AMD: what went wrong?
Posted on 17 Feb 2012 at 15:56
Five years ago, AMD looked set to topple Intel. Now its very existence is under threat. Mike Jennings investigates what went wrong
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, we speculated, it could “smash Intel’s chip monopoly” for good.
Its chief rival, meanwhile, was having a torrid time. Intel’s first-quarter results saw a 40% drop in profit, revenue was falling, and it shed around 10% of its workforce: a whopping 10,000 jobs.
The underdog from Sunnyvale was on the verge of toppling Intel, and it felt like the computing tide was about to turn.
The underdog from Sunnyvale was on the verge of toppling Intel, and it felt like the computing tide was about to turn
Although 2011 has seen eerie parallels with 2006, the shoe is now definitely on the other foot. This time it’s AMD that’s cutting 10% of its global workforce, after its share price plummeted from a high of $9.44 in February 2011 to $4.53 by October.
How did AMD end up surrendering such a advantageous position – and was it given an unfair shove on the way down?
The extent to which AMD has contributed to its own downfall can hardly be exaggerated. Back in 2006, AMD sought to capitalise on its success by announcing two major products: the Barcelona desktop architecture and Fusion, its much anticipated first collaboration with ATI.
Both were plagued with glitches, delays and production problems however, which extinguished much of the early enthusiasm.
Barcelona was officially finalised in the summer of 2006, with president Dirk Meyer promising a summer 2007 launch for the new chips, branded “Phenom”.
But the August launch was postponed until September and, when AMD finally put its chips on the table, they received a frosty critical reception – our tests concluded we were “underwhelmed by its performance levels” – thanks to a flaw that required a key internal buffer to be disabled. Fixed chips didn’t arrive until April 2008.
AMD’s failure to capitalise on its technology lead was brutally (and as we’ll see later, unfairly) exploited by Intel. Just months after fully working Phenoms finally materialised, Intel unveiled its Core i7-920 – and we said “there isn’t a doubt in our minds that tomorrow belongs to Core i7”.
Its November 2008 launch was followed by the mid-range Core i5 almost a year later and, in January 2010, by Core i3 – two series of processors that waded straight into AMD’s traditional budget stronghold.
ReviewAMD FX "Bulldozer"
Amazingly, it took AMD three years to respond to Intel’s first-generation Core processors, with Bulldozer. The architecture, launched as FX in 2011, is built using dual-core modules, and pre-release talk was bullish: leaked AMD documents saw high-end FX cores squaring up to Intel’s Core i7-2600 series, and AMD unveiled prices that were much lower than those of Intel.
The move was welcomed by IDC analyst Mario Morales, who says AMD should “establish its own position rather than compete with Intel”. AMD’s own spokesman Mike Silverman told customers to “let go of the old AMD versus Intel mindset, because it won’t be about that any more.”
Yet comparison is inevitable – and not very complimentary. Our review concluded that “Intel still holds all the cards”, with pricier AMD FX processors delivering benchmark scores synonymous with Intel’s mid-range Core i5s. The verdict was unanimous; our sister title bit-tech dubbed the FX-8150 a “stinker”.
Light was shed on Bulldozer’s problems when ex-AMD engineer Cliff Maier spoke out about manufacturing issues during the earliest stages of design. “Management decided there should be cross-engineering [between AMD and ATI], which meant we had to stop hand-crafting CPU designs,” he said.
Production switched to faster automated methods, but Maier says the change meant AMD’s chips lost “performance and efficiency” as crucial parts were designed by machines, rather than experienced engineers.
AMD’s latest chips haven’t stoked the fires of consumers, either. Martin Sawyer, technical director at Chillblast, reports that “demand for AMD has been quite slow”, and there’s no rush to buy Bulldozer.
Sales are partly propped up by die-hards who only buy AMD because they don’t like Intel
“With no AMD solutions competitive with an Intel Core i5-2500K”, he says, “AMD is a tough sell in the mid- and high-end market.” Another British PC supplier told us off-the-record that sales are partly propped up by die-hards who only buy AMD “because they don’t like Intel”.
AMD is more positive. Dina McKinney, corporate VP for design engineering, told us that “we’re happy where [FX] landed”, after the “herculean effort” to design and launch the chips. “We’re on a very rapid pace of improvements”, said McKinney, “and things only get faster and more refined”.
Fusion was another project that gave enthusiasts plenty of reasons to be excited back in 2006, but it took a staggering five years before it was launched in the form of the Llano and Brazos APUs.
