AMD cuts 10% of workforce, leaves one group untouched
This is going to hurt a lot
Nov 3, 2011 in analysis, Finance, Microprocessors, Opinion, Rumors
AMD (NYSE:AMD) has just decided to lay off about 10% of it’s employees according to a press release, but one group of people is strangely untouched. While our condolences go out to all the ex-AMDers out there, we do fear for the remaining ones.
The short story is simple, 10% are gone, told you so. Slightly longer, the press release was entitled, “AMD Optimizes Cost Structure to Enhance Competitiveness and Accelerate Growth“, and shows that AMD has a lot of feeling for those let go. I am pretty sure they will all be glad to know that they are good optimizations, the sentiment almost brings a tear to my eye. The best list of high level ‘optimizations’ I have seen is at Icrontic, including a nice link to the internal memo itself describing the ‘optimizations’.
There are a few things that come to mind here, the lists we have seen, and a few that asked not to be publicly named, are some of the best and brightest at AMD. Having lost their CEO under exceptionally stupid circumstances, then their COO, VP of corporate strategy, and quite possibly their best remaining executive asset at that point, Rick Bergman, you have to wonder what is going on?
According to Icrontic, the list of ‘optimized’ employees is going to hurt. Carrell Killebrew, the man credited with Eyefinity, probably the largest advantage AMD has in the graphics market right now, VP of corporate strategy Patrick Moorhead, and many others. Add to this the brain drain that has happened over the last few months, people like Dr. Gamal Refai-Ahmed, David Hoff, and dozens of others, you have to wonder who is left.
That is completely unfair, there are lots of good people remaining at AMD, but the sheer number of VPs, Fellows, and other high ranking people that were just optimized is rather shocking. If you don’t think this is the end of an era for AMD, here is something to think about, Patrick Moorhead was the last VP at AMD from the Sanders era. Of the dozens of VPs at a company of AMDs size, he was the last one of the old guard remaining. For good or for bad, this is a new chapter for AMD.
Updated 11/3/11: Fixed spelling, Moorhead has no second e.
With all of the optimizations flying around, you have to wonder about one group that has been totally untouched by this round, the board. When talking about the buck stopping somewhere, you normally go up the food chain until you can’t think up an excuse to point the finger at the guy with the next largest office.
The number of CEOs for the first few decades of AMDs existence was one. In the last decade, there have been three more, so the buck probably doesn’t stop with that post anymore, nor the fully optimized ranks just below them. You have to wonder when shareholders are going to question the top level when everyone else that can be blamed is gone by now?S|A
Rory Read has a stock option bonus for driving the stock up a little bit. He played the cheapest and easiest page out of the old investor appeasement handbook, and at first he got a little 5% return, for it to level off. Rory Read walked in on day one with a plan to cut 10% for a nice round number for a press release. He can’t know what “optimization’s” AMD requires after sixty days.
Think! What exactly does AMD do well? What valuable engineering designs does the company actually own?
AMD’s software divisions – driver support for their graphics chips, programming support for all the new parallel programming paradigms that exist on the GPU cores, drivers for the CPUs- are completely ROTTEN and INCOMPETENT. And yet, the putrid software engineers at AMD often hold positions of significant power.
Likewise, those behind the new hardware designs (bulldozer and the VLIW4 GPU design in the 6900 family GPUs) have created new products significantly worse than the previos generation designs they were intended to replace. Sadly, fanboys keep denying this, because the crap new products may high slightly higher performance, if one ignores the issue of terrible yields, massive increases in transistor counts, and massive increases in power draw.
Blaming GloFo doesn’t work, because GloFo have nothing to do with the problem of a chip that maybe doubles in numbers of transistors used, but struggles to show even a 15% performance boost (even when one ignores massively increased power drain).
AMD has great consumer products in the Phenom2 class cores (including Llano and Zacate), and the 6850/70 GPU design. These teams need to lead AMD, not the blue-sky FAIL teams behind the bulldozer and 6950/70. Sadly, AMD is betting the bank on its FAIL teams, and doing nothing to improve its driver and parallel-programming support.
