How AMD's Failures Are Triggering An Intel Monopoly
AMD Held The Future In Its Hands. And Dropped It.
There was a time only a couple of years ago when the much smaller AMD was able to totally dominate its Goliathan competitor Intel in virtually every benchmark and measure of performance for the dollar. During those halcyon times, AMD accumulated an extremely enthusiastic group of evangelists who touted their team's superiority and enjoyed nothing more than rubbing Intel's Big Blue face in it.
That behaviour was completely justified at the time. What AMD managed to pull off was somewhat analogous to Kia blowing out Toyota, or the (dearly departed) XFL knocking out the NFL. The war was won on the retail counter, and there is no way to dispute that the junior sized ankle biter AMD certainly had the goods over the bloated Intel conglomerate.
That all changed in July, 2006 when Intel introduced its Core 2 series of processors. To put it into a layman's perspective, it was as if a new SUV had been introduced that cost $3,000 and ran on solar power. Jaws dropped all over cyberspace and the popular rush to this new processor series resembled the stampede for tickets to an Abba reunion concert. To say that AMD was left in the dust would be an understatement. It might be more accurate to compare it to the look on the Coyote's face when he realizes that he's flown off the cliff and there is nothing but air between him and the canyon floor.
So what did AMD do? Well, on exactly the same week that Core 2 debuted, they decided for reasons that will be forever debated, to purchase leading Canadian videocard manufacturer ATI. There were many good reasons for the takeover. Videocards utilize a Graphics processor (GPU) rather than a CPU to do the taxing data work of constantly repainting the pixels on your screen. For reasons too technical to cover in a layman's article, these GPUs process data in a very different manner than the CPU that runs your computer, but they run that data wickedly fast. New technology is on the horizon that would allow a GPU to process the types of data previously reserved for CPUs, albeit with restrictions, as fast as video data is currently being processed. Think: NASA Space Shuttle vs. Millennium Falcon.
For AMD to get its greedy paws onto some of that supercalifragilistic ATI technology was undoubtedly a brilliant strategy. Not only was ATI a popular videocard market leader with a healthy P&L sheet, but the potential of fusing that blindingly fast GPU with AMD's series of CPUs could create the Holy Grail of comuputing: Supercomputer power on the desktop. Not only would that spark a paradigm leap in capabilities for the computer users of the world, but it would also boot Intel back to the Stone Age where they would have to lithograph circuits onto granite with chisels.
There is something about the plans of mice and men that applies here... What looked so good on paper turned out to be an unmitigated disaster in execution. And yes, execution would be just about what the AMD executives might deserve for blowing what could arguably have been the best corporate move in computerdom since Bill Gates convinced IBM to ditch CP/M in favour of Quick & Dirty OS... er, sorry... MS-DOS.
What happened in the next twelve months would make a novel far more riveting and bloody than any wild fantasy of Stephen King on acid. Wall Street viciously spanked the new AMD-ATI (anagram: DAAM-IT!). The market was not at all ecstatic about AMD taking on billions of debt on rather unfavourable conditions to swallow up ATI and reduced its market cap to a measly $7 billion, which was just about a billion dollars more than the price AMD paid for ATI. It was as if the stock market said to AMD your company is worth next to nothing, the only asset you have is the value of ATI that you have just purchased. Again in layman's terms, it was not too dissimilar to buying a house for $200,000 in Cleveland's zip code 44105 which has the highest rate of foreclosures in the country, and watching your new property devalue to $50,000 within months.
What must have been going on in the office of AMD's Dr. Hector Ruiz would probably have been a top rated reality TV show. With all the aplomb of the Titanic's helmsman, Dr. Ruiz and his crew loaded the dinghys with some of the most malodorous compost to have been issued by any company since Enron. They were scrambling to prove that they had something, anything up their sleeves and just managed to humiliate themselves even more, if that were possible.
AMD tried to deflect some of the criticism of their outdated CPU range by launching a fully irrational and baseless hype campaign for their new K10 architecture which was going to devastate Intel's Core 2. Unfortunately even AMD itself had to admit in its own releases that their new CPUs would outperform Core 2 by 2%. Two percent! Given the fact that Intel was just to introduce its own new architecture named Penryn, followed up by Nehalem, which represent quantum leaps ahead of Core 2 performance levels, the K10 challenge is completely laughable. The ability to beat Core 2 by 2% was like showing up at the dragstrip in a car that would definitely beat Don "Snake" Prudhomme's 7-second 1970 Funny Car by a hair to the lights, but coming up against Tony Schumacher in his sub 4.5-second 2007 Mustang.
On the first anniversary of Core 2's launch, AMD still had nothing but empty blather on the store shelves. The K10 had not been launched, and not even one independent benchmark had been released. That led some pundits to comment that it is very difficult to benchmark air, as