Still, reviews were positive: we called the on-chip graphics “undeniably impressive” and dubbed Llano “a great all-in-one processor for the casual user”. Tom’s Hardware concluded it’s “a cool piece of tech for your mainstream mom or dad”.
But delays meant it had competition, with Intel’s integrated Sandy Bridge graphics capable of media playback and cheap discrete cards from AMD and Nvidia handling basic gaming and image work.
There’s little hope of drastic performance boosts any time soon, either. AMD recently scrapped its plans to shrink the low-power Brazos range from a 40nm die to 28nm, and switched manufacturing from GlobalFoundries, the fabrication plant it helped create, to rival Taiwan Semiconductor Manufacturing Company (TSMC).
AMD issued a statement confirming its 32nm “Trinity” APUs will be manufactured by GlobalFoundries but the statement offered no information about 28nm and beyond. With Intel set to release its 22nm Sandy Bridge revision, Ivy Bridge, in early 2012, the delay could be catastrophic.
Demand but no supply
Even at the height of its powers, AMD had problems getting enough chips onto the market to satisfy demand. In October 2006, we reported that AMD was “starving” supplies of its high-end parts to smaller OEMs to meet commitments to Dell, with only a “handful” of processors available to less fortunate, or influential, system builders.
Even at the height of its powers, AMD had problems getting enough chips onto the market to satisfy demand
Three years later, the launch of ATI Radeon HD 5800 Series graphics cards was undermined by TSMC struggling to produce the 40nm cores, with AMD spokesman Matt Davis admitting “the design is sound, [but] it’s a matter of trying to get to where it [TSMC] can yield”.
GlobalFoundries has also faced production problems. CEO Rory Read said in October that “we were disappointed with yields in the 32nm space”, and AMD’s performance in the third quarter of 2011 was affected, with all of its headline processors reliant on this process. McKinney echoed these concerns but remained upbeat.
“We’re very early in 32nm [and] trying to ramp as soon as possible. GlobalFoundries is doing a great job, and [it’s] a critical partner.”
Chillblast has confirmed it currently has “zero availability” of Fusion chips, with no confirmed date for new stock. In fact, since Llano’s June launch, Chillblast has been able to secure precisely two of these chips, and Sawyer said one supplier he uses had managed to source only ten of these sought-after processors.
A spokesman for another major British firm told us off-the-record it’s had issues with both Fusion and Bulldozer, saying “we’d struggle to make a high-end gaming system” because of a lack of availability.
Chillblast has confirmed it currently has zero availability of Fusion chips
CyberPower’s production manager, David Scott, tells a similar tale. “Both the A8-3850 and FX-8150 are in low supply, and these are what enthusiast customers want. They’re not selling in huge numbers, but it would be nice to have the chips more frequently,” he says.
Bullied by Intel
While there’s no doubt AMD has brought problems on itself, it’s also suffered at the hands of its larger and ferociously aggressive rival.
Intel didn’t claw back its sizeable share of the market by just improving its processors; it also threw money at the problem – by offering incentives to a host of PC makers to “postpone or cancel” AMD-powered products.
AMD’s court filings from June 2005 labelled Intel as “predatory”, and the firm issued subpoenas to Microsoft, Dell, HP and IBM to uncover crucial evidence. The findings were shocking.
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.
In a bid to win new custom, AMD offered HP a million free chips – but it was only able to accept 160,000 because of its agreement with Intel.
Sony dropped its AMD purchases “from 30% to zero within a matter of months”, while Lenovo postponed the launch of an AMD desktop PC and limited its promotion. “In its [Intel’s] perfect world, we wouldn’t exist,” said AMD general counsel Harry A Wolin in an interview with CNET in 2009. “I think they would absolutely like us dead.”
Intel was exposed when a Japanese court described its behaviour as “illegal monopoly abuse”, and after investigations in the UK, Germany and Italy, the European Commission imposed a record $1billion fine in May 2009.
Competition commissioner Neelie Kroes said “Intel has harmed millions of customers by deliberately acting to keep competitors out of the market for many years”, claiming that “everyone but Intel was worse off”.
Intel responded with audacious claims that the European Commission was trying to harm the company, and that AMD was using distraction tactics to cover its “business failures”, but it finally admitted culpability before another damaging court case in the US.
In November 2009, Intel settled all of its outstanding legal disputes with AMD by paying it $1.25 billion, with then-CEO Dirk Meyer saying the two companies were now “competing on a level playing field”.
AMD’s court filings from June 2005 labelled Intel as predatory
The payout meant the fourth quarter of 2009 was AMD’s first profitable period since 2006. It soon dipped back into the red, though, with its second quarter record revenue of $1.65 billion in 2010 still boiling down to a net loss of $43 million.