The PC marketplace is getting tougher at an exponential rate. The best players in the ARM marketplace are making money hand-over-fist. It just seems that AMD and Nvidia lack the nimble ability to move quickly (see the massive number of Tegra2 design losses, driven entirely because the idiots at Nvidia did NOT put proper video-decode engines or NEON into Tegra2, are are extremely late offering Tegra3, their part which fixes both deficiencies). By now, both AMD and Nvidia should be market leaders in SoC ARM designs.
“Blaming GloFo doesn’t work, because GloFo have nothing to do with the problem of a chip that maybe doubles in numbers of transistors used, but struggles to show even a 15% performance boost (even when one ignores massively increased power drain).”
While I do think that Global Foundries is getting an unfair portion of the blame here – its pretty hard to sell a chip when the yields are getting crushed by GF’s inability to process their 32nm SOI. Lets not forget a portion of performance is raw frequency which has also been an issue with global foundries – their transistors are straight up slower than their competition.
I also dont know how you consider the llano project to be a non-”FAIL” team when its suffering from some of the same global foundry issues as BD.
With all of that said, its pretty clear that BD is 2-3 years late which doesnt help matters either.
BD/FX is less than stellar but for many desktop users it is still an upgrade from an older AMD product so it’s good that there is demand for this CPU. Obviously BD based Opterons have been in demand and GF has been unable to meet demand on Llano, FX or BD based Opteron so this hurts AMD in every possible manner.
IMO AMD needs to get Trinity laptop and Piledriver desktop based CPUs out the door in Q1 and both had better be a reasonable upgrade from Llano and FX respectively.
As we can see from insider comments there are always different perspectives and perceptions. Unfortunately we do not live in a perfect word and we could never agree on what that is anyway.
Let’s hope that when the dust settles that AMD moves forward in a positive direction that continues to deliver quality products to the masses – us. There are many roads that can lead AMD to where it needs to be. Let’s hope they pick the best and most expeditious pathway.
I just got an email from John Swinimer (former PR Manager). He’s gone as well.
Shame no one has mentioned the engineering “VP of Core Development” who was shown the door.
Hi all, Brian from Icrontic today. We’ve added a couple of names to the confirmed list of people who were let go yesterday: Craig Lakey and Hao Pham from FirePro marketing.
Time will tell if AMD has shot itself in the foot or if new blood will revitalize it’s Biz. My condolences to those dismissed. Life really is unfair sometimes. The current board at AMD could be as much at fault as the upper management? We simply won’t know for awhile. What we do know is AMD does have some good current and future products and they need to come to market in volume NOW. That would be my #1 priority while working to fix the secondary issues. Thus GloFo needs to get it’s act together NOW.
Products.. products. Which ones will be left and which will be canned.
AMD’s new boss wants a change in direction… low powered devices and the cloud.
Low power is mobile and I suspect AMD will go against ARM/Atom… good move?
Cloud I suspect is servers, where margins are relative high. Good move if they can compete.
That leaves GPUs, CPUs and APUs.
APUs (ie: Llano, etc) do very well. Once the fabs get their act together, they will do even better.
Which leaves us with? It will be interesting to see what kind of skill sets the new hires will have and the implied direction for the company.
Congratulations / All this was well predicted in January article -
So what happens to AMD now?
With Dirk out, a world of possibilities are opened up
End of an era for sure,I think trinity may well be the future of the computer industry.
eventually culminating in a highly efficient
vector processing approach.
I hope these, um, “optimizations” trickle down to the chip level. And please, fire their ad men and spinners!
Judging by the level of marketing expertise, at least the severence package should be minimal….a few bunches of bananas should cover it.
Back to the circus boys!
Dear Charlie,
Your judgements are based on the BS that some of the AMD PR guys (I can name names, but that won’t be professional, just think C.H.)
Despite what you might think, while Eyefinity had a halo effect, the so-called advantage was none, at least in consumer markets. Carrell was dumb enough, like the indecisive Rick Bergman, not to keep it exclusive to FirePro, preventing AMD from top margin dollars like what Nvidia has been doing with Quadro for years, eating AMD’s lunch on Radeon line pumping dollars to game developers.
