AMD Layoffs Maul Marketing, PR Departments

MojoKid writes "AMD's initial layoff announcement yesterday implied that the dismissals would occur across the company's global sales force. While that may still be true, it has become clear that AMD has slashed its PR and Marketing departments in particular. The New Product Review Program* (NPRP) has lost most of its staff and a Graphics Product Manager, who played an integral role in rescuing AMD's GPU division after the disaster of R600, also got the axe. Key members of the FirePro product team are also gone. None of the staff had any idea that the cuts were coming, or that they'd focus so particularly in certain areas. These two departments may not design products, but they create and maintain vital lines of communication between the company, its customers, and the press."

Comments

Bye markedroids

[Anonymous Coward]

Honestly they haven't been performing and it's understandable they got the axe. Maybe now AMD can focus on product rather than image.

Re:Bye markedroids

[ackthpt]

Honestly they haven't been performing and it's understandable they got the axe. Maybe now AMD can focus on product rather than image.

In my experience image sells more often than brand. Particularly image establishes brand, for what it's worth.

These look like the sort of cuts of a company which may be in particular stress. Not encouraging.

Re:Bye markedroids

[Chris Burke]

In my experience image sells more often than brand. Particularly image establishes brand, for what it's worth.

Yes, but the people doing that for AMD haven't exactly been doing a stellar job over the years... Their marketing messages have been constantly changing, and each version was a muddled mess.

Not saying they deserved to be sacked or anything... Just, marketing is not one of AMDs strengths and I don't think this will cost them as much as one might think.

These look like the sort of cuts of a company which may be in particular stress. Not encouraging.

I'm encouraged that they cut heavy on marketing and less on R&D, as opposed to the opposite. That'd imply they aren't planning on being competitive, ever, and just want to suck as much money out of existing products as possible before the end. Reducing the marketing but keeping as much R&D as possible implies they want to be around for the long haul.

Of course you do need both R&D and marketing, but if you gotta cut (and with their current revenue and cost structure, I can't say they don't) then this seems like the smarter way to me.

I'm even more encouraged that the riffs supposedly hit the executives disproportionately as well. AMD has a kinda top-heavy structure and I'm sure that there's some chaff that can be cleared out, and with better buck-for-bang.

Re:Bye markedroids

[rbeef]

/me thinks they will contract marketing out. Cutting marketing and outsourcing is the lesser of the two evils.

Re:Bye markedroids

[MagusSlurpy]

Honestly they haven't been performing and it's understandable they got the axe.

The products haven't been performing lately either - hell, the marketing people did their job too well, and got us thinking that Bulldozer might actually be worth waiting for. Oops.

Re:Bye markedroids

[hairyfeet]

Actually if you have been keeping up with events of late AMD has been selling so many chips that the only thing holding them back has been trouble from the manufacturing getting up to speed on the latest die shrink so I really doubt marketing when you are selling out of chips already is REALLY needed that much, do you?

I mean look at the AMD Brazos line, they have those things in everything from netbooks (here is the one I personally sold my MSI Wind for after picking up a few for customers, it hold 8Gb of RAM and is sweeeet, both on performance and battery life) to laptops like this one with B-Ray to cool HTPC designs to these really awesome all in ones which I found make pretty killer SOHO/small business and family PCs.

So actually I'd say AMD was on the right track when their CEO announced that they were slowing their desktop output to ramp up mobile chips to try to fill the demand. Frankly even if they hadn't had the problems with the supply chain I doubt seriously they'd be needing much in the way of advertising ATM. Right now with the economy down prices trumps just about everything and the bang per buck was in the AMD camp even before the APUs hit, now you have machines that'll play WoW and smooth HD video at frankly insanely cheap prices and get 6 hours on the battery for the mobiles and not heat up the house on the HTPC and all in one. Seems like a good combo to me.

Of course this isn't even bringing up the next "big thing" from the AMD camp which most of the number crunchers and programmers here ought to be drooling over and that is the switch from VLIW to vector in their APUs and GPUs which should bring floating point math a hell of a speed boost. Oh and for all you FOSS lovers out there AMD is switching to Coreboot so you'll have a system that can be open and modified from the BIOS layer on up.

I do have a question about TFA though....what was wrong with R600? Sure the 2xxx and 3xxx series didn't slaughter Nvidia but they also didn't crank the living hell out of the heat nor did they have the whole "bumpgate" issue Nvidia had at the time. Their IGP version of the 3xxx was also quite nice for HD video and the HD38xx was pretty sweet and was easy to crossfire. So while I wouldn't call it a second coming or Nvidia killer it certainly wasn't up to the level of the Nvidia 5xxx, aka the Hoover card fiasco.

Re:Bye markedroids

[V!NCENT]

The people that run the business at AMD are all geeks, and very good ones.

That said AMD is probably the first company that gets it (next to Apple, for that matter) in the sense that the product that you're making should sell itself.

I'm not a fan of Apple at all, but the point here is that I've never bought something because of an emotional response (marketing) but rather because I wanted something for a reason, like: I need something to do x, what is it and where can I get it?

AMD has already done lots of thing that keep me a customer, like for example their GPU and chipset documentation, helping FLOSS developpers create GPU drivers and Coreboot.

We all know Intel pwns the market. That's marketing you cannot beat in the short term. But the best marketing is word-of-mouth and is a long term investment anyways. Companies that are in it for the long term end up being the healthiest and the best.

And it's not as if PC gamers weren't checking the benchmarks first in order to buy the best card. Same with the admins who want to have a BIOS replacement.

Re:Bye markedroids

[tqk]

... I really doubt marketing when you are selling out of chips already is REALLY needed that much, do you?

I'll be watching to see if this has any effect on their sales. If this keeps up, then performance and reputation is what sells (as I've always suspected), not pissing money into the wind on advertising and PR.

Re:Bye markedroids

[Runaway1956]

Personally, I hate marketing. I hated commercials when I watched television, I hate adverts in my newspaper, I hate them on the tubez. Today, with this wonderful internet we have, when I want something, I start searching.

My youngest kid decided that a bike would be cool. He thought about a Harley. I told him that A; Harley is overpriced by an order of magnitude, and B; V-twins suck ass. He did some research, he half believed me, but still, the offer of a trade was just to good to pass up.

Now, three months later, he's believing me, and he's researching AND verifying his research with test rides on other bikes.

He came home last week, exclaiming about how much POWER just a little bitty Kawasaki 250 has, and how FAST it is, and how SMOOOTH it rides. He went on and on, and on some more.

Marketing? Screw marketing. Unless I get to ride the girl in the commercials, marketing means shit to me - and the son is learning the same thing. Just give us the internet, so we can compare specs, features, and prices. The corporations would do better to save all that money they are wasting on bubble headed blonde wannabe actresses who sprawl half naked across their products.

Re:Bye markedroids

[hairyfeet]

I'd say the big advantage right now that AMD has and really needs to pursue is the mobile division which has just about the perfect combination of power+price+ energy efficiency. As I said I sold my MSI Wind (which actually was a nice netbook) and got one of the new Brazos EEE netbooks. When I was stuck at the doc's for half a day i took it with me and ended up with half the patients AND the nurses all wanting to ask questions and get a look. It has a great 720p picture, nice keyboard, even running videos for hours it barely got warm to the touch, just a really nice unit. And when everyone heard what I paid for it ($350 WITH an 8Gb of RAM upgrade AND a briefcase for it) I ended up having to write the make/model and website half a dozen times for even the nurses wanted to know where they could get one.

So I'd say they really hit it out of the park on the mobile line and I can see why they were having trouble meeting demand for the chips. They are low power, cheap, and do the things folks want to do which is play the occasional game or watch HD video. while they haven't had much luck on the desktop front with Bulldozer (I predict like the original Phenom it'll take them a little while to get the bugs out of the new design) I'd say both their mobile APUs and Phenom/Athlon mobiles are in the sweet spot when it comes to the all important Price+Performance+Efficiency.

