Slashdot Mirror
AMD: What Went Wrong?
Barence writes "In 2006, AMD could seemingly do no wrong. Its processors were the fastest in the PC market, annual revenue was up a record 91%, expansion into the graphics game had begun with the high-profile acquisition of ATI, and it was making exciting plans for a future where it looked like it could 'smash Intel's chip monopoly' for good. Now the company is fighting for its very survival. How did AMD end up surrendering such a advantageous position – and was it given an unfair shove on the way down? This article has plotted AMD's decline, including the botched processor launches, the anti-competitive attacks from Intel and years of boardroom unrest."
It's really simple--Intel made better products. Once Intel abandoned the dead end of the Pentium 4 and changed tacts with the first low-power Core chip, AMD never had a valid response. The article details some predatory behavior on the part of Intel which was eventually settled, but I don't think the outcome would be different today had that not occurred.
Of course, Intel better watch its back with ARM around.
Intel has had its share of buggy and bad designs, and that's even without going into discussion of the HMSS Itanic. Some AMD chips do great job of bang for the buck, my laptop has a nice dual core one that made the cost much less than comparable Intel chip would.
Still, AMD needs to get more risky with heavy investment into more advanced design and fab. mediocrity just isn't tolerated in processor design.
Intel just succeeded HUGE. They few years of AMD dominance were more a result of intels missteps. The p4 didnt clock as high as they wanted it to, so they had to scramble up a new design.
...Intel had basically sat back on its a**, AMD make some great design decisions and Intel said "Oh sh**, how'd we suddenly become second...?" From that moment on there was only going to be one winner in the PC chip space.
Loading...
AMD is not that much newer than Intel less than a year, and in comparison Apple is just a child
A firm with a Monopoly has multiple, permanent advantages. That there is little/no interest in breaking it up is another story.
http://www.maxineudall.com/2010/02/should-economists-be-sued-for-malpractice.html
Let's not forget the underhanded tactics that Intel used. They were forced to pay a minimal $1bn to AMD for it. I always thought its too small an amount for losing their position as leaders in the CPU market. And now look how things turned out...
AMD's budget range is still better than Intel, when compared at a constant price against Atom
But with the netbook/nettop market starting to flatline (or so I've heard), maybe they just made a wrong stratey decision
Also, the botched Bulldozer launch: they should have used the no. of complete modules in the processor name, instead of the number of Integer units
That way they wouldnt have a 6 core which was actually 3 core, but rather a 3 core which performed better in Hyperthreading than an equivalent Intel
Getting the driver issue sorted out before launch would have helped as well
AMD will never fail because Intel won't let it fail since it is their DOJ defence against being a monopoly. The couple of times AMD got ahead of Intel on technology, like x86-64, Intel started a money losing price war to put AMD back in its place. When AMD is struggling, Intel raises profit margins on its products to help them out. There are also less advertised ways Intel helps keep AMD afloat: Patent sharing, employee no-stealing, joint tools development like OpenAccess, etc. Having worked in that industry I was always surprised that the DOJ never came down on them for those agreements. The patent sharing and joint tools ones are official even though Intel puts like 10X more into them as AMD does. I left that industry after 5 years since I saw it as a dead end since you only have a few companies competing for your skills. As my manager at Intel told me, "I won't give you a raise since you only have one other place that would even care about the skills you picked up here, AMD and we really control them too."
Right from the start, you've lost me. If memory serves, 2006 would've been Athlon XP Barton core era. At this time they were numbering their CPUs in a way that indicated what P4 they could beat. But who was responsible for rating the CPUs for speed when they came off the line? AMD. So really they asked themselves, "Is this processor faster than a P4 1.6Ghz? Yes? Then this one is a 1600."
Yeah, you can stay that they were faster and be right, but only because the processors were marketed in a smart way.
... their merger with ATI didn't make sense to me because I knew that was going to divide the companies focus. You begin to lose focus the more stuff you try to take on and do yourself unless you have significant resources to buffer you against screw ups. Now AMD has great graphics cards but extremely poor cpu's that were extremely late and not even competitive with the previous generation of cpu's. I imagine this schizophrenia has hurt amd's focus.
AMD really doesn't know what kind of company it wants to be and it needs to find out because trying to do too much without the talent or resources ends in mediocrity.
AMD might have made okay CPU's but their partners made junk. You simply can't buy quality motherboards for AMD. All of it seems to be low-end crap with weird flaws. Every AMD system I have put together I wound up regretting. Things would crash randomly, freeze randomly, or just act downright strange. With Intel-based systems, I rarely have this problem (though I always pair it with a boring, plain-vanilla intel motherboard).
Bottom line, I simply cannot recommend AMD-based systems. Sure it costs less, but you pay for it in frustration.
AMD has done this many times throughout the years.
The only reason they boom is when Intel makes a mistake. In the mid 2000's Intel bed on that crappy Pentium 4 line. This allowed AMD to gain a huge foothold. It was only temporary until Intel figured out they goofed and corrected. AMD sat on their hands and didn't invent the next thing so Intel just stomped all over them.
This isn't the first time this has happened. The same thing happened in the days of 486 and 586's. AMD gained a huge share then lost it all as Intel corrected they're mistakes and AMD failed to continue to innovate.
It's almost like AMD shows the way then Intel does it better. It will probably happen again assuming AMD doesn't eventually just die.
And I think ATI was one of the worst. They were already loosing the competitive edge against intel before ATI but it seems that ever since they really have been putting all their efforts into ATI and boom now here we are in 012 where a slower, hotter AMD chip based on 5 year old designs cost more than the intel.
If you are asking what went wrong, then you are not part of the 1% that cleaned up. Didn't you cash out your stock options and use your golden parachute?
Site was slow for me (not quite slashdotted), but got through OK.
Am I the only one who's getting an error at PC Pro? Is it paywalled or something?
Thats what went wrong!