As for Moorhead, have you ever dealt with the guy? He’s an idiot who was paid a salary of V-P just because he was from the times of KFC Sanders. As for David Hoff, don’t get me started, everybody is happy that he’s at Intel now! Good luck with his BS.
I used to respect your articles, but I see that you react like an emotional guy who just lost his contacts for leaks from AMD.
Will Rory Read succeed? I don’t know, but one thing I give him credit for, he’s decisive, unlike Bergman, he’s cleaning house, brought a new credible V-P of engineering, for me it gives me hope, but the future will tell.
This is the real World, and it was about time AMD had a CEO who cracks the whip.
Completely agree w/ “From Inside AMD”. I feel bad for the folks that got booted, but most of these folks deserved to get booted. Little outputs, lots of politics. Rory is decisive enough to take out the broom for a much needed spring cleaning…
In fact, I believe there should be > 10%. Still a lot of incompetent, politic playing people remained. I have to deal with these folks everything, and it is no joke…
Clean them all! :)
Please enjoy your time working on buggy whips. You are probably great at it but over time no one will care about them. I cannot believe that you buy into the execution-execution-execution BS. Yes, you need execution, but you also need vision. Execution only isn’t going to lead any company into any glory except if you are a low cost manufacturer. Was AMD64 or Athlon or Opteron more about execution or a combo of vision plus execution? I wish there had been a RIF list so I could have gotten a package. I’m just biding time before I can jump into ARM SOC designs, anyways. AMD, please RIF me soon so there’s still money to pay me.
Execution is what makes money buddy! You showed your real color! Basically you claim that these changes will produce mediocrity, yet you will stay to take a package… this shows well your standards of quality.
I can certainly agree that there seems to have been severe ‘execution’ deficiencies at AMD – but honestly it doesn’t seem like that has changed yet either.
Personally I think it would be a shame if AMD goes completely b2b and signage – there are only 2 players in x86 and AMD happens to be one of them ( for now? )
What I honestly don’t understand – and I would appreciate some comments on – is why the ?#$! AMD hasn’t pushed OpenCL out to developers? Yeah – I saw the ‘competition’, whose idea was that? The conditions ruled out migrating ‘legacy’ code to OpenCL – Would strike me as a better idea to have a FOSS based competition “Who can issue the most good pull requests for OpenCL enhancements” – Although.. that would probably mean that you guys would have to work on your Linux drivers some – bringing back gDEBugger for linux would also be pretty nice of y’all (whose idea was it to discontinue the linux branch?).
1 year ago it seemed to me like AMD was on a perfect trajectory to very neatly leapfrog the competition, so what happened?
Wow, some serious inside fighting happening on a public forum.
It’s funny – I’ve never seen Intel employees do this kind of stuff.
Vision + execution are both important.
You mentioned you wish you were on the RIF list. Why the F are you still working for AMD? If you are that good, then go out and get another job. Why bother sticking with a sinking ship if you are not happy with the company’s strategy?
Sounds to me you are just one of the old blood that needs to be cleansed, as they cannot find a better opportunity elsewhere…
Eyefinity is something I took into account when deciding what graphics card I wanted next.
So the proof is in the pudding, I’m a consumer, it played for me as a plus in deciding.
You have to have something special that differs from the competition.
Sorry, but one customer is not a measure of the success of a business
I also bought one – HD5770, Jan 2010, for the same reason. Upgraded from a HD4850 that was less than 1.5 years old – simply for Eyefinity. I was otherwise intending to wait until probably the 7xxx or 8xxx series before upgrading again.
I push AMD GPUs for two reasons: it pairs up with the onboard GPUs of the systems I push (leaving IGP for secondary+ monitors when gaming, leaving discrete card for full gaming capability) – and they don’t rebadge the same chip for 4 generations. (Although then they went and pulled the 6xxx rebadging, doh!)
The other options are nVidia, but that isn’t integrated – so takes more power, and Intel GPUs only come with Intel CPUs – and I prefer AMD’s – feature set is across all but the bottom chip (Sempron) and supports ECC without a server-grade chip, as well as not abusing their market position, etc.