So I'd say it isn't so much performance + rep as you say but the metric I outlined above. Folks want a good deal, with a machine that will let them play their FB games or watch HD Youtube videos while not breaking their wallet on either the price or the electric bill. my oldest decided he wanted something bigger so he got a Phenom II Mobile 17 inch and all the kids at college have been quite impressed by his. He's sent more than a half a dozen of them to me to "hook them up" with similar laptops, as it lets them do their classwork and then kick back in the lounge for a little TF2 or Second Life for the ladies.

And in the end it ALWAYS came down not to raw performance, but would it do what they wanted to do with it smoothly without breaking their bank accounts. And that is where I think AMD has the advantage, their mainstream chips are a much better deal while still giving good video and gaming performance. If you haven't tried one of the Brazos netbooks give one a spin, they really are quite good.

What is the disaster of R600?

[Etherized]

I'm missing the context here; could somebody explain what this disaster was and how it threatened the existence of the GPU division? A quick google returns nothing.

Re:What is the disaster of R600?

[Anonymous Coward]

R600 was a huge, hot, and expensive design. It had to be delayed due to it being impossible to release on the 65nm process that was available at the time, and it barely fit on the 55nm half-node either.

All AMD (ATI) cards released after R600 have been build from the ground up to target the mainstream market, whereas in the past they would create big monolothic dies and then cut them down to fit the lower markets. The enthusiast slots from AMD are now filled by dual-GPU cards.

A parallel would be Intel moving from the P4 to the core design, where they explicitly design the chip around efficiency, not speed.

Re:What is the disaster of R600?

[Anonymous Coward]

After spending lots of area and design time on the R600 to make this "ring-bus" to get good memory performance, basically someone at Ati f'd up and accidentally implemented the design of the R600 ROP w/o a pipeline (basically get a batch of pixels, crunch on it, output it, instead of pipelined like get a batch of pixels, crunch on it, and get the next batch of pixels, output the first batch, crunch on the second batch, get the third batch, etc, etc.). Although perfectly functional, the perf sucked big time (compared to the nvidia 8800 which was available about the same time and didn't make that kind of silly mistake).

Through lots of software hacks and their marketing group twisting developer arms (having developers do massively custom AA modes or huge shaders where the abysmal rop performance didn't matter as much), they managed to salvage the situation from their crappy design mistake... This was highly fortunate as OEMs that purchase the midrange chips often use game benchmarks to select cards for various price points and if the game benchmarks showed say 1/3 the perf of a comparable nvidia card, they wouldn't sell many cards. That would have probably happened if all the benchmarks were ROP limited and they didn't use lots of MRT hacks to get better perf out of their ROP.

Since ATI was losing money at that time, it may have been the end of the rope for them. They had just made an aborted R500 design (which they eventually salvaged by selling it to MSFT for Xbox360) and they were hoping to have a killer product on their hands and suffering through the illusion that nvidia wouldn't show up with a unified shader DX10 part. The resultant R600 wasn't good for ATI (bad slow rop made bad benchmark scores and nvidia G80 design was unified dx10 despite what the pundits thought at the time), but saved them long enough to be bought by AMD...

-Anon

Re:What is the disaster of R600?

[Anonymous Coward]

R600 (Radeon HD2900) was even made on 80nm, Radeon HD24xx and HD26xx were made on 65nm and 55nm were used by the HD3000 and HD4000 series

Re:What is the disaster of R600?

[Anonymous Coward]

"basically someone at Ati f'd up and accidentally implemented the design of the R600 ROP w/o a pipeline"

Reference link, please? I don't remember that it happened quite that way.

Re:What is the disaster of R600?

[Anonymous Coward]

Of course they wouldn't publically admit to this problem, but many reviewers were puzzled by the choices that ATI made at the time to implement fast path AA resolve when samples were fully covered (for example here). These choices were basically because that path in the ROP was not pipelined properly (and thus slow)...

Anyhow, they made a diving catch for that and that's water under the bridge now days...

Good?

[Xanny]

AMD's weakness is not in getting brand recognition, every major PC carrier knows who they are. They need a competitive product. That requires engineering investments and hard work to catch up to soon to be Ivy Bridge. Servers don't want a power inefficient processor, and power users want top class for the price, and AMD is delivering neither right now on the CPU front. They also shouldn't try entering any other markets, I imagine that is what they are thinking though, try to get out of the x86 business since they are falling behind. Hopefully the Radeon 7000 series does really well, this next GPU generation is shaping up to be a huge force in the massively parallel server market, and AMD better realize the opportunity they have right now to earn back some cred with a rock solid GPU lineup. It doesn't help that Nvidia, AMD, and every ARM manufacturer are all basically waiting on TSMC for bulk 28nm transistors. They are all starting to feel the heat for depending on one company for all their silicon for this next gen of graphics hardware.

Re:Good?

[blair1q]

TSMC isn't the only fabber.

Rumor is that AMD and ARM may team up. But this means they might be thinking of an ARM/ATI combo chip. Which would be verrrrry interesting. But it would leave AMD's x86 department out in the cold for the future of computing.

It's also a clue as to why AMD dumped the marcom hacks: these are the people who are supposed to tell the bigwigs what the Next Big Thing is going to be, and they have consistently been 1-2 years behind the curve.

The only place AMD has been approaching the bleeding edge is in graphics, where the ATI engineers are merely advancing their skillz as fast as they can. No need to guess where their market is going, since there's always a call for more cores and more clock.

Re:Good?

[Sir_Sri]

AMD and its graphics subdivision also rely on globalfoundries, which used to be the production arm of AMD.

If anything, in the long run, the ARM business might help AMD and nVIDIA even indirectly, a big market for another high end foundry company gives them more potential suppliers.

But I agree, marketing handles itself if you have a good enough product right now. The 6000 series is good, but not revolutionary, but I suppose one could say exactly the same thing about the 500 series from nVIDIA. I wonder if they're both waiting for dx12 or next gen consoles to do much with. When the biggest games launching this year all still have dx9 codepaths except battlefield 3, you know you've got a tough market to break into with the next gen hardware. (COD3, star wars TOR, Skyrim and others all have dx9 support in them, though they also have dx11 paths).

Re:Good?

[hedwards]

Competitive products aren't the issue. Well, right now they are, but they do have issues with name recognition. PC carriers are not going to integrate AMD products of there isn't demand for them. And I've been shocked that AMD still doesn't bother with advertising the way that Intel does.

There have been periods where AMD chips were better than Intel chips, and yet that hasn't ever been reflected in market share. Right now, the move is the correct one, cut marketing and focus on developing better products, but history suggests that focusing on products alone isn't going to be enough.

Re:Good?

[0123456]

There have been periods where AMD chips were better than Intel chips, and yet that hasn't ever been reflected in market share.

It's kind of hard to massively ramp up market share in a short time when that requires spending billions of dollars on building new fabs to churn out those new chips. If AMD had built new fabs when the Athlon-64 proved to be significantly better than the space-heater P4s, they'd probably have finished them sometime after Intel release the Core line and blew them away.

Re:Good?

[marcosdumay]

"No need to guess where their market is going, since there's always a call for more cores and more clock."

What is different at the CPU market? More cores, more clock, less power consuption, price (you forgot the later two). Ok, there was the change to 64 bits, and the doubt about ARM vs. x86*, but if they had an entire team just to figure those, they were really throwing money away.

* ARM vs. x86 is easy. What plataform will be better in cores, clock, power consuption and price? Each will win the markets that need whatever of those three they are best at. If you are the one designing the chips, it changes from a prognostic to a decision.

Re:Good?

[leathered]

Sadly for AMD I feel their time has come and gone, for a number of years they were well ahead of Intel but we all know why they could never press home their dominance. A billion euro fine; despite being a record, was chump change for Intel and in reality they profited greatly by shutting AMD out of the market.

Re:Good?

[Chris Burke]

AMD did start building fabs when the Athlon64 and Opteron were kicking ass all over, and when their projections of market share showed that they would be fab limited -- which for a while, they were.

The problem is that when they opened up the flood gates on their production capacity, the market share didn't follow. It bumped slightly, but not nearly enough to justify the massive investment in the fabs, wrecking their financials and ultimately forcing them to spin off the fabs as Global Foundries. This is due to the backroom deals Intel had with OEMs limiting the amount of AMD parts they could sell.

This is the essence of AMD's lawsuit against Intel and the anti-trust rulings by Japan, North Korea, and the EU.