It’s impossible to say how much damage Intel’s illegal tactics did to AMD’s market share and profitability during the second half of the last decade, but it seems almost certain that – all things being equal – AMD would otherwise have done far better.
Lack of leadership
AMD’s battles with Intel have overshadowed a more insidious problem: a chronic lack of leadership. Eight senior executives have departed in the past four years: Henri Richard, the chief sales officer, left shortly before Barcelona’s debut, and Phil Hester, AMD’s chief technology officer, left soon after.
The former head of ATI, Dave Orton, left in July 2007 and 2008 saw Stephen DiFranco, the corporate vice president of consumer sales and marketing, fall on his sword. CEO Hector Ruiz left in 2009 after an insider trading scandal.
Ruiz’s successor, Dirk Meyer, left in January 2011, and he was followed out by Robert Rivet and Marty Seyer – AMD’s chief of operations and senior vice president of corporate strategy. September saw respected products group general manager Rick Bergman depart.
Another key signing could be CTO Mark Papermaster – responsible for the physical development of the iPhone and iPod
AMD has at least brought in well-respected faces to fill those desks. New CEO Rory Read is described by McKinney as “refreshing and energetic”, and chief information officer Mike Wolfe has previously worked at Motorola.
Another key signing could be CTO Mark Papermaster – the man responsible for the physical development of the iPhone and iPod. Previously at IBM, he was so highly regarded that his old employer actually sued him for breach of contract when he departed for Cupertino in 2009. An eventual agreement allowed him to head to Apple, but he was subjected to two evaluations to make sure he hadn't disclosed IBM's trade secrets.
A gloomy future
Below-par products, legal issues, boardroom politics and supply problems: AMD is in a bad way. Morales thinks it needs to “try to find itself for a couple of years” by “becoming architecture-agnostic, working with ARM and diversifying away from PCs”.
Most think AMD must expand into mobile chips, and McKinney agrees, saying that AMD is a “great design house” and that “the chips we’re working on today will go nicely into [the mobile] space”.
It’s going to take time and resources, though, and they’re both in short supply in Sunnyvale: AMD’s traditional competitors have already made strides into mobile markets, and a move into this area brings AMD into competition with experienced rivals.
Take Nvidia, which is gearing up to release its third-generation Tegra chipset. Morales describes older versions as “more PR than reality”, but Tegra 3’s presence in upcoming devices from Asus, Acer, Lenovo and Samsung certainly indicates its potential.
Perhaps the biggest threat to AMD comes from British chip designer ARM. It already has the lion’s share of the smartphone and tablet processor market; now it’s expanding into PC territory too.
Not only is Microsoft adding support for ARM into Windows 8, but the ARM-based Tegra 3 is targeting laptops. ARM’s executive vice president of marketing, Lance Howarth, has said that by 2020 “there’s only going to be two [chip designers] – ourselves and Intel”.
Morales echoes this sentiment, saying he expects AMD to become “a shell of a company” in the near future, as “tablets and smartphones move extremely fast and hardware companies can’t keep up”.
Perhaps the biggest threat to AMD comes from British chip designer ARM
The need for change is reinforced by AMD’s performance in its core market. Chillblast’s Sawyer explained that “back in the Athlon 64’s heyday, we sold 100 chips for every Pentium – but now we sell 100 Intel chips for every AMD processor”.
AMD has even had to deny it’s leaving the x86 processor market entirely after spokesman Mike Silverman said AMD is at “an inflection point” and CEO Rory Read spoke about AMD’s plans to “refine our focus on lower power, emerging markets”.
McKinney reassured us her design team is “working on the fourth generation” of Bulldozer, and implied that the core has plenty of shelf life – its predecessor, she reasoned, “was around for ten years”.
But if leaving the PC processor market is a step too far, drastic action is nevertheless required. AMD is 42 years old; it will need something extraordinary to avoid early retirement.
Author: Mike Jennings
Wrong info on Mark Papermaster
Mark Papermaster was sued by IBM when he left them to work with Apple. See http://en.wikipedia.org/wiki/IBM_v._Papermaster
By huxley on 19 Feb 2012
It appears Llano availability has been improving.
I'm not sure if this is true in the UK, but Llano seems to be available on North American retailer Newegg:
By Andruin on 19 Feb 2012
Buying it through retail or etailers hasn't been a problem in the UK. If I understand correctly, the issue seems to be that AMD weren't able to produce sufficient working units for OEM Manufacturers to buy and install in systems so lost out dramatically that way to Intel with the introduction of i3 and i5 processors at the same price.