The article is pretty explicit about how AMD dug its own grave. I don't think blaming an Intel monopoly is all that convincing.
Really? The article mentions how Intel managed to get Sony money to cancel ALL AMD shipments, and how they paid Dell roughly 3/4 of a billion dollars in a single quarter to not use AMD chips. But I'm sure you're right, I'm sure keeping AMD out of all of the major OEMs(except to some extent HP) had nothing to do with it.
If i had one dollar for every brain you dont have, i would have $1.
That campaign really had a lot of success. The only people who buy AMD are geeks who only do it when it gives a good price performance ratio. It does for me, going AMD simply means you can spend your budget on a fast SSD which will do a hell of a lot more for your performance then a faster more expensive intel CPU with a regular HD.
But people like me are the exception and AMD never really managed to remove "a computer has an intel inside" from the consumers mind. Just try your local electronic store.
Netbooks were a chance, AMD didn't put restrictions on its netbooks but they failed to push high end netbooks before Intel again stole their thunder with smart books. My netbook has got 8gb in it, it makes it a very smooth machine, just light and cheap enough to lug around and not worry about it getting dented or worse, stolen. Netbooks partially failed because they sold with slow HD's and tiny amounts of memory, hurting their performance no end.
AMD just never had the clout to sell its chips on even terms. And it is sad because Intel dropped the ball completely when they believed they had no competition. There is a reason that 64 bit linux is report as AMD64. Intel failed and AMD delivered but for AMD to have truly broken through they need a long string of victories and no losses like Bulldozer.
If AMD wants to succeed, they might consider something that Intel is also thinking of doing. Intel is having trouble gettings its chips into tablets and phones especially, so they have considered making their own... AMD could do a lot better getting their CPU's in PC's if they started selling them. Control the whole supply line and pass the savings on to the consumer and beat Intel and Intel Inside PC makers on price. Intel can't do that for fear of pissing of all its customers but AMD doesn't have many bridges to burn.
Yes, making PC's is a very low margin industry but that is partly because you are buying all the parts from third parties. AMD wouldn't be doing that. The profit on the CPU inside the PC would be part of the profit of their PC. The profit on the graphics card would be part of the profits on the PC.
Risky and unconventional but unless THEY build the PC, they are always going to have a hard time getting their CPU into the PC.
MMO Quests are like orgasms:
You may solo them, I prefer them in a group.
What're those things? Big, loud boxes. There's usually lots of them in a big, cold room together. Oh yeah, servers!
I think those are probably quite important, too.
Syllable : It's an Operating System
Dell – then the world’s biggest PC maker – received billions of dollars to “remain monogamous” with Intel. At their peak in the first quarter of 2007, payments from Intel made up 76% of Dell’s quarterly operating income: $723 million against a total of $949 million.
And I really wonder why Intel hasn't been gutted and salted for monopoly abuse, with its CEO and main backers arrested. How can it not be MORE clear than that ?!?
Non-Linux Penguins ?
Weird. It's working fine now.
Only the geeks still worship pc type computers
Most of us moved on to smartphones and tablets.
Only dumb people worship smart phones.
There you are, staring at me again.
I've read at least twice that throughout its latest history (the last ten years) AMD managed to create only two new CPU architectures, K8 and Bulldozer.
All AMD CPUs between K8 and Bulldozer are more or less the same design, and that fact alone explains that even Phenom2 CPUs offer modest improvement in IPC and power consumption over original Athlon64 CPUs which were released over 8 (!) years ago. All these CPUs share the same functional blocks, the same cache hierarchy, the same number of core blocks, etc.
Meanwhile during this time Intel has gone through Merom, Conroe, Wolfdale, Kentsfield, Arrandale, Clarkdale, Lynnfield ... the list goes on and on. Every 2-3 years Intel offers some radical improvements which made Intel the king of the hill since the advent of the Core 2 architecture.
Also we have to bear in mind that Intel's R&D's budget equals AMD's entire revenue, and since the x86 architecture is one of the most complicated computing architectures (at least from what I've heard), maybe the fact that AMD is always trailing Intel CPUs is that AMD just lacks resources to innovate and invent (actually resources are there but senior managers in AMD have indiscreet bonuses and salaries which means they don't have as many talented engineers as e.g. Intel can easily afford).
Restructuring? Incidentally, the year mentioned in the OP is 2006, which was also the year that AMD spun off Spansion, which used to be AMD's positive margin company, but happened to be losing money competing against Intel. That was when AMD managed to hit a home run w/ AMD64, and also spun of Spansion. Intel followed suit much later, spinning off Numonyx, which later got acquired by Micron.
AMD is in the SSD business? That's news to me!
I've never had problems with AMD processors--it's always been the supporting hardware. Basically, every AMD supporting motherboard/chipset I've tried to use--be it MSI, or Asus, Gigabyte, or ABIT has just been a steaming pile. My last "attempt" to use AMD was building a media center box, and it more or less ended in failure since the board didn't play nice with the 5770 I was trying to use. RMA'd the card, same problem when it came back. RMA'd the board, same problem. Switched to Intel chipset board and CPU, no more problems. Every time I've tried to give AMD a chance--it's ended in similar results and I can only conclude that board MFGs just don't give AMD hardware the same level of QA as their Intel lines. They're ok with "good enough". On the Enterprise side, AMD ignored virtualization--they've never done well in the server market (or even tried very hard). Intel saw the trend and backed VMware and won big. AMD stuck it's head in the sand and decided it was OK to eek out an existence making emachines and QVC specials it's primary target audience.
The CEO of Intel always said he wanted to be the McDonalds of CPUs. That is, efficient produce mass quantity of CPUs. They spent far more on the manufacturing facilities then AMD. Whenever you here Intel people speak almost always focus on their Fab plants, not their chips.
So, you are going against the a company that's 10 times your size and can produce large run of chips cheaper then you. Intel could always discount it's chips more then AMD and still have more money then Intel.