Re:Good?

[Anonymous Coward]

Some nights I miss my old K133 and its solid decoding of 128kbps mp3s.

Re:Good?

[blind biker]

...anti-trust rulings by Japan, South Korea and the EU, I presume.

Re:Good?

[superfastdude]

Please tell me more about the anti-trust case in North Korea.

Re:Good?

[vakuona]

AMD should have bought a computer manufacturer, or a brand. AMD can't compete against Intel by trying to get into Dells etc. They can compete with Intel by making their own desktop and server lines, targeting them well and making decent profits by cutting out the middleman. The profits would allow them to continue to invest in their engineering.

AMD need a strategy that allows them to make money off a small market share without sacrificing their ability to invest in their engineering.

Re:Good?

[Rockoon]

Show me the Intel parts that can compete with AMD's C-50 and E-350. The cost and/or power consumption will be a lot higher to reach performance parity with these two parts.

AMD still doesnt have the market share in the segments where Intel should really be getting their ass handed to them. Why would any manufacturer pick inferior Intel for a netbook or notebook? Makes you wonder.

Re:Good?

[Anonymous Coward]

Because Cores and Clocks is not an effective strategy in CPUs. GPUs perform nothing but massively parallel operations, where every core is used and no pipeline ever experiences a bubble*. A 64 core 6ghz CPU would never get above 1% of its "theoretical peak" in any real world task**, but it would heat your home effectively.

*This has become less true with shaders, which can branch. It is still true the vast majority of the time, however.
**There are rare exceptions, which fall into two categories: the category of "any CPU does this so quickly it doesn't matter" and "people actually use huge farms of multi-core processors, because it is efficient in this case". For more on the extreme case, check out the early days of the Sparc T1 and its descendents.

Re:Good?

[Anonymous Coward]

CPUs have a lot of issues that GPUs don't have. Performance per clock is more important for CPUs because a lot of workloads are single-threaded, whereas GPUs' uses are all embarrassingly parallel. Memory access patterns in CPUs are quite different, and you have lots of issues around the memory model (e.g. cache coherency, different NUMA designs), while GPUs can be optimized for only certain types of memory access. Branch prediction is another issue CPUs have that GPUs don't.

When designing a processor, you have to make tradeoffs on which problems to devote silicon to (e.g. focus on new vector instructions, or focus on faster cache?) and different markets will want different things.

Re:Good?

[Anonymous Coward]

Seems like a load of bullshit to me.

Then what is the purpose of 4x 12-core boxes? Why do some want 48-core servers and could use with a lot higher number of cores?

A 64 core 6ghz CPU would never get above 1% of its "theoretical peak" in any real world task**, but it would heat your home effectively.

The entire "cloud" runs on more than your single-core P4 you seem to cherish. Secondly, your math sucks. Even if you only have 1 core of your 64-core processor loaded, that's 1.6% of your CPU. 1.6% > 1%. QED?

Re:Good?

[Anonymous Coward]

Actually that was not exactly what happened. AMD did have enough cash on hand to build a new manufacturing FAB at NY given the expected government subsidies. The problem is they blew all that cash buying ATI above stock market prices just before the market tanked in 2008. That aquisition nearly ruined AMD. The fabs were spun off in order to raise cash. They basically sold them to a bunch of people in Abu Dhabi. The architect of this "master plan" was Hector Ruiz who would be in jail right now if he wasn't testifying for the prosecution in a massive securities fraud case.

Re:Good?

[Chris Burke]

Your pretty accurate reference to Hector's role aside, that description of events sounds suspect. AMD bought ATI in 2006, when the housing bubble hadn't even burst yet and the financial crisis was way over the horizon. The lack of value of the ATI purchase was factored into the AMD share price long before the market crash -- stock being the main thing they bought ATI with, not cash. There's another thing, too.

You don't pay for a fab with cash, either; they're just too expensive to build and update. It's about financing and investors, and they want to see that money invested in the fab specifically is going to pay off. ATI only affected that to the extent that the promise of Fusion wasn't enough to convince investors, but that's only a minor factor compared to the fact that the previous investment in fabs wasn't paying off for the CPU business. If AMD's market share had grown as it would have been expected to in the absence of Intel manipulation, then the presence or absence of ATI and Fusion wouldn't have mattered.

Which is why AMD didn't sell the fab for cash since cash wasn't the problem. They spun it off because they couldn't get investment in the new fab tech that they absolutely need as long as said fabs were tied to making only AMD products. As a foundry, though, the technology had value. At least to those in Abu Dhabi. The point was someone was investing in the fabs again.

Re:Good?

[splerdu]

I would mod you informative for your post, and funny for north korea.

If I had the points.

Not that bad then

[Anonymous Coward]

Better marketing than engineering I say

Re:Not that bad then

[hedwards]

Ultimately, they need both, but right now they need engineering far more than they need marketing. Intel has a significant advantage in that they sell far more chips and can afford to spend more money on bribes and development

Re:Not that bad then

[MagusSlurpy]

Considering how much higher quality Intel chips have been the last two years, they don't even need to bribe anyone.

Re:Not that bad then

[Zancarius]

Considering how much higher quality Intel chips have been the last two years, they don't even need to bribe anyone.

That's true.

The unfortunate truth for Intel, though, is that their chips have historically been fairly overpriced in contrast with comparable offerings from AMD. Of course, AMD is beaten into a pulp by Intel's high end offerings and can't even compete in that market segment, but I can't think of anyone much who'd fork out $1,000+ for a desktop processor unless they had a business-related reason to or had more money than sense. Certainly not in this economy clime.

I recently spent a fair bit less than $200 on a CPU upgrading my desktop to AMD's then near top of the line offering. Sure, it wasn't near as capable as Intel's chips (and I've historically always purchased Intel CPUs), but the problem is that Intel's comparable offering was over $100 more expensive--never mind how much more expensive Intel motherboards were. I'm not a regular gamer, I often write code in my spare time, and I have no real need for anything ridiculously fast (just more cores). I did want something reasonably future-proof without breaking the bank, which is why I went with AMD for the first time ever. I know of at least one friend of mine who recently did the same thing. I've been extremely pleased with the results.

On the other hand, I know of someone in my circle of acquaintances who purportedly forked out almost $4,000 to build an Intel rig with their latest CPU offering at the time. Yet, in spite of the investment, he never played anything other than WoW on his hardware. To each his own, mind you, but it seemed like something of a waste to me. He had his reasons, and that's fine.

I expect Intel to continue dropping the price on their low- to medium-end offerings in order to compete, but I also don't expect to see them drop very far since 1) the low end has tighter profit margins and 2) Intel has volume (in terms of production capabilities) plus market share in their favor--don't expect that combination to allow for much generosity on their behalf.

Re:Not that bad then

[MagusSlurpy]

I expect Intel to continue dropping the price on their low- to medium-end offerings in order to compete, but I also don't expect to see them drop very far since 1) the low end has tighter profit margins and 2) Intel has volume (in terms of production capabilities) plus market share in their favor--don't expect that combination to allow for much generosity on their behalf.

On the other hand, now that Intel has AMD up against the ropes, Intel might be content to put the squeeze on them, and take a hit (or just break even) on the low-end chips to push them even further out. Intel is certainly pulling in more than enough cash to buffer a bit of a hit in that market.

Re:Not that bad then

[hedwards]

They don't need to right now, because they successfully drained AMD of customers long enough for AMD to have funding issues with their R&D. The incompetence of over paying for ATI didn't help, but the consequences of Intel's anticompetitive activities are going to take years for AMD to overcome.

Re:Not that bad then

[hedwards]

Intel needs AMD, otherwise it faces even more antitrust suits and a closer eye from the DoJ and whatever agency in Europe does the same thing. I'd expect that Intel won't make the mistake of squeezing AMD completely out of the market.

Intel definitely could do it, but they'd end up being broken up if they managed to succeed. There's only so much anticompetitive activity that the DoJ can turn a blind eye to.

Vital?

[vux984]

These two departments may not design products, but they create and maintain vital lines of communication between the company, its customers, and the press."