Oh yeah! He cracks the whip! We’re gonna be a tiger! We’re a predator! We just have to execute in an executionary way, then we’ll be executin’!
Could anyone with a brain have witnessed Rory’s intro speech at AMD without barfing?
Are you worried you can’t perform and get cut?
Maybe he just doe not like the idea of the only high-end alternative to Intel left being IBM Power …
I’m worried that certain projects that were sort of understaffed are now back to being woefully understaffed, because the staff we hired not even 6 months ago in an attempt to rectify the situation are now gone.
I guess it’s because new hires don’t get as much severance?
Of course we’ll be told to pull in schedule, because it’s what you do when you can’t actually give your staff a wedgie.
All you smug types had better pull in your schedules too, you hear? I don’t want to get yanked near the end of my project to bail out your shit.
LOL
Credible VP of engineering Papermaster got shitcanned from Apple and only stayed at Cisco a year – that’s hardly a ringing endorsement.
Once again, ignorant people relying on press to make an opinion. Do you know people who work at Apple? I do, and the ones I know all respected him, the issue was a personality clash with a Jobs, not competence issue. As for CISCO, he was not fired, he was hunted by AMD. Oh wait, how long was he driving low power strategy at IBM, you “knowledgeable” person?
I didnt say he was fired from Cisco.
Also, per your trivia questions – I’m not one to post on a board without knowing what Im talking about. Papermaster was/is a large IBM name with obvious connections to the IBM crowd of executives and directors in the engineering space they have been hiring for the past 5 years. One doesnt have to be terribly connected in Austin uProc development to have met him.
I knew Morehead and Carrell, who were two of the most future looking guys I knew. Morehead was ahead of the game in his thinking, which you probably just didnt understand because you are stuck making the buggy whips. He called the MHz BS, netbook scam , user experiences and called tablet success. Also, he was one of the only guys left with balls and passion. Carrell had guts, too, and insight enough know that room computing was the wave of the future. Eyefinity will go down as one of the technologies that defined the next gen computing. I’m sorry to see them leave.
How many 6 output Eyefinity boards sold if you’re a real insider?
All Moorhead was doing in the last few years was tweeting… Moron!
As for Carrell, he was big indeed on forward looking technologies like Microsoft Excel (real insiders know what I’m talking about)!
True dat, don’t get me started on hoff either, ah f8ck it you got me started, seriously how could that guy stand up in front of the press and tell them how good AMD products were supposed to be when he’d come back inside and tell the execs how much our software sucked, I mean that’s total bs, we can’t be as bad as he kept saying we were compared to CUDA and Optimus and Quick Sync Video, seriously that’s not even fair, we’ve been like working on gpgpu since ctm 5 years ago, that’s so unfair to compare us to Jensen’s team that is expected to make money, we don’t have to show any business for our efforts and it’s not fair to compare OpenCL to CUDA since it’s well open and we’ll get at least 5 years of some more of the great industry pay and promotions without any of those mean pressures other companies have, and he actually scared several nice software types into leaving, that’s such BS, where else will they go to find a deal like this, and he would tell Dirk and anyone who would listen that there was no chance we’d have accelerated Apps for Fusion the way we were heading and that’s BS! No one actually reads the specs like 400jigawhateverflops, they buy from us because everyone hates Intel and we’ll always be here since the someone will always jump in to save us. People will always buy our stuff,even BullDozer no matter how bad, and we’ve got Clarity and Win which means they’ll surely wait for Tahiti since no one actually cares that Arkham City only ‘looks’ better on GeForce, they’ll wait because Jensen’s not as nice as we are. The dude is such BS and these layoffs are seriously a good thing as I do trust our execs got exactly the poeple we don’t need and I’m very, very glad we still have you From Inside AMD, thanks for being here.
eyefi was one of amd’s smartest moves. Looks like we’re left with nv, arm and intel.
…and AMD.
Actually most of the personnel cuts are in marketing/communications/PR. This will boost the financials and make Wall Street happy for the next year while AMD ramps BD based Opterons, Trinity and Piledriver based desktop CPUs.
The real problems are mostly with TSMC and GF who have been unable to deliver 28nm and 32nm chips in a timely fashion. I actually believe AMD is going to do just fine over the next five years and surprise quite a few folks along the way.