The only way AMD could stay ahead of Intel was to always be faster, smarter, more daring. And for a while they were.
they paid Dell roughly 3/4 of a billion dollars in a single quarter to not use AMD chips
When they're buying back multiple billions of dollars in product, 3/4 of a billion dollars off is called a discount - not a monopoly. Businesses use that exact strategy all the time. "If you buy a huge order of our product instead of the competitor, we'll offer you $____ off!" They also have the option of sending as many 'promotional' free products as they want in order to convince the potential customer, even if it is half of the customer's order.
At the same time, Intel is not at fault because their products are more functional and desirable to the general computer user than the alternative - at least, no more so than Apple is at fault for having an enormous following in the "I don't know computers, I just want it to go" market. If AMD wanted to compete seriously in the consumer market, they could - but they aren't.
If the only way you can accept an assertion is by faith, then you are conceding that it can't be taken on its own merits
Right now they have to few pci-e lanes and adding TB, USB 3 to boards just makes it suck even more.
No, I didn't say AMD is in SSD business. I am saying they will have to restructure, do something, concentrate on something and find what their core competency is or should be.
You can't handle the truth.
Only dumb people worship.
PlusFive Slashdot reader for Android. Can post comments.
Their tendency to over-reach their supplier's manufacturing capabilities. While this excess of ambition has served them well with the push into 64-bit and multi-core computing, it can sometimes lead to them painting themselves into a corner. And it's happened several times in their history.
Their tendency to come of with what are, at the time, hugely over-complex chips.
Their inability, due to their lack of actual fab space to implement in a timely fashion.
The last leads to a disheartening tendency towards paper launches.
Their piss poor support of what they DO have out. Especially early on. They struggle to get stuff out the door, but usually all the amenities aren't in place at that point.
They essentially went from fighting a single-front war with Intel to a two-front war with Intel and nVidia in two of the most cut-throat markets in computer electronics. And they were doing neither of them well.
The fact that the acquisition of ATI ate up valuable capital that could have gone towards pushing their CPU architecture forward (and out the door) in a more timely fashion.
Their continued need to support ATI continuing to eat into valuable capital.
The fact that the market, while bigger than ever, is more cut-throat than ever due to the economy.
It's not real hard to come up with the reasons why AMD's in such trouble.
Chas - The one, the only.
THANK GOD!!!
The movie character Gordon Gekko, got famous for his words "Greed is good" in the movie Wall Street, but he was wrong, and AMD has proven it - as so many others have before them, greed is indeed NOT good, it's a destroyer of all things good.
Why?
Because AMD was Warner Brothers when Disney always bet their money on Cute & politically correct. AMD appealed to the young student generation, the people that wanted POWER but didn't buy into the heavily advertised Intel hype. Sure - nothing wrong with Intel, I was an avid Intel fan myself, the AMD processors where notorious for overheating, and several issues on certain math performances, but AMD overcame those issues, and produced some absolutely AMAZING processors that even outperformed their competitor at a staggering 3rd of the price back then, it was a no-brainer, every geek wanted an AMD in their computers, many of them where excited about overclocking their AMD cpu's to unseen speeds, it was indeed the "rogue" choice, but people (like me) loved it, and certainly took advantage of it.
But anyone who gets up there, get's taken by GREED, it's kind of like Nintendo who just couldn't understand why no one wouldn't pay the same price for their toy, when it was 3 times slower than the competitor, it's like Sony who simply didn't understand why no one wanted their proprietary formats and couldn't understand the need to have an open platform, when they could be in total control instead...
Yep, story of our lives as computergeeks & users, history repeats itself, and it never fails to tell things like it is.
What this world is coming to - is for you and me to decide.
Who has a monopoly of what? Intel dominates the market for desktop and laptop processors, but is way behind in the mobile markets, which are totally dominated by ARM licensees. Breaking up Intel, like most break up efforts, would come too late. Remember when the Government tried to break up IBM?. A few years later, IBM was almost irrelevant.
In the land of the blind, the one-eyed man is king.
In my opinion, AMD has always been the poor man's Intel. You'd really get what you paid for. If it's the performance you wanted, then go for it. However, often times a lesser Intel chip would match performance (if it didn't out perform in some applications) of a newer AMD chip for much less cash.
Chewbacon
The Bible is like Wikipedia: written by a bunch of people and verifiable by questionable sources.
Then when you look at the server market, they do not compete at all anymore. Performance per watt and per dollar both lag badly behind the Xeon.
That's untrue in my experience and has been for 5+ years. In the higher-end 2-4 CPU server range, AMD has had the best performance/price ratings for a long time because competitive Xeons are much too expensive. For example, the highest-performing 4P Linux servers (e.g. SPECint2006 rate) are currently Xeon E7-4870 based, followed closely by Opteron 6282 SE, but for the Xeons you'll pay 4 times as much as for the Opterons. A typical server configuration with 256GB memory will cost you ~8000 EUR ($10500) if you go with the Opterons and ~20000 EUR ($26000) if you go with the E7-4870 (and if you can actually find one on the market). More affordable Intel-based servers are not competitive performance-wise with the 6282 SE. If you don't need much parallelism and a lot of RAM, you might be able to get a more affordable offer using Xeons (with 2 of their 4C CPUs e.g.), but even there C32 based Opterons will offer much better performance per Dollar at comparable or lower TDP even (e.g. 2 x X7542 vs 2 x Opteron 4238). We've always been comparing closely when purchasing beefy 1U/2U servers over the past 10 years and Intel has not had competitive offers since their socket 604/Clarksboro Xeons when they allowed decent amounts of RAM (24 FB-DIMM sockets) in 1U compared with socket 940 Opterons (at somewhat sane prices). YMMV if your CPU needs are different, although I'd like to know how ...