Better to cut marketing and the "vital" line of communication to the press, than to cut product development and not have a new product next quarter... because then having lines of communication to the press won't seem so vitally important anymore.

Still it sucks for anyone to lose their jobs.

Re:Vital?

[v1]

It can't be easy to determine where the cuts are going to be made if you've decided to not just do an even across-the-board cut. Most divisions within any sizable company could have good arguments made for not selecting them for cuts.

Unless your company has a "wing" ripe for picking anyway. Something that can be cut out all at once like a tumor without much effect to the rest of the company. You just have to hope you have quality bean-counters working closely with the company directors/visionaries to determine the best places to trim the fat. Mistakes get made when one or the other of those two groups has too much say in the decision. One way you shoot yourself in the foot, and the other way you just slow the trip down the drain.

Re:Vital?

[Anonymous Coward]

Hi, a member of the press here (proper press, not a blogger);

So the issue is that the NPRP was responsible for providing support for us when reviewing products at launch. This meant tracking down bugs, letting us know about internally known issues, and getting drivers issued for important bug fixes. The fact of the matter is that pre-launch hardware (particularly for a new architecture) is practically beta testing, and we're the beta testers. The risk AMD takes by not having a well staffed NPRP is that if we encounter issues on future products and are unable to get the fixed, then those problems are going to end up being published.

Or to put this another way, the NPRP was the last line of defense against a bad review. In the short term this won't hurt AMD, but in the long term it makes botching a launch more likely, and that's a stigma that's would stick to a product for its entire life.

Re:Vital?

[Grave]

NPRP is the last line of defense? Cool. Glad they didn't cut the first line of defense, which is the engineer team that is responsible for creating a superior product in the first place.

Re:Vital?

[del_diablo]

It is easy: You cut down to the entire companies managmants wages to lowest engineering wage, and no bonuses. That includes the stockholdes, CEOs and other "high positions".
It wouldn't surprise me one bit if that would earn them a really nice surplus of cash, which again could be used to massive amounts of R&D.
Of course, no corporation these days wants to sit down and do what needs to be done.

Re:Vital?

[lexman098]

that's a stigma that's would stick to a product for its entire life

I think that depends entirely on the consumer. Personally, I go for the best (current) performance for the lowest cost. I realize that getting your CPU in a larger share of mainstream computers from Dell etc may be affected by general perception, but AMD is already losing on that front. Intel has a much larger desktop market share and, anecdotal as it is, most people I talk to favor intel for seemingly emotional reasons. I believe that the perception of the "enthusiasts", however, are ultimately what sways the PC market and it seems these are the people that AMD is now aiming to impress.

In short: CPUs aren't exactly like vacuum cleaners when it comes to marketing.

Re:Vital?

[FreakyGreenLeaky]

If that's the way you want it to work, then move to North Korea.

Re:Vital?

[Anonymous Coward]

Honest question, do you know anyone who would manage for the same salary as their workers? I know I wouldn't. Management is really hard. I would personally much rather be doing the programming rather than managing programmers, so I wouldn't want a management job if it only paid the same. I have the impression that all my bosses are the same way. They would rather be programming.

Re:Vital?

[Anonymous Coward]

If the product is bad, it deserves a bad review.

If one "journalist" gives a bad review for a good product, others might not. And in the end the one with a bad review will lose credibility.

Why not just let the product speak for itself? Let the customers taste the chicken soup instead of trying to wave some rubber chickens and make chicken sounds.

Re:Vital?

[Anonymous Coward]

stockholde(r)s don't get paid wages.

Re:Vital?

[Anonymous Coward]

It's actually a pretty easy decision. I worked at a startup during the financial crisis. When we had to let people go, we followed the maxim that a company only really needs two kinds of people: people who make the product and people who sell the product. Consequently, the prime target for layoffs was the marketing department, because they were neither engineering nor sales, nor were they already on a skeleton crew like our operations department (though our layoffs eventually turned marketing into a skeleton crew).

I assume AMD is following the same rationale.

Re:Vital?

[vux984]

You cut down to the entire companies managmants wages to lowest engineering wage, and no bonuses. That includes the stockholdes, CEOs and other "high positions".

Or maybe if they cut the engineering wages to that of the lowest janitors wage and no bonuses. What do you think that would accomplish?

A big fat surplus of cash? Of course not. All the engineers would find new jobs and quit.

What makes you think management would be any different?

I agree executive compensation is way out of whack, but you can't just cut their wages to below market norms or they'll just go find new jobs like anyone else.

Re:Vital?

[vux984]

Or to put this another way, the NPRP was the last line of defense against a bad review.

Not having a new product to review in the first place is far worse.

Marketing of tech is almost free.

[blair1q]

Seriously, with 14 bazillion bloggers fighting to get clicks to their webpages, all you need is one guy with a copy of the datasheet and a twitter account, and you'll have your part's nomenclature showing up on every RSS feed in the world within minutes if not days. And, if you're lucky (or just know where to put the typos), you can get /. to send your favorite blogger enough clicks to buy an iPhone.

Re:Marketing of tech is almost free.

[hedwards]

Yes, but those aren't the people that AMD needs to be reaching. To people that know little about computers, Intel is a name brand that has been associated with quality. The problem is that it isn't always true, there are periods where Intel is doing really good work and there's periods where AMD chips are better, but you don't really ever see that in the market share, in large part because for the most part you have to build your own computer if you want AMD parts.

Not quite so much now that AMD does GPUs as well as processors, but an AMD processor can be relatively hard to come by in a mass produced machine.

Re:Marketing of tech is almost free.

[westlake]

Seriously, with 14 bazillion bloggers fighting to get clicks to their webpages, all you need is one guy with a copy of the datasheet and a twitter account.

If this were true why is Linux clinging by its fingertips to a bare 1% market share?

You need people who can negotiate OEM system installs, retail placement and sales promotions. Your bazillion bloggers aren't as useful as the one man or woman who knows how to cut the right deal with Walmart.

Re:Marketing of tech is almost free.

[blair1q]

Okay, so in addition to your twitter guy, hire a guy to call OEMs and say "use our chips and we'll give you free stickers."

And a guy to negotiate for a NASCAR team. Because, fuck, man, this is Amurrca.

Fact is, if AMD had ONE THOUSAND FOUR HUNDRED people doing that job, they were wasting about $140 million a year on dead wood.

Re:Marketing of tech is almost free.

[hedwards]

It's not a fact. That's how marketing works, it doesn't matter how incredibly brilliant and affordable your product is if nobody knows it exists. And for most people, AMD products just aren't available when they go to the store. And they don't know about them because there's no marketing and they often aren't carried by the stores.

Re:Marketing of tech is almost free.

[Anonymous Coward]

Oh shit son, did they just unthaw you?

Linux doesn't have 1% marketshare of anything. It has closer to 10% desktop marketshare and dominates the mobile, server and embedded spaces.

Re:Marketing of tech is almost free.

[Kjella]

The problem is that it isn't always true, there are periods where Intel is doing really good work and there's periods where AMD chips are better, but you don't really ever see that in the market share, in large part because for the most part you have to build your own computer if you want AMD parts.

Well that, and you have the problem that it's not that easy to ramp up or ramp down CPU production. Building fab capacity is started years in advance so by the time AMD actually brings a processor to market they got a fairly narrow percentage of the market they can supply. Good chips mean high prices, poor chips means low prices but they can't take that much market share in one generation. And if you overextend yourself you risk that Intel pulls a very good processor out of the hat and you're left with way too much fab capacity and has to dump prices to get it moving. Or maybe that you spend too much money on fabs and not on R&D so that you got plenty capacity and a poor product next generation.

That and one thing that matters to producers, not so much to consumers and that is Intel has often been a half to one generation ahead in processing technology. That translates directly to higher margins and more money to make new factories and more R&D. Even when AMD had successes they weren't making great money. And in the CPU war the winner is the one who can keep the pace of improvement up, not who wins at the moment. In a year that generation is gone and it's a different battle.

Re:Marketing of tech is almost free.

[Anonymous Coward]

gonna call bullshit on "10% desktop marketshare"

Re:Marketing of tech is almost free.

[fluffy99]

Oh shit son, did they just unthaw you?