By mr_chips on 19 Feb 2012
Huxley is right
And the article should be corrected. I'm normally not so anal, but i think in this age of apple vs. everyone else, it's important to be factually correct.
By Ranread on 20 Feb 2012
The Intel problems highlighted were from the late 90s and the first half of the previous decade, when AMD were at their height.
It no doubt damaged them, but it didn't really contribute to their lack of competitiveness in physical products or lack of yield.
Also, you seem to gloss over why AMD became so prominent.
During the early part of the last decade, Intel was making power hungry and inefficient processors, whilst AMD came up with a more efficient design (they were faster per clock cycle than an equivalent Pentium, so they ran slower and cooler and used less power).
It took Intel nearly half a decade to respond to that threat. By dumping the Intel and using the Israeli design team that came up with the Pentium M architecture, they didn't only catch up with AMD, they blew straight past them with the Core and Core2 processor lines, then really pushing it home with the 3rd generation of Core processors (the Core i chips).
AMD caught Intel napping, when it introduced the Athlon and Athlon 64 chips, but it then rested on its laurels and was caught napping by Intel.
With the Core lines of chips, did Intel even need to keep bribing their customers?
You could argue, that the damage had been done by then, that AMD's profit was stiffled, when they were "leading" the chip race. Did that capped profit mean they couldn't invest in the right people to design the Phenom, Llano and Bulldozer? Or would they have been struggling with the internal politics and supply problems, even if Intel hadn't played dirty?
I went through a string of AMD based machines between 1999 and 2006, but with the advent of the Core series, I have been using Intel again.
I am not loyal to either side. If AMD could match the Core performance using less power, I'd switch back for my next machine...
By big_D on 20 Feb 2012
Dumping the Pentium
Ups, just seen, that should read:
By dumping the Pentium and using the Israeli design team...
By big_D on 20 Feb 2012
AMD's demise was clearly bad management on multiple senior levels.
Having intimate knowledge of AMD during this time frame, there were bad decisions on multiple fronts.
The ATI acquisition was too costly. Not necessarily in dollars but in retaining ATI talent and integrating ATI into AMD. Too many upper managers trying to protect their fiefdom rather than implementing best practices from both companies. HR did it the best but it was led by an ATI VP with acquisition experience. Unfortunately, he left not long after the acquisition.
Product wise, ATI was acquired during one of its cyclical re-engineering and process change cycles. This would usually mean lower margins and lower sales. Under AMD it meant bottom line losses.
The most common phrase I heard from AMD upper management to ex-ATI management was, `We acquired you`. Not a great way to build best practices. This from a company that didn`t have acquisition experience.
The graphics product group at AMD went on to do its engineering update and process change while the CPU/chipset group couldn`t launch new processors. Even a die change was out of their grasp.
AT one point, the AMD Graphics Product Group had a stronger working relationship with Intel than it did with AMD processor and chipset groups.
It was, and still is, a train wreck.
With the shrinking of the PC market, it is debatable whether Intel will continue to let AMD survive.
By longtermtech on 21 Feb 2012
Thanks for pointing out the mistake about Mark Papermaster - that's now been corrected.
By Mikey_Jennings on 21 Feb 2012
I have only stayed loyal due to cost. Processors & Mothers boards basically. Memory 2 sticks instead of 3. It all comes down to what you can afford.
By roberttrebor on 20 May 2012
For more details about purchasing this feature and/or images for editorial usage, please contact Jasmine Samra on email@example.com
- Google I/O live stream and blog: how to watch 2014 Google I/O keynote speech live
- Google testing its own domain registration service
- Adobe announces first hardware: Adobe Ink and Slide
- Vote now in the PC Pro Excellence Awards 2014!
- What’s new in OS X 10.10? Apple Yosemite’s new features
- Samsung Z Tizen phone helps loosen ties with Android
- Microsoft rumoured to launch smartwatch this summer
- LG G3 launched: LG takes the wraps off smartphone that offers “more with less effort”
- LG G3 launch live video stream and blog: as it happened
- Apple fixes iMessage lock-in for Android switchers
- How Google Glass ruined my lunch hour
- Smartphone battery packs: can a USB power pack beat the festival battery blues?
- Windows Easy Transfer – not so "easy" in Windows 8.1
- Formula 1: what a difference virtualisation makes
- Office of the future: comfy chairs and tablets everywhere
- I went to Glastonbury and the only thing that got high was my smartphone
- Meet the robots helping teach children
- PaperLater: would you pay to print the internet?
- Amazon vs Kobo: how much to make the ebook switch?
- Phishing emails: how I nearly got caught out