"I love my job, but I hate talking to people like you" (Freddie Mercury)
Intel still has the advantage of QuickSync for super fast decoding/encoding/transcoding. Besides gaming, the most CPU intensive applications is probably centered around mobile media. E.g., transcoding that blu-ray you got from Redbox so you can watch it on your iPad. Sure the quality isn't great, but AMD has no answer for it in the near future. http://www.tomshardware.com/reviews/sandy-bridge-core-i7-2600k-core-i5-2500k,2833-5.html
AMD were extremely slow to push out low power chips.
I think the article is a bit premature. We've yet to see where Bulldozer can go and it's definitely a design aimed at a 6+ core future.
The big question has to be: why are AMD losing money? That's not something the article really answers.
To me, it seems like every market AMD is in, Intel (and soon ARM) are in -- with predatory pricing. Where AMD has chosen to not compete, Intel can charge monopoly prices.
The article makes an interesting point: Intel bet heavily on fabrication plants which are way ahead of the competition. AMD relied on others and has consequently had problems.
AMD had a wonderful technical position, Intel bet the farm on Itanium and NetBurst. AMD countered with an x86 architecture that was much much more efficient than NetBurst, a 64-bit implementation that didn't break backwards compatibility, and to further embarass Intel an affordable NUMA architecture with on-package memory controllers. For all this, 'Intel Inside' *still* carried some marketing weight despite the horrible tech behind it at the time. AMD failed in two ways:
-They failed in marketing execution to erode the value of 'Intel Inside'.
-When they did succeed, they didn't really come up with any *new* game changing plays. Intel's QPI was catch up to hyper transport, but since then Intel has continued with superior fab technology, advancing performance per clock, more memory channels per package, and incorporating features for particular sore spots like AES and h264 encode/decode. AMD's biggest advantage at the moment is that Intel GPUs are relatively poor and the Fusion line can quite thoroughly embarrass intel at gaming. The problem being the gaming market is very comfortable with discrete GPUs and this difference matters for a relatively small slice of the market.
XML is like violence. If it doesn't solve the problem, use more.
Oh, their core competancy is undoubtedly CPUs - they abandoned every other segment they were previously in - programmable logic, flash, and a few other things from 2005 that I've forgotten already. As for ATI, both they & NVIDIA were going from graphics chipset design to complete CPU chipset design. Previously, AMD used to build their own chipsets for their CPUs, and later, went w/ the ones from Via, SiS and ALI. Once NVIDIA and ATI entered this market, there were new complements and competitors, but it made sense for AMD to acquire one of them so as to fill a niche they weren't addressing - the GPU. Personally, I think they'd have done better marrying NVIDIA.
I agree w/ a post way above - that AMD missed the bus by betting on desktop CPUs continuing, instead of focussing their CPUs on laptops and power saving designs. I think they've gotten it right by now, but just wonder if they've missed the bus. They're probably trying to hedge their bets by stating that they're open to offering ARM CPUs as well, but if a tablet or computer vendor is shopping for ARMs, why would they go to AMD, as opposed to far more established players in this market, such as TI, Qualcomm, NVIDIA or Freescale?
I think they have to focus on solutions that beat Intel on the performance per Watt metric and price them on par. Oh, and given how well they've done w/ some of their recent low powered CPUs, work w/ some tablet vendors in getting them to offer Windows 8 tablets w/ Fusion CPUs - those will certainly be better than ARM in that they'll be more likely to run legacy Windows apps.
Intel absolutely sucks for integrated graphics.
People who care about 3D don't rely on CPUs with integrated graphics. Intel's IGP is designed for people who watch Youtube videos and play Facebook Flash games, not the latest Call of Duty.
Saying 'my new integrated GPU is now half as fast as the slowest discrete card!' is not a great marketing win. If you want to play games well you need something better and if you don't care about games you don't care how fast the integrated graphics are.
It's total bullshit. This guy has been following my posts for literally months. The list of people's he's accused me of being grows every time. I'm now apparently almost 20 people astroturfing Slashdot for every company that he doesn't like.
His claim about me karma whoring for mod points is also bullshit; I can't even get mod points because I posted in "the post" many years ago. I don't even know what the mod interface looks like because I've never seen it.
Normally, I wouldn't care, but my post was originally at +5, then this troll got modded up to +5 Interesting instead of -1 Offtopic like it deserves, and now my innocuous post is modded down for no reason. In fact, I think he's the one with multiple accounts abusing moderation points and subverting the comment system.
Robert Rozeboom at geek.net has told me that Slashdot's moderation system is one of many things they plan to revamp soon. It really can't come soon enough.
Posted anonymously because this is all totally offtopic.
- bonch
Intel seems to be winning because of marketing. Their top end CPUs out perform AMDs, but few people actually buy those.
Intel's MID-RANGE CPUs beat AMD's high-end, even though the AMD CPUs are 50% larger. That's a recipe for disaster, because AMD are forced to sell their most expensive CPUs for less than Intel's mid-range. Few people can see any good reason to buy a slower, more power-hungry AMD chip instead of an i5 unless the price is low enough to justify that.
I wouldn't bet against Bulldozer in the long term because the benchmarks I've seen seem to indicate some kind of unexpected bottleneck in their hyperthreading implementation; if that's the case then a new generation could actually make some use of all those extra transistors. But for now it's hard to see how they're going to make enough money from it to fund development of the next generation.
If AMD goes down, is that going to 'end' ATI too?
ATI would presumably be sold off; from what I've read ATI's GPU profits seem to have been keeping the CPU side of the company alive for the last few years.