Linux doesn't have 1% marketshare of anything. It has closer to 10% desktop marketshare and dominates the mobile, server and embedded spaces.

Recent surveys by respectable groups like Gartner put Linux at 1.1% share on the desktop.
http://en.wikipedia.org/wiki/Usage_share_of_operating_systems
http://marketshare.hitslink.com/report.aspx?qprid=9&qpcustom=Linux

Linux is still hovering around 20% in the server market, hardly what I'd call dominating. Sure Android is based on a linux kernel, but Iit's a stretch to call it Linux. I will believe you on embedded systems though.

Re:Marketing of tech is almost free.

[hairyfeet]

Uhhh...where have YOU been? Go into Walmart, Best Buy, Staples...you see a fricking TON of AMD machines being sold. Hell last time I went in there I think more than half of the netbooks and laptops had AMD Vision stickers on them.

I thought the whole "vision" idea is damned smart but whomever they had in marketing dropped the ball. Most folks don't know gigawhatitz and megawhositz they want to know "What can I DO with the thing? and as a small retailer I can tell you THAT is what you use to sell machines in this economy! You show them how easy it is to transcode video for their mobile devices, you show them how nicely it plays Farmville and WoW and TF2, you tell them how this netbook that weighs only 3 pounds and is cheaper than the other guy's stuff lets them watch THREE hour and a half movies in HD on a charge!

Frankly the marketing bunch deserved to be fired because they had this great idea and dropped the ball. All I ever saw about vision was a sticker telling a couple of bullet points, that's it. I'd have partnered with the OEMs and retailers to have those display models playing a little cool video telling the folks all the things these chips can do FOR them. That was the mistake here folks, marketing wasn't doing their jobs!

Re:Marketing of tech is almost free.

[Anonymous Coward]

Err, what? A system based on the Linux kernel is the *only* reason to call anything Linux. Of course Android is Linux.

Re:Marketing of tech is almost free.

[qualityassurancedept]

If you have a superior product, then the approach you outline might work. Just let the bloggers tell everyone what they honestly think. But when it comes to Intel vs AMD practically everyone would say Intel is better except for those who are think price is a decisive factor. So, when you go to Best Buy and you see row after row of laptops with Intel chips and the sales people are telling you the battery will last 6 hours and they processor speed is this or that and its all because of the super efficient Intel chipset then who on earth is going to go for that refurbished Gateway with an Athlon II in it? The marketing team didn't so much fail to sell the consumer on AMD but rather they failed to sell the major computer assemblers/resellers on AMD. So now everyone wants an i3, i5, i7 in a new computer and that right there is a result of the failure of AMD's marketing team. They OUGHT to be held accountable for that failure.

Re:Marketing of tech is almost free.

[mikael]

A simple CPU performance comparison chart would work

But how can anyone tell whether one CPU is going to be fast or slow for their purposes? There's all that hyperthreading, cache size, memory size to consider as well.
You'd really need to know how many MHz and Megabytes a particular application requires in order to run smoothly.

My parents have a laptop (1.5 GHz) that is so slow on starting up, that they just basically keep it on all the time, and don't bother shutting down the applications (E-mail, spreadsheet, web-browser). On the times that they do switch it on, they go off to make a cup of coffee, make some telephone calls, and when they're done, the PC has booted up.

Re:Marketing of tech is almost free.

[gbutler69]

because there's no marketing and they often aren't carried by the stores.

So, just WTF were the 1,400 marketers doing then? Apparently not their job. Maybe that's why they were purged.

And...

[GameboyRMH]

Nothing of value was lost.

Re:And...

[Anonymous Coward]

The laws of gravity still applied and AMD had no where else to go but down!

Re:And...

[Anonymous Coward]

...and it's true in this case as long as the products remain good enough to sell themselves. AMD/Radeon has to keep doing what it's been doing in low-budget value segment. Word of mouth does wonders.

Marketing is only really necessary if you have yet to establish yourself as a business, or if you have many players vying for your piece of the pie. But when it comes to CPUs and GPUs, its not like you're going to get drowned out of public conscious anytime soon.

The rest is P.R., but that can take you only so far if you don't have anything worth selling.

What exactly would you say you do here?

[Anonymous Coward]

.........

Does this remind any one of...

[catmistake]

Office Space?

These two departments may not design products, but they create and maintain vital lines of communication between the company, its customers, and the press.

Bob Slydell: What would you say ya do here?

Tom Smykowski: Well look, I already told you! I deal with the goddamn customers so the engineers don't have to! I have people skills! I am good at dealing with people! Can't you understand that? What the hell is wrong with you people?

Amazing

[liquidweaver]

Here it is, 2011, when CEO's live and die by 10K's and stock prices, we have a company that layed off marketing and PR and kept their engineers. How much AMD stock can I buy? Sign me up!

Re:Amazing

[CAIMLAS]

Absolutely. This is a "bold" (ie, different) move. I've always liked AMD as a company, as their business decisions have always at least been long-game driven. "It hurts now, but two, three quarters from now, we'll like it".

I was starting to think AMD was going to fall way, way behind. Now, I think they're going to pull ahead of this one. (Hell, it took Nvidia one major revision in their cards to get from "poor performance and high power use" to "the head of the pack by a bit". From what I understand, Bulldozer isn't designed poorly - the implementation is just lacking. Sounds to me like they pushed a beta product out for quarterly product presence, but the real product isn't far behind...

Re:Amazing

[hedwards]

I'm not sure what's so funny about this post, it's clear that they understand that engineers in a company that depends upon creating new products can't cut back on engineering indefinitely.

Re:Amazing

[Anonymous Coward]

In most cases, losses to Marketing are not a big deal, hell Berkshire Hathaway could probably cut down half their advertising budget and you'd see only slightly less Geico and Dairy Queen commercials, but you'd still know they sell insurance and ice cream.

AMD sells CPU's, and GPU's. I know they sell other stuff, but nobody ever talks about the other stuff. I think that tells you all you need to know.

From an engineering POV, the only reason I buy the Radeon's but not the AMD CPU's is because the CPU's tend to be one to two ticks behind Intel. The only time they were ahead was during the Pentium 4 /Rambus boondoggle. If AMD can see the future, they should apply their CPU engineering to releasing a CPU/GPU part similar to what HP is going to use in their servers, but release it in a end-user package so users can start developing ARM software. Intel doesn't build ARM parts, but they did at one point for a short period of time when PDA's were hot. Tablets are just PDA 2.0 . I mean, if they could build a 99$ CPU+GPU that a DDR3L module can be dropped onto a motherboard, there might be some serious development. Right now people are waiting for Windows 8. Linux already runs on ARM, FreeBSD runs on ARM. Stack 150 of these hockeypuck sized things into a server and you consume less power than one Intel server.

Re:Amazing

[marcosdumay]

Except for the repetitive redundant part where you keep saying "engineers, engineers, engineers", lots of companies seem to not understand that at all.

Re:Amazing

[gman003]

From what I understand, Bulldozer isn't designed poorly - the implementation is just lacking. Sounds to me like they pushed a beta product out for quarterly product presence, but the real product isn't far behind...

Actually, a huge part of Bulldozer's problem is marketing lies. The architecture is very interesting - it's based on a "module" made of an instruction fetcher/decoder, two integer cores, a floating-point core, and two levels of cache. The effect is comparable to Intel's Hyper-Threading, even if the implementation is different. A four-module Bulldozer chip is comparable to a hyper-threaded quad-core Intel chip - it can ALWAYS run four threads at once, and can theoretically reach eight.

The problem is, AMD didn't market it that way. They market their four-module chips as 8-core, and their two-module chips as quad-core. Which isn't, technically, lying - they do have that many integer cores - but that marketing caused problems when benchmarks came out. People saw "AMD 8-core chip beaten by Intel 4-core chip" and thought "man, those cores must suck BALLS. And since even I know that a lot of programs are still single-threaded, it really makes no sense for me to buy an AMD chip right now".

It's almost justice, seeing the marketers fired for this. They stretched the truth beyond what the public would believe, and it bit them in the ass.