Engineering got hit reasonably hard as well from the list of names Ive seen.
BD being 2-4 years late (depending on how you measure – internal vs external) has little to do with global foundry’s 32nm yield issues.
I agree BD is a disaster after being two years late but what I’m saying is damand far exceeds supply of BD based Opteron, Llano laptop chippies and even FX chips in spite of the less than stellar performance. It looks like laptop Trinity chips may show on time or early which can only help revenues.
If I was being unkind I would say that’s because the supply is nearly non-existant!
Well, Cray is now completely disinterested in future “supply” from AMD after they destroyed their super business. So that should help with demand.
No doubt other “customers” are equally pissed off about “supply” as well.
Now is a good time to buy some AMD stocks.
Why? Are the stock certificates 2-ply?
$ Savings 180 mil or half years profit. However is IT Time to cut workers when llano sold out & probably from bulldozer core thru trinity, already sold out, seems NOT Quality issue.
Bet RED Chinese are behind double dealing.
Missiles To Launch, Test #1. Blow RED china. yeeessss, DaT Will Be Better.
BTW GloFo is term now Sylvia has STomped off to unknow places, RCR UnPlugged apparently ended mid september. in few short years Ms Barak made 9 pages of substantial serach pages. Charles should do search on self article.
Seems SAD Day when Best & Brightest Are thrown Into STorm.
Morris Day.ton Drashek….stormy weather fpate camp 13.vn
I expect AMD stock to remain undervalued even if they execute perfectly for the next decade. Intel has too much clout and market control.
As long as the money saved is dumped into improving products, this will turn out fine. AMD never had much marketing anyway, and won’t need it until they release another monster.
The next 5 year were NEVER in the question.
The investments for thus were mostly made so they will be OK at worst for the next 5 years even without making the right strategic decisions.
The question is more what then, whether they will go the way of VIA or be able to keep themselves on top.
Sadly, the VIA path is looking more likely by the day…
is very very good the restructuring there.
The staff was too relaxed there and less preocupied for the destiny and image of the company.
Are there still a lot of traitors that have to be removed somehow
Seems to me like this is a good thing. Seems like people were being given a little too much leeway with their projects which led to rising development costs and delays. Rory is obviously tightening things back up, that’s just the consequence of letting things slip. The fact that they actually have a CEO that is willing to make these kinds of tough decisions bodes well for the company.
It’s going to be tough losing all those great people, but I think the company will do better in the long run.
You do not fix a process (i.e. people getting to much leeway or whatever) issue by changing the staff.
You actually make it worse since the new staff will very likely have to make the mistakes over again.
Also, there is one VERY important point. AMD had (for a 40 years!) a culture that summary executions are just not acceptable.
This is the FIRST summary layoff in the whole history of AMD.
Now just think for a minute which signal this sends to anybody who was thinking “company first, personal profit second”. And I would bet there were and still are a LOT of such people. Mostly the “20% do 80% work” type.
Considering all above, I would expect the brain drain Charlie mentioned to only multiply next year. The board will simply not allow for people to be paid better (since now doing more work) and they just eliminated loyalty and pride as motivators for the top people to stay.
As it is, AMD is done. I am very sad to have to agree with Charlie on this. But unless Abudala grows balls and trows those bastards down the drain, I see no light at the end of the tunnel.
:(
You fix the issue by tightening the restraints. The firing part comes along because the previous people in charge let things go late and expensive. So you have to position the company to be profitable as quickly as possible.
New staff? It’s a layoff. The staff is the same, minus a few people.
You don’t seem to understand. It’s definitely a big deal that this had to be done. However, in the grand scheme of things, it’s a good thing that it did because things had to change in order for the company to improve.
Great, AMD has sinked into the oblivion, I like it. So, Intel\’s agent in the board has been worked successfully. This is the first step to make my dream becoming a reality when Intel owned all. Well every losers should be sacrificed, so the god blessed winners can control every human to obey them.
May Intel throw you over a barrel and have their way with you as God and you NOT intended. While vacuuming your wallet and bank account. Enjoy!!!