"The average user does not need more cores right now"
cores are the only way to improve performance because the GHZ increases are reaching the limits of the physics of the universe as currently understood by science. There are not going to be 15 GHZ chips because it is simply not possible unless you use cyrogenics and i don't think the "average user" wants to have a jar of liquid nitrogen sitting in their house.
there are no regulators anymore. the agencies that were supposed to watch over this stuff were gutted and staffed with clueless yes men whose job was to make sure their employees didn't do their jobs.
a few books for you:
"The Asylum", Leah McGrath Goodman
"The Big Short", Michael Lewis
"The Sellout", Charles Gasparino
"All the Devils are Here", Bethany McLean, Joe Nocera
"Colossal Failure of Common Sense", Lawrence McDonald + Patrick Robinson
"Lost Trust", Lang Gibson
"Diary of a very bad year", Anonymous Hedge Fund Manager + Keith Gessen
the article describes this practice going on at many, many computer suppliers, not just dell.
of Intel chips, does not have 'intel inside' anywhere on its stores, logos, brands, cases, etc. they didn't even used to say 'intel core duo" processor, they just said 'core duo' processor.
are ok? wow. interesting moral theory
the result was modems, the BBS, and, oh, something called 'the internet'. it was what people used to browse the web on before "Wal-Mart Mobile" took over.
This article would have made a lot more sense several months ago, right after the poorly executed bulldozer launch. At that time, AMD was having very bad supplier problems with Global Foundaries producing their Llano APUs, and they had just launched a new architecture that was disappointing at best. Since then, they've had a number of things happen that have substantially improved their outlook.
First, the layoffs. While layoffs are rarely a sign of corporate strength, their layoffs seem to be concentrated in marketing and PR, two areas where, quite frankly, AMD should have been firing people a long time ago. Intel's "Intel Inside" and "Centrino" campaigns have firmly established them as the better chipmaker in the minds of most people. AMD has allowed themselves to become the "cheap" option rather than the "value" option. That's marketing and PR. Additionally, as disappointing as Bulldozer is, it would have been much better received had AMD not hyped it as an architecture that would bring Intel to its knees. Had they marketed it as a forward looking architecture that would be capable of scaling with future software, and had they not marketed modules as 2 cores, they would not have had the massive deluge of negative press ("8 core AMD barely competitive with 4 core Intel" is a horrible headline for AMD). If AMD is truly using the savings from these layoffs to devote more resources to product development, that will be a good thing down the road.
Second, their APUs. both Bobcat and Llano, do hold up quite well against the Intel competition. Bobcat is flat out better than Atom, and while Llano comes up short vs Sandy Bridge on CPU performance, with its superior graphical capabilities it provides an enticing option as a more balanced option for people that want a general use laptop. AMD has been constrained in this space by their supplier issues with Golbal Foundaries. Recent reports, however, indicate that 32nm yields are improving. If they can launch their Trinity APUs on schedule, they should be well positioned to take market share in the laptop segment, including the growing "ultrabook/ultrathin" segment.
Finally, Bulldozer. Bulldozer is a disappointment right now, but it doesn't appear to have a fundamental flaw that can't be fixed. Recent reports indicate that Trinity will have a substantial improvement in IPC vs Bulldozer. Add to this that windows 8 is expected to launch this year, (with a Bulldozer-aware scheduler - reportedly good for a meaningful boost in performance), and you have a much better positioned product. While it almost certainly won't catch Intel at the very top, Trinity (and Piledriver on desktop) should be able to compete throughout most of the budged and mainstream market segments. On the server side, it seems Bulldozer is actually selling relatively well. Its design is meant for the heavily threaded applications used in server workloads, and the compiler/system tuning necessary to get the best performance out of bulldozer is much more practical in the server space than it is for desktop users.
Overall, while AMD does have its risks going forward, it is in one of the stronger positions it has been in. They are profitable (and have been for 8 straight quarterly statements) They have competitive products in most segments of the market (with the major exception of high-end desktops and laptops). They do have technology that positively differentiates them from Intel in some key segments (the graphics capabilities of their APUs, the CPU performance of their Bobcat processers vs Atom). And, they have a modern architecture that they should be able to add to and improve upon for several years to come.
The Intel compiler, widely regarded as the best compiler available for x86, still produces code designed to make Intel chips look better than any others.
http://www.agner.org/optimize/blog/read.php?i=49
That page was posted three years ago. Scroll to the bottom, and read the latest additions to the discussion there: "New Intel compiler version - still the same!"
http://www.agner.org/optimize/blog/read.php?i=49#179
This makes it difficult to be sure how much better Intel chips really are than AMD chips. When the Intel chip scores higher on a benchmark, and the benchmark includes Photoshop, was the Intel chip actually better or was Photoshop compiled with the Intel compiler?
Sadly, I think Intel chips really are better now; given that Intel is leapfrogging past AMD on process technology, they have major advantages so their chips ought to be better.
But I still buy AMD. Yeah, I'm giving up some increment of performance... but the chips these days are so fast, I can survive on only 90% performance or whatever. And I prefer to avoid doing business with a company that continues to sell a compiler that sabotages performance on competitor's chips.
Personally, I would love to see AMD sell a line of processors that return "GenuineIntel" for the CPU ID, and thus run Intel compiler code at full speed. When Intel sues them, they can argue that this is necessary for full compatibility with the code produced by Intel's own C compiler. (Yeah, I know. It will never happen. It's a fun daydream but that's all.)
Even if AMD doesn't have the top performing chips, they continue to score very well on price/performance, and the performance is good enough for me. And they are less evil than Intel. So I remain an AMD customer.
steveha
lf(1): it's like ls(1) but sorts filenames by extension, tersely
HP convinced Intel that it was not going to be possible to scale performance of the x86 architecture any more, and that the only options was a new architecture that exposed massive processor resources directly to the compiler which would use its broad scope of compile-time code visibility to explicitly schedule the code optimally for execution. The result was the Itanium architecture,
Unfortunately (for them) this turned out not to be a win, because even the best compilers can't predict what's really going to happen at runtime. They also took WAY too long to get Itanium working, and also what they were doing was pushing a hardware problem back up into software with an apparent belief that a silver bullet would be available to slay the problem of a ridiculously complicated piece of software that they needed to develop (the compiler).
If not for the delays and one other little problem they probably would have succeeded in replacing x86 with Itanium, whether or not it was a technically viable idea.
