package - buy X% of Intel systems of specific breakdown Y to sell in your products, and guess what you'll see on TV? Your models next to our logo at these retailers!

Crisler is exposing some points that AMD can address, but he's not thinking about their overall strategy, they're investing their marketing and focus in the new expanding markets they're shifting to compete in, not the markets they're still in right now and struggling in. This is because they're resource limited, for reasons Dr. Z, outlined as well as their own internal failures. At some point you've got to acknowledge the realities of the markets and say this is what we're going to deal with, we can't throw our hands up and cry foul and hope everybody just says 'oh, ok'.

Performance desktop market is shrinking but will be around for another 10 years, big GPU's will be needed but not necessarily big CPU's. You might argue that the latest FX vs IVB results clear indicate otherwise but as gaming platforms become HSA the PC market will have to become HSA compliant as well.

Performance desktop is more than enthusiast gaming but both enthusiast gaming and performance applications can benefit from adopting OpenCL and native parallelism, AMD will benefit with the HSA platform providing their place in the market. In the long term people making money from performance desktop applications now will find it more cost effective and offer better returns to move to either compute cloud offerings, either local private or hosted private. More and more new companies are starting without IT, they're going straight to the cloud for their business operations from day of VC funding.

Note - some people seem to think when I post things like this that I think it's a good thing. If you want to know my opinion, ask me, don't tell me I'm shill/apologist/fanboy. All I'm doing is posting my opinion of what I think AMD are doing based on my experiences. Take that for what it is, some guy on the intarwubz.

[shadow001]

Performance desktop market is shrinking but will be around for another 10 years, big GPU's will be needed but not necessarily big CPU's. You might argue that the latest FX vs IVB results clear indicate otherwise but as gaming platforms become HSA the PC market will have to become HSA compliant as well.

Performance desktop is more than enthusiast gaming but both enthusiast gaming and performance applications can benefit from adopting OpenCL and native parallelism, AMD will benefit with the HSA platform providing their place in the market. In the long term people making money from performance desktop applications now will find it more cost effective and offer better returns to move to either compute cloud offerings, either local private or hosted private. More and more new companies are starting without IT, they're going straight to the cloud for their business operations from day of VC funding.

I'll hate the day when i can't have my own high end personal home computer since i am a hardware enthusiast, to do whatever i feel like and whenever i feel like doing it, and not relying on cloud services because what's available by then for end users isn't powerfull/fast enough to get it done.

But i do realise it is coming regardless if i like it or not.

[caveman-jim]

we might have to find new hobbies in a decade and a half.

[Dr. Zhivago]

I'll hate the day when i can't have my own high end personal home computer since i am a hardware enthusiast, to do whatever i feel like and whenever i feel like doing it, and not relying on cloud services because what's available by then for end users isn't powerfull/fast enough to get it done.

But i do realise it is coming regardless if i like it or not.

we might have to find new hobbies in a decade and a half.

[shadow001]

I guess that when that day aproaches where one can no longer be a hardware enthusiast, i'll buy the biggest, nastiest system possible that is well above any software requirements needed in 15 years time, and it'll be my final system for everything and i shall flip the bird to any cloud services offered...


Then again by that time i'll be nearly 60 years old and not far off from kicking the proverbial bucket, so it'll be the younger generations that will have to deal with the cloud crap anyhow...