Not big on recognizing sarcasm?
HINT: The post above yours is an AMD troll attempting (badly) to employ reverse psychology
LOL what do you mean “Intel’s agents”?
Just wait until AMD’s agents on INTEL’S board takes action. Then Intel will be sunk along with its fleet of Itanics.
This is some of the strong stuff. The main question remains: How much of it is bloated machinery, how much is vital engineering?
In the Roman army, when discipline was too low, they decimated the unit, meaning that every (randomly chosen) tenth man was executed. That served as a good motivation for the others.
Seems like AMD’s new Centurio wasn’t too happy with discipline either…
Well, I did mention that the new boss had a broom.
Question is, did he sweep to much talent out the door? More importantly, did he sweep the 20% that do 80% of the work?
how about russian roulette? it sounds better to boost the low employee morale.
Reality is AMD not doing well has a lot to do with reliance on Global Foundries, BD fail and some awful business decisions (eg selling smartphone business to Qualcomm for $65M when the competitor has a market cap of $1.9B). Who was in charge during the time BD was being designed, who made the decisions? It wasn’t RR. It was DM and RB. The layoffs are on them.
you should study up on BD design architects and managers; you really mean CM, BG, & DM.
Management must Manage and be held accountable.
Management manages (to run the company into the ground) and have never been held accountable (before.)
If you’ll recall, Intel had a lot of problems with 32nm. Granted, Intel’s whores in the media did a lot to brainwash you and everyone else into thinking it was intentional, but it took them an incredibly long time to get a quad-core 32nm CPU out the door.
Now, I understand that whenever Intel misses a deadline or fails to significantly improve a product, it’s always due to “lack of competition”, but do you ever stop to think that these advanced nodes are getting significantly harder for everybody? 45nm was the last “business as usual” node, everything below that has been a nightmare for everybody, every time.
Oops, I never mentioned GloFlo. The point of that rant is that GloFlo isn’t doing badly, as they only really took over at 32nm/28nm. You will never see another smooth new node from anybody, it just requires too much atomic-level precision now, and has too much probability for error. Intel will do their best to make it look easy, and downplay their delays and quality issues.
“but it took them an incredibly long time to get a quad-core 32nm CPU out the door.”
ROFL!
Intel launched 6-core & QC 32nm Xeons in March 2010, along with 6-core desktop parts!
Yeah, they had LOTS of trouble!
GloFo blows, and shows every indication of falling further and further behind.
Nice try. They intentionally priced the desktop parts at $1000, so that nobody would actually buy one, because they couldn’t make very many, it was marketing tricks.
Clarksdale Xeons were dual-core, not quad, and the Gulftown Xeons were equally unavailable in significant volume, just as the desktop parts were.
Yeah, yeah. Because GloFo can’t fab 32nm to save their ass, Intel MUST have had similar problems…
It is not that they “have had to have” it is they HAD.
Same for GloFo and same for TSMC before that (40nm TSMC is pretty close to Intel 32nm in cell size).
How do you reckon they can’t fab 32nm? I see Bulldozer and Llano in stock everywhere, and there’s nothing wrong with the performance of the silicon. That’s really pathetic FUD, to suggest there’s a shortage when there is none, or that I should care what percentage of the wafer was viable. Analysts also complain when there’s too much stock, you just can’t please some people.
I have been in hardware, software, and telecoms for 20 years.
If I could sum up what I have learned thus far it would be a couple of simple lessons:
1) Fish rots from the head.
2) When a leader leads, the followers follow.
Read, from ibm, well, beam me up, softee’. Pbvious, ploy like GM, maybe 100 bil or trill might do. Something like the Inquirer, taking hits everywhere, is ror meg whitman. obvious question.
Just as AMD is Getting Good. Striking Board might be Nice. seems enlightened or new dumpster com. somebody call in clowns.
how about stanley barton, reeffect CONTROL DATA, make drashek, ummmm:GOD.
Foreshorten tailwind, ForeSail, mast, Gibber Timbers, theSHIP is SINKING.
Bulldozer II, Bulldozer II. Fanbois’ Splish Splash,
Hey, Go Retail.