That one other little problem was AMD, who decided that yes you could push the x86 architecture quite a bit further along than where it was when Intel effectively abandoned it during the Itanium development. AMD pulled some rabbits out of hats and basically took over x86 development from Intel for a while. Eventually Intel was forced to get back into the x86 world in a serious way, when it became clear that AMD was going to pretty much take over the market with 100% compatible processors (and with 64-bit capability) where Intel was going to be stuck pushing an overweight, incompatible, late, and under-performing alternative.
Unfortunately for AMD, once Intel finally got the boat turned around in the right direction, they had the money and the engineers and architects to do x68 even better, and they have proceeded to produce a series of incredibly impressive implementations which squeeze the most out of both process improvements as well as architectural advances.
I think that without the Itanium detour there would never have been an opportunity for AMD to do what they did, but without AMD we would probably all be struggling with the baroque complexities of Itanium-powered PCs (which is too bad because then I would have been able to make lots of money hand-coding Itanium machine instructions for people :)
G.
whoosh.
>>> The only people who buy AMD are geeks who are aware of computers beyond Dell-made ones. >>>
Get informed: Dell's entry level servers are AMD.
The article looks back at the year 2006. 2006 was also the year when Apple switched to x86 computers. At the time, there was obviously the choice between AMD processors and Intel processors. And many people at the time said that Apple should have gone with AMD; we now know that would have been a big mistake.
In 2006, AMD processors were better than Pentium 4 processors. Pentium 4 design goal was "highest possible clock rate" with complete disregard of actual performance, because customers bought GHz. AMD countered by using processor numbers. Instead of a 3800 MHz Intel chip you could buy an "AMD 3800" chip which many customers thought was the same as the Intel chip, but in reality had much lower clock speed and slightly higher performance.
At the same time, the old Pentium 3 had much better performance per Megahertz than Pentium 4. Pentium M was a slight improvement of that, and Core Duo was again an improvement. Apple built a few Pentium 4 3.6 GHz Macs for developers. My first MacBook with 1.83 GHz Core Duo (May 2006) ran faster.
I think with the introduction of Core Duo, and with Intel getting rid of the abomination that was Pentium 4, AMD was beaten. They just didn't know yet for a long time. Reading the description of Pentium 4 internals, all I could think was "WTF". Same with Itanium (which was WTF squared). Athlon was in comparison a clean design, which was why it performed so much better per Megahertz. So was Core Duo, and since then Intel managed to stay with clean design and improve performance bit by bit.
Intel has enough of it that it can, without much trouble, rebound from a stumble. When AMD stumbles, their revenue stream suffers, which impacts their ability to produce the next generation of products, which causes their revenue stream to suffer...
We've yet to see where Bulldozer can go and it's definitely a design aimed at a 6+ core future.
Throwing cores at the problem isn't really a solution for the desktop. Most desktop apps are still single threaded and even games are usually unable to use more than four cores.
The big question has to be: why are AMD losing money?
Becuase intel are both bigger and technically ahead (both better designs and better processes afaict). This means a few things.
1: Intel can almost certainly produce equivalent/better chips to anything AMD can make at a lower cost.
2: Intel can produce chips that are faster than anything AMD can make. These chips can be sold with no competition (at prices that go up by big chunks for each minor step-up in performance).
3: Intel can spread their R&D costs over more units.
AMD got ahead of intel briefly because intel went up a dead-end with the pentium 4 but intel fixed that with the C2D and afaict AMD CPUs havebeen behind intel ones ever since. Afaict AMD has an advantage in integrated graphics but Intel is working hard to try and destroy that too and any serious gamer will probablly go discrete anyway.
Where AMD has chosen to not compete
I'm pretty sure that if AMD COULD compete in the high end desktop market they would. The very existence of the FX brand implies that they want to.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
I do. For geophysics, video, image processing, engineering design (finite element analysis) and a pile of other things that are very easily done in parallel it's a definite win. Those are very large niches.
A large but trivial example is applying the same filter to twenty million seismic traces (ie. that many tracks of digitised sound). Something like that is no harder to do in parallel than in serial but it finishes several days earlier if you can send it out to all the cores you can get to.
They didn't do that because they thought fabs were boring, they did it because they're expensive and they didn't have the revenue to support them. In part because of Intel's anti-competitive behaviour.
SuperMicro are junk?
AMD has never been saying that. When Llano came out, it was on par with mid-range discrete cards. You could clearly play quite a few modern games on it: http://www.anandtech.com/show/4444/amd-llano-notebook-review-a-series-fusion-apu-a8-3500m/11
Never underestimate the bandwidth of a 747 filled with CD-ROMs.
Real men have fabs. - Jerry Sanders.
AMD was not behind for a short time with Shanghai but noone noticed.
And with those mod points, the people operating the bonch account spend them on modbombing corporate-unfriendly opinions and upmodding other shill accounts controlled by their public relations firm,
What has happened to Slashdot? When such nonsense conspiracy theories get +5 moderation and they don't even seem tangentially related to the topic at hand.
Why posted as Anon? Don't want people to recognize some bizarre pattern of rants?
This is the second time today I have seen trolling/conspiracy theories modded +5.
We may as well be Engadget with the quality of discussion/modding recently.
I was actually interested to see if some had insights into AMDs stumbles. Instead I get nonsense conspiracy crap.
3/4ths of a billion dollar was their operating profit. I'm not sure how you quantify that as a discount. That would mean essentially Dell was losing money on every chip sold prior to that quarter. So lets not beat around the bush, Intel had the money and willpower to forcibly squeeze AMD out of the truly lucrative markets.
For the record though: Intel CPUs are faster but in every day business use AMD chips were probably a much better buy for office machines. In fact if you look at the cheap CPUs for the last 5 years AMD prices (at market) are substantially lower for business-level machines so really the idea that Intel CPUs are faster is true but false in the sense of their quality vs. cost analysis. In other words, Dell, IBM, and HP had no reason to keep buying Intel CPUs for their business lines except for the fact that Intel was cutting them deals where Intel made pennies on the dollar or lost money but made up for it in volume sales of more expensive chips in the other consumer lines.