The other problem with Bulldozer is pricing - Bulldozer chips, at least right now, are ~$30 more expensive than the comparable Sandy Bridge processor. Sure, you'll quite likely save twice that if you're upgrading, since Bulldozer is mostly compatible with older motherboards while Intel is still thrashing sockets, but that's not going to be the case for everyone.

Re:Amazing

[the linux geek]

I can't think of very many situations where having a four-instruction decoder in front of two four-issue cores would be a good idea, unless you're doing something weird like issuing nothing but vector ops. Additionally, cache latency is through the roof compared to everything else in that space right now, including previous-gen parts.

Bulldozer is a bad design, and AMD's current roadmap (10-15% improvement per year) doesn't look very reassuring. Still, I'd like to see them get their shit back together, because having Intel as the only high-end commodity manufacturer is bad.

Re:Amazing

[Nemyst]

AMD has another problem: its own Phenom line. For all intents and purposes, an older and dirt cheap Phenom II X4 will more than please 99% of the population. For the remaining 1%, performance is an important criterion, at which point they'll probably go for a top-of-the-line Sandy Bridge or Ivy Bridge processor since they're rebuilding every three years anyways, negating the upgrade discounts on AMD's platform.

That AMD is refocusing its strategy towards APUs and possibly teaming up with ARM should be a telltale sign that things are looking dire on the x86 front for them. That's truly a shame, since at the height of their game they could challenge Intel toe-to-toe easily, which is probably what ended up fostering Intel's tick-tock process and the arrival of the Core series (ie very, very good news for consumers). Whether you like AMD or Intel, you need both strong for capitalism to do its job.

Re:Amazing

[MagusSlurpy]

From what I understand, Bulldozer isn't designed poorly - the implementation is just lacking. Sounds to me like they pushed a beta product out for quarterly product presence, but the real product isn't far behind...

I don't know jack about this, so I'll just quote what an acquaintance of mine wrote on my gaming community forum:

The main issue is that the Windows 7 scheduler doesn't understand the effective use of the modules, which drastically cuts down on the ability of the processor to turn off cores and run in turbo mode.

The bottom line in the end unfortunately will be that most desktop work loads still will not take advantage of the Bulldozer architecture. It'll fair much better in the server world but it's going to be a while before the desktop software truly shifts to the style of programming Bulldozer requires. A long while. Probably a lot longer than 5+ years. . .

[I ask a stupid question]

. . .Intel also disables cores. The idle cores are turned off to reduce the thermal foot print while running the active cores at higher clock speeds. Bulldozer's method is a little more complicated than Intel's though, and Windows 7 doesn't understand how to deal with it. For example, for two integer heavy threads with shared data, it should schedule them to a single module and throw the turbo on. Two integer heavy threads, with non-shared data? Two modules. Two floating point threads? Two modules. There's a lot of conditionals about the work load, based on the new longer pipeline, the dual integer unitss, but only a single fp unit.

Piledriver will improve the FP a lot since it'll have the GPU on die, which is what AMD is really aiming to do. They don't really want FP units at all in the module, they want to push that work to GPU-style modules. Which makes sense, as they're a hundred times better at it. But programs aren't written to take advantage of that yet.

I have no idea if he's correct, but that's the extent of my understanding.

Re:Amazing

[lexman098]

Bulldozer is mostly compatible with older motherboards

I think you've got that backwards. The *older* chips (Phenom/Athlon II) are mostly compatible with *newer* motherboards.

Re:Amazing

[gman003]

Goes both ways:

"Some manufacturers have announced that some of their AM3 motherboards will support AM3+ CPUs, after a simple BIOS upgrade. Mechanical compatibility has been confirmed and it's possible AM3+ CPUs will work in AM3 boards, provided they can supply enough peak current. Another issue might be the use of the sideband temperature sensor interface for reading the temperature from the CPU. Also, certain power-saving features may not work, due to lack of support for rapid VCore switching. Note that use of AM3+ CPUs in AM3 boards may not be officially supported by AMD." - Wikipedia

Re:Amazing

[Anonymous Coward]

Tell that to the Linux kernel engineers they laid off. The spin isn't accurate: the layoffs are essentially random.

Re:Amazing

[Kjella]

Except if they called it a quad core, then it'd be the world's biggest and most power hungry quad core. It's got the transistor count and power consumption of an octo core, without actually delivering that performance. I'm not so sure marketing it the other way would have looked any better.

Re:Amazing

[saint0192]

And they recently took the world's fastest processor record for Guinness, I believe. They used liquid helium cooling at -220 C... The actual chip and package are made to withstand high clock speeds and tortuous conditions. They are gonna tune these and make them faster. They need their R & D team to make better sense of their instruction set implementation. That's where Intel is killing them in Ops/second... I still have hopes for AMD, 'cuz competition is good for us, but I use Intel right now...

Re:Amazing

[Anonymous Coward]

It's a fairly effective analysis, with the caveat that you still want to put threads that share data on the same core, even if they're floating-point oriented, as long as they're memory-limited (bandwidth or latency) and not actual throughput-limited.

Kinda reminds me of EPIC -- "As long as you hand code all your assembly and make sure you have no pipeline bubbles and are always using all execution units, performance is FABULOUS" -- yeah, well...thanks? How about a processor design that I can actually use tools, like a compiler and operating system, with? Oddly, with EPIC, the result was AMD's market dominance (and Intel's illegal bribing of OEMs). This time around, the stupid is on the other foot.

Re:Amazing

[tompaulco]

I agree. I have always felt that the ROI for marketing and advertising is somewhere under $1 per $1 spent. The only reason that you have to have marketing is because your competitor has marketing, and even though the competition loses money on every dollar spent, you also lose some amount of money for every dollar they spend because their marketing has SOME effect on your business.
In other words, marketing is necessary because marketing exists. In the same manner, we all hate lying salesmen, but for the same reason, salesman HAVE to lie because other salesmen lie.

Re:Amazing

[Anonymous Coward]

It was mostly in PR and marketing, but not entirely. I'm an engineer who worked there until thursday... They had a room dedicated to firing engineers all day too.

Re:Amazing

[Anonymous Coward]

Well, you also have to ask yourself: how much does a motherboard really matter -- most people don't upgrade their CPUs, they buy new motherboards. So new designs really don't factor in except to the extent that you have more options and cheaper options when purchasing. Systems... In the consumer space, future compatability of motherboards matters a lot more than in the business space, where people don't upgrade their systems piecemeal, they buy new boxes.

Re:Amazing

[jthill]

Getting an OS to observe which threads play nice together is not new technology.

Re:Amazing

[Rudeboy777]

This has turned out to be rarely true in practice. Only about 3 models each of MSI and Asus 800-series motherboards can run a Bulldozer CPU with a BIOS update and NO Gigabyte 800-series boards will except for the very last hardware revisions of about 8 of their models.

The motherboard OEMs did a much better job of supporting new CPUs on their existing boards for the previous AM2 - AM2+ - AM3 transitions.

Re:Amazing

[CAIMLAS]

So, for gaming, why would I want one of these? That's still the question on most people's mind.

Most of the remainder of geeks think: but Bulldozer uses a gobton more power than the performance-equivalent Sandy Bridge chip (and it costs more). My existing AMD system out-performs it for single and double core workloads. I don't want one.

The common consumer buys what's almost cheapest, usually, thinking it's the value proposition. Sometimes that's a good deal, and sometimes it'll be AMD. Or, at least, that's the case historically: how is that going to be possible with the AMDs being more expensive now?

Good.

[Anonymous Coward]

marketing and PR does not a good product make.

Always liked amd because they DIDN'T plaster the world with 'amd inside' too. I don't want to help pay for that shit when i buy a cpu.

Advertisers and marketing people might be the lowest form of life too.

Re:Good.

[qxcv]

Advertisers and marketing people might be the lowest form of life too.

I'm pretty sure that title goes to IP droids as far as corporate lifeforms are concerned. At least marketers and advertisers create something of value (usually).

Re:Good.

[Dunbal]

Advertisers and marketing people might be the lowest form of life too.

Eh, I am sitting here uncomfortably as my wife is head of market research (not marketing) at a Fortune 500, and my sister in law is regional VP of a large media/advertising corporation...

Re:Good.

[edxwelch]

In all fairness, that's a pretty low bar. Nothing to brag about if you say at least you are better that patent trolls

Re:Good.

[MagusSlurpy]

I don't know how long it's been since they've done this, but when I built my very first system more than ten years ago, I put a 1.4 GHz Athlon Thunderbird in the system, and it came with a 1"x1" (or whatever) AMD sticker to put on the case where the "Intel Inside" stickers always went.

Unfortunately for AMD, my case also came with an "Antec Outside" case, which suited my sense of humor much better.

This was back when they were a seperate company

[Sycraft-fu]

The reason it was a disaster was the nVidia GeForce 8800. ATi was pretty sure that nVidia was still going to be back on teh old style of cards, with separate shaders, for their first DirectX 10 part. That is allowed, though not ideal (the programming interface has to be unified, not the hardware). ATi already had experience with unified shaders from the 360.

So from all accounts their not-so-great GPU that was up and coming was going to be fine against nVidia. Then out of the blue nVidia drops the 8800, they did a real good job keeping a lid on it. Fully unified architecture that was fast as hell. We are talkign twice as fast as previous generation stuff often and that was on DirectX 9 stuff, never mind what it'd be able to do with the newer APIs.

So ATi had to delay their release a bit and try to get something to compete better. When the R600 did launch as the Radeon 2000 series, it wasn't good competition.

However ATi recovered very well with the Radeon 4000 and 5000 series. The 4000 series were extremely competitive cards. Good prices, good performance, low power usage, etc. Then the 5000 series were the first DX11 cards on the market by a number of months, and also great performers.

Re:This was back when they were a seperate company

[Hadlock]

Excellent point. It's also worth pointing out that the 8800 survived for five years as a very viable card. Released in 2006, it's still listed as a minimum requirement for many games today (including Battlefield 3). That's quite a feat considering how fast technology matures in this market. In 2009 the 8800-class cards were still selling north of $120, and while not mind blowing by today's standards, were pretty much the gold standard until mid-2008. It's hard to compete against that kind of technology.

Re:This was back when they were a seperate company

[UnknownSoldier]

> However ATi recovered very well with the Radeon 4000 and 5000 series. The 4000 series were extremely competitive cards. Good prices, good performance, low power usage, etc. Then the 5000 series were the first DX11 cards on the market by a number of months, and also great performers.

Yup. What's "ironic" is that the 6970 is significantly _slower_ then the 5970 *, since the 6000 series was even more about low power usage.

It will be interesting to see if the 7000 series focuses more on power or efficiency ....

The 5770 is an amazing bang/buck card.

* "Tom's Hardware : Review battlefield-3 graphics performance"
  http://www.tomshardware.com/reviews/battlefield-3-graphics-performance,3063-8.html

Re:This was back when they were a seperate company

[Anonymous Coward]

Yup. What's "ironic" is that the 6970 is significantly _slower_ then the 5970

Because one is a Dual-GPU part and the other is not. The 6 series equivalent to the 5970 is the 6990, not the 6970 like any sane person would expect.

Re:This was back when they were a seperate company

[MagusSlurpy]

The 6 series equivalent to the 5970 is the 6990, not the 6970 like any sane person would expect.

And who's to blame for that? Marketing.

Re:This was back when they were a seperate company

[rtb61]

Still even counting all high performance graphics cards those number aren't really the big numbers. Keeping focus on design rather than marketing AMD especially after buying ATI, might be focusing on a CPU with a high powered embedded GPU, with separate memory caching but combined main memory usage. The demand is pent up and growing for high powered portable graphics based devices. Tablets, netbooks, notebooks even more efficient smaller desktops. That's where the big numbers are. The first to crack a really good combined GPU/CPU will be on a winner for quite a few years and likely will pick up the lions share of the market.

Re:This was back when they were a seperate company

[saint0192]

And Linux Support came. And Eyefinity rocks. AMD is finally ok with me :)

Re:This was back when they were a seperate company

[hvm2hvm]

Yeah, I have a 9800GT and it never disappointed me yet

Re:This was back when they were a seperate company

[Anonymous Coward]

It is worth noting that the 5970 is a dual GPU card while the 6970 is a single GPU card, so it's not really worthwhile to compare the two (especially considering the 6970 is a cheaper card).

Suspicious?

[Anonymous Coward]

Is anyone else suspicious that this piece of news may have been written (badly) by someone laid off from the PR department? I mean, this line alone:

"These two departments may not design products, but they create and maintain vital lines of communication between the company, its customers, and the press."

Re:Suspicious?

[Anonymous Coward]

Why is the truth suspicious?

AMD execution

[Anonymous Coward]

AMD's most recent shareholder meeting revealed that it has been missing production targets and shortages are occurring. Customers aren't getting the parts they expect, the new stuff AMD has developed, and they're walking away.

Fusion is good. AMD seriously needs to jump on some Taiwanese necks and get the fab stuff fixed pronto. Developing new chips doesn't create a profit if you can't actually build them.

Re:AMD execution

[Anonymous Coward]

AMD seriously needs to jump on some Taiwanese necks and get the fab stuff fixed pronto.

Most of AMD's chips are fabbed by GlobalFoundries. GF is Chartered Semiconductor merged with AMD's former fab operations (spun off to raise cash when AMD was circling the drain a couple years ago). GlobalFoundries doesn't have anything in Taiwan. Try Germany, Singapore, and upstate New York.

Oh yeah?

[Dunbal]

None of the staff had any idea that the cuts were coming, or that they'd focus so particularly in certain areas.

And this is precisely why they were fired. I mean duh, this is not news that marketing is among the first areas to be axed in a dying company. There's quite a bit of precedent in the business world. If those employees didn't even know this, and had no situational awareness as to how their brands were doing, I can just imagine how they were handling their day to day work.

Re:Oh yeah?

[MagusSlurpy]

Actually, a lot of companies are hitting R&D the heaviest, since they require lots of space, supplies, and equipment (overhead) and are often the highest-paid non-management/attorney positions. Usually a terrible, terrible plan for any company as a whole, since THAT'S HOW A COMPANY STAYS COMPETITIVE, but it keeps the stock healthy for a long-enough period that management can cash out and then get the fuck out of Dodge before the house of cards comes tumbling down. For a thrill, keep an eye on the big pharma companies over the next two to three years, it's not gonna be pretty.

good product + marketing = sales

[Anonymous Coward]

take a look a apple... marketing wizards. you may love their products, or hate it, but their growth and sales tell the story for itself. this may not end up to be such a good move for AMD in the long run.

Re:good product + marketing = sales

[Zancarius]

take a look a apple... marketing wizards. you may love their products, or hate it, but their growth and sales tell the story for itself. this may not end up to be such a good move for AMD in the long run.

The problem is vastly different, as many posters here are pointing out: AMD has (had?) a lousy marketing department.

The illustration you're making is apples (heh) to oranges. Apple has an extremely strong and talented marketing wing--so much so that the actual real world quality of their products almost doesn't matter (much). AMD is quite the opposite. Their marketing department has dropped the ball and pretty much failed the company.

PR at its finest

[Anonymous Coward]

Is this their marketing and PR department at work?

Not even close to reality

[Anonymous Coward]

Yes, I am an AMD employee. The base "article" is crap. Maybe it was a troll. And no, this is NOT, repeat NOT a company sanctioned view. And no, I am not a company shill. I'm an engineer who must (and will) remain anonymous. The initial post: "The New Product Review Program* (NPRP) has lost most of its staff and a Graphics Product Manager..." Oooh - a bloated-process product manager got let go - who cares? And "key members of the FirePro product team are also gone". Ok, that happens in non-trivial RIFs. "These two departments may not design products, but they create and maintain vital lines of communication between the company, its customers, and the press." In other words, overhead got chopped. Sorry, I'm skeptical of the base gripe. Yes, it sucks to be RIF'ed. Been there, suffered that. So what if Dr. so-and-so left engineering, or VP useless-whattheirname got sacked? I've seen some articles referencing various individuals as "critical" by web articles that I've never even heard of before. Sounds like a self-serving PR department at work on those. Sort of along the lines of "Paris Hilton sex tape with an Intel CPU" silliness.