A GPU only wins if the software is written to run on a GPU. Of the commercial products the company I work for uses there is not a single one that can do so, and only one of those vendors has a product that will (but it's for situations that don't apply in my part of the world).
The leading product in seismic processing still has portions that only work in 8 bit colour FFS so it's going to be a very long time before it will run on a GPU. There's nothing that does the same things from any other vendor that currently uses GPUs either.
Intel has made great success out of the different designs originating in their Israeli operations. The Core chips and others, for example. These designs gave Intel a tremendous product for years and years and made them a lot of money and made a matter of pride for Israel. This is not noticed so much in the US where nobody thinks much about where the design came from. But it matters back where the product was born.
And it was not unnoticed by Israel's enemies the Arabs. Nor was it missed by AMD. When AMD needed funding, it went to those Arabs and played investing in AMD as both a way to make money and also a way to smack at Israel via Intel. The Arabs bit. AMD got cash and turned out some decent products and there were collaborative efforts too like the Arab investment in Ferrari which went as far as Arab firms and AMD sponsoring Ferrari's Formula-1 team. The stickers on that car tell the story.
But the results were not enough, the fabs were woeful, and Intel came out with even more better stuff like the Core i3/5/7 series and AMD's Arab backers saw that they weren't going to win with this horse and refused to keep pumping in money with out something to show for it. They wanted a major win and got a lot less.
The Arabs have a lot of cash but they tend to be shrewd about it and demand results. AMD failed to deliver.
That's how AMD lost. The money to fight a war has gone away. All they can do now is fight small skirmishes.
It's a pity. As and AMD fan, and a fan of underdogs, I wish they'd continue to stick it to the man. But at the same time, even as I think highly of my quadcore PhenomII x4 965 AMD desktop, I see it get whooped in the ratings by similar Intel products. The only place AMD is winning is in being cheaper.
Sig for hire.
When I pointed out that some portions will only run in 8 bit that was nothing about whether a GPU rendering anything but instead a comment about development proceeding at a glacial pace (and most devlopers with a clue about that software not being available to update it over the past decade or more).
So yes, something that runs on a GPU would be nice and this has turned into another argument for open software instead of closed near-abandonware.
Intel had backed itself into a corner with the asinine idea that you could run fast with the P4's 256k cache, and the same on their multiprocessor Xeons. The fast but high latency memory bus would somehow take care of it. And they pushed hyperthreading at the same time, which puts even more pressure on the cache
AMD's caches were still decent size. No hyperthreading. And the Athlon's had a memory architecture which made each CPU responsible for part of main memory instead of choke-pointing everything through a single path.
Net result: AMD's CPUs spent a far higher percentage of their time doing work instead of waiting on a cache fill.
Then Intel figured it out, ramped up to 12 meg caches, copied AMD's memory per CPU approach and, for a while, ditched hyperthreading. And achieved nice power consumption / heat generation improvements to boot.
AMD's caches weren't that big. And still aren't.
Moderating "-1, Disagree" is simple censorship. Have the guts to post your opinion.
First is as a Linux users I always went with AMD, most of the machines with the AMD moniker had an AMD processor and kickin nVidia graphics on-board. So most AMDs at the time just fell into the "Just Works" category.
Now since AMD's aqisition of ATI - it's AMD with Radeon, so the "Just Works" appeal is gone. ATI support in Linux is still commonly flaky (yeah you can get it to work, but "just works" isn't the catchphrase associated with ATI.
Second is the new branding AMD Fusion and AMD Vision - which is just as incomprehensible as intels i# labeling in my book - can't easily remember which one was betther than the other, is it Fusion or vision, and how do those stack up to a Phenom II CPU??
AMD lost their customers because they made their customers "lost". Just this month my Linux box died... got too confused with the AMD branding so finally went with an Intel i5, made sure the MB supported PCIe and popped in my nVidia card, and am not looking back... Maybe by the next purchase they will have their ducks in a row so I can do some informed shopping... Until then, well... I just don't really know what is what with AMDs.
"Enjoy what you're doing! If it becomes drudgery, you're doing it wrong!" - Jim Butterfield
AMD + ATI seemed like the merger of underdogs to me.
I think if NVidia had instead merged with AMD the result might have been stronger, also thinking that it would have NVidia CEO in the drivers seat as he seems more driven and ruthless.
ATI would likely have been too weak to continue solo after that and might have been swallowed by Intel.
Come on, this isn't that hard to find: http://www.pcpro.co.uk/features/372859/amd-what-went-wrong/print
You know the thing about UDP jokes? I don't care if you get it or not.
That is patently false.
AMD's forward EPS growth is strongly positive in the short, intermediate, and long term, and AMD stock is up 33% in the last couple of months. Analysts are extremely bullish on AMD.
They have also exceeded earnings estimates for the last 9 consecutive quarters.
...botched processor launches...
What, you really need more than that?
"People who do stupid things with hazardous materials often die." -- Jim Davidson on alt.folklore.urban
This is hilarious. I've been on this site for more than 10 years (joined up during first undergrad degree in 1999 or 2000) and you've listed me as one of the shill accounts?
I suspected this was total bullshit before, but that simply confirms it as far as I'm concerned.
What? You think Apple/MS/Sony etc have been paying me for a decade? I wish! Then my student loans would be more manageable.
For the record, this is my only account, no one else has access to it, I've had it since first registering it, no one has ever paid me (or otherwise instructed me) to post anything. Just so it's down in writing.
Join the club - he's accusing you of being me, so I know it's all nonsense. This is getting out of hand.
You are the most paranoid and mental person I have ever seen.
Do you yell at your own shadow for following you around?
Seriously, you'll have every slashdot account owned by one guy by the end of the week if you think anyone who disagrees with you is a sockpuppet.
Let me guess, you're also one of the elite group that owns wikipedia? Same sort of crazy paranoia and conspiracy nonsense.
They didn't have the funds to both continue investing in development and run their own fabs.
Would you really argue that the right decision would be to keep the fabs and have nothing worthwhile for them to produce?
I don't know, but I'm still loving their 12-core Opterons.
"Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
I'm still surprised how people still pay much attention to benchmarks on Anandtech or Toms Hardware. All of these benchmarks are compiled with Intel compiler, and ran on Windows. They are relevant if you are transcoding videos or playing latest games. And if you are doing software development on Linux workstation with an SQL DB, none of this is relevant in any way. Do some tests for your own workload, then make decisions. Check out Phoronix- they at least run canned synthetic benchmarks on Linux.
And anyway, CPUs no longer matter- they are all plenty fast unless you are doing some heavy computation. I'd rather spend some money on more RAM and a fast SSD- this will improve system performance by much much more. Eclipse (Java IDE) responsiveness improved by ~4x when I installed a SSD in my machine. And currently I have a nice Llano laptop- plenty of CPU power in those 4 cores, enough IO with a SSD, and it didn't break my bank. It has a Radeon 6550 GPU, so it will run most games as well.
--Coder
> Most desktop apps are still single threaded
Maybe tru, but I could easily be compiling, listening to music, recoding a video and browsing the web at the same time.
...and I'm surprised it isn't used more often. I've used it to upgrade my PCs and I'm fairly certain an equivalent Intel solution would've cost me about £100 ($150) more. The Llanos seem to deliver upper i5 levels of performance with reasonably capable graphics before adding an independant graphics card, so maybe someone can explain to me why they aren't shifting more of them??
Donte Alistair Anderson Roberts - hi son!
Karma: Chameleon
A quote from an engineer who used to work for AMD
,which meant we had to stop hand-crafting our CPU designs and switch to an SoC design style. This results in giving up a lot of performance, chip area, and efficiency. The reason DEC Alphas were always much faster than anything else is they designed each transistor by hand. Intel and AMD had always done so at least for the critical parts of the chip. That changed before I left - they started to rely on synthesis tools, automatic place and route tools, etc. I had been in charge of our design flow in the years before I left, and I had tested these tools by asking the companies who sold them to design blocks (adders, multipliers, etc.) using their tools. I let them take as long as they wanted. They always came back to me with designs that were 20% bigger, and 20% slower than our hand-crafted designs, and which suffered from electromigration and other problems.
What did happen is that management decided there SHOULD BE such cross-engineering
That is now how AMD designs chips. I'm sure it will turn out well for them [/sarcasm]
source: http://forums.macrumors.com/showpost.php?p=9746191&postcount=619
more:
http://www.insideris.com/amd-spreads-propaganda-ex-employee-speaks-out/
It's difficult to be taken seriously for any sort of pro-anything stance when you come off like a creationist.
A Pirate and a Puritan look the same on a balance sheet.
The Capital District (Albany and surroundings) of New York State has been trying to rebrand itself as "Tech Valley." There are some good educational and research facilities in the area, led by RPI. But the centerpiece of the transformation was supposed to be a new, ultramodern chip fab plant. AMD's manufacturing subsidiary, GlobalFOUNDRIES, got huge tax breaks as an incentive to build there.
Construction took years longer than first estimates, and slowed to a halt a few times. Rumor says that lack of money was the cause. The plant is supposedly nearly complete, but they still haven't ramped up to anywhere near full production capacity.
Linux and BSD developers generally recomend Intel when it comes to video.
The matter is simple: nvidia only provides binary blobs, ati provides *some documentation* and crappy drivers, and intel provides full documentation.
The linux/bsd crowd may not be large, but they lost most of it. With intel manufacturing more powerful videocards, *nix users where no longer interested in amd/ati.
And of course; if you have intel video, you'll have an intel processor (I've yet to see a motherboard with intel video and amd processor).
I realize this is a very small portion of the market, but they lost is *completely*.
Aditionally, the netbook market exploded, and then the notebook market grew enormously. AMD did have powerful stuff, but they where never really and good at power-saving, or cooling this very well.
Intel is a chronically abusive corporation ; abusive towards the market, abusive towards its employees and abusive towards its customers. We NEVER buy Intel processors no matter what the deal or how hard we have to work to get an AMD system
Anyone who doesn't know about Intel's anti-competitive practices is asleep at the wheel. Their "forced ranking" where every tenth employee is fired every year - in Roman times this was called "decimation" and was considered barbaric back then, creates a work atmosphere of pure political backstabbing where the only real goal is to ensure sure you're not the one who is selected to die. In that environment, work product quality counts for one chip and sucking up and undermining others counts for ten.
Anyone who cares at all about either the free market or a fair workplace has exactly zero reason to buy a chip from Intel.
They wouldn't be allowed. They very nearly lost the farm trying to spin off globalfoundaries because they have a licence from intel for the x86 instruction set. They don't even make their own CPUs any more, GlobalFoundaries does.
Besides, what evidence is there that the 'open' community of CPU developers would design better CPU's than they already make? There aren't exactly a lot of people out there making CPU's on the side for the fun of it, (whereas there are with software). Designing a modern CPU is a really serious engineering challenge, and anyone with the knowledge and experience to make a CPU probably works for AMD or a competitor already.
That is still only four things and two of them should have pretty negligable CPU load. Yeah going from a single to a dual would probablly benefit your scenario a lot and going from a dual to a quad might help a bit but a fast dual will probablly still be better than a slow quad and a fast quad will almost certainly be better than a slow 6-8 core chip.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
Point taken, although pure CPU load would not be the only measure: good reaction time is important for music, and useful for web browsing.
3 cores could support 1 for video, 1 for compiling, 1 for web/music. Except GCC already supports parallel building: I am not sure what the maximum useful number of cores would be !
I would also agree that "I have N cores" may become the new "my camera has N-megapixels" !