As a result

[rossdee]

forthcoming chips from AMD will just have numbers instead of names

AMD = Axe More Demoers

[Anonymous Coward]

Other choices for the third word can be used to describe subsequent waves of layoffs.

Typical Slashdot post

[Anonymous Coward]

Whatever I do, should be highly compensated because you're getting the end result of many years of education and professional experience.

Whatever the other fella does, should be given away for free as it was probably worthless crap anyway.

Indeed

[unity100]

there are actually people still using 4870 cards, to play recent titles in high settings. i noticed that if i had had bought a 4870 instead of a 3870 back in 2007ish, i wouldnt need to change my card last year. even then my 3870 was holding very well, but the dual slot fansink was rather noisy (they always are). so i moved onto a 5670 arctic cooler sapphire, which is so silent that i didnt hear any fan noice for the last year despite overclocking it in taxing games.

ati seems to know how to produce gpus.

Re:Indeed

[Anonymous Coward]

Ironicly, the Datsun 1800 my dad purchased in 1978 (used for ~2grand) produces more megagiggles today.

We Lament the Loss of the Titanic but ....

[Anonymous Coward]

How many Hitlers, Mossuliniis, Geroge W. Bushs and Barak Hussien Obama IIs went down to Davie Jone locker and good riddance to them the bastards all who should never have been born to this Earth.

-- -- -- -- :-)

How many marketers to screw in a lite bulb

[Anonymous Coward]

you'll have to provide your own punch line.

we're out of PR people at the moment.

Some good info Finding Denims

[aprilhappy1122]

Denims have been in design given that Eighteenth century. Right now there's a lot of designer label reach bootcut skinny jeans offered as a result of known brands which have made the mark while in the tradition. Skinny jeans of the specifications lower and even color selection might accommodate just one and then the comparable feminine otherwise, from there exists simply no two equal sets associated with denim jeans.Other than, don't forget that that after you are sporting a pair of jeans for a little bit that they turn into more substantial, that's the reason why you have got to almost certainly obtain trousers a little lesser bigger. Navigate yourself the size of jeans, it is usually showcased by way of a small fraction of only two statistics, the primary a particular this means Moncler Womens Jacket that stomach dimensions, cost-free a particular . . . size. Endure in mind see your face custom-made jean materials company is targeted by itself buyers. This is the reason trousers created by Hand gun, Hugo Superior together with Klaus Montana style manufacturers flows to are perfect for those who prefer conventional units, Versace bluejeans are designed for many who will talk obtaining long decrease thighs and leg, and also Lee Philadelphia and even Ford mustang Unique-Erotic jean materials beats they fit for by individuals that would like to start looking captivating. You may examine any soundness associated with jeans Moncler Men Coat tone running a coordinate trick. Go on a go with, bathe that on mineral water after which it indicated through corduroy. Should the coordinate can get tone, your current trousers will probably take dye while you are having on them. High quality overdyed denim jeans will get rid of hue in the process.To be able to not necessarily attraction the strain account so that you can negative aspects of your respective establish be certain to discover basic guidelines connected with denim jeans assortment. The following are a number of creative ideas that will very well be beneficial to people. Therefore expanse pants end up not being for you to slip on. They're reduce Moncler Vest Women during the higher about half and are usually starting to be tight on the bottom. Trouser limbs could also be upright, convert or complex. Light kit is proposed. Nevertheless, be thoughtful together with lean trousers in addition to magic formula jeans brands. Navigate yourself in the pinnacle specifications certain for the jeans content label.Most women which are needing aspect will need to place on skinny jeans styles which unfortunately beautifully stretch out the particular thighs and legs. Thin, too long together with dishevelled units are it's possible that yet another break down selection for you. To be able to by simply worn pants, favor these kinds of clothes the spot worn components won't be quite clear and are usually nestled adequate inside middle . . . it result will make thighs and legs beautifully more. As well as leave out high heel pumps! Those that have restricted hip and legs in addition to toned glutes ought to put on hipsters * small or possibly excessive-waisted. Those that haven't got some sort of safely and securely proclaimed midsection ought to decide peaceable reach bootcut denims fashions with the help of specifically trouserlegs. These lengthen bootcut denims might aesthetically create your middle look thinner.The best opportinity for an individual should be to invest in skinny jeans a lot of events much bigger around way of measuring, as a consequence of small-sized trousers could really be as well tight for your needs in the waistline a part. And continue to added onto males's jeans fashions : they might meet an individual when nicely because they please have various proportions compared with ladies put on. Soon after studying manner, human body varieties, slashes, flushes and

good riddance?

[Anonymous Coward]

a Graphics Product Manager, who played an integral role in rescuing AMD's GPU division after the disaster of R600, also got the axe

Good to know some psych/communication/visual arts major was the hero in the resurrection of a venerable brand.
Words that he singlehandedly perfected the TSMC 65nm process with his smooth, soothing, baritonic voice.

Yes marketing is pointless!

[Anonymous Coward]

I agree with the sentiment here, marketing and PR teams are so completely pointless in this age. Better to design a product so good that it sells itself and have a few devs do the content and post it across the web and see it flying out of shelves, how much effort can it take to take a product to market?

That's on account of renumbering BS

[Sycraft-fu]

For whatever reason, AMD decided that they wanted to mess with their scheme. In the 5000 and 4000 series, the x8xx part was the high end single card, the x7xx was the lower range (like half as many shaders and so on) and the x9xx part was the dual GPU part: two actual GPUs on one board.

Well for the 6000 series, they changed it. The x8xx range is the same as the x7xx range from the 5000 series and the x9xx the same as the x8xx. So the 6970 is now the highest end single GPU card, and is equivalent in the lineup to the 5870.

Turns out performance wise they are a bit faster for almost everything (different shader setup) though not a ton.

That's where the confusion comes from. If you look at pricing, you can see what they did. Annoys me to no end but it is what it is. For some reason the graphics card companies can't stay happy with a naming scheme for long.

This is a good thing for AMD.

[indymike]

The future market for GPUs is not a bunch of gaming enthusiasts and the design cycle for new devices takes sometimes years. The future is in tablets, phones, TVs, and other connected devices. Having a GPU that fits in a smartphone is more important than having insanely good rendering performance. Having a GPU that draws little power and does not need three fans and a big heat sink is crucial. AMD's marketing team made a colossal branding mistake by killing the ATI brand when millions and millions of gaming devices were sitting in people's living room with little red stickers on them and on the splash screen. ATI had become a household name thanks to the Will. ATI was a much more recognizable brand than AMD ever was. When you add together the amount of change in the GPU market (desktop down, small device up) My guess is senior management did not like the fundamental marketing strategy and wanted to change the direction. After marketers lock in to a branding strategy, it's very hard to get them to change direction, just like it's really hard to get a Java development team to port their product to Lisp, Add to that the sea change in who is buying GPUs and what devices they go in, and you have a situation where AMD needs to totally reposition. To everyone here who is saying marketing people are not as important as the engineers, I believe that you are talking about a chicken and egg problem. To succeed in the market, you have to have a product people want (good engineering) and the marketing make sure that the appropriate level of demand develops in the marketplace. Marketers do a lot more than hang around the coffee pot wearing black turtlenecks and discuss the merits of sumatra vs. colombian coffee.

sounds like a good move to me

[Anonymous Coward]

how many liars do you need on the payroll?

Re:Vital? yes Vital

[gr8_phk]

It can't be easy to determine where the cuts are going to be made if you've decided to not just do an even across-the-board cut. Most divisions within any sizable company could have good arguments made for not selecting them for cuts.

AMD is a Fabless chip company now. That means they design chips. They are behind in performance on the x86 side, are about to be behind in low power when Intel uses FinFETs (sorry Tri-Gate). The last thing they need to cut is their core design business - it's what the company does.

Small correction

[gr8_phk]

The Bulldozer FPU is shared between the two cores in a module. It sounded like it's really only a single FPU for the new 256 bit instructions. Remember when the K8 used a 64bit FPU and still kicked Intels butt? IMHO they waited about a year too long to upgrade it to 128bit. Do we really know where BD bottleneck is?

wasn't just marketing/PR...

[Anonymous Coward]

many (most?) of our engineering departments saw 5-10% layoffs too. odd that this is under-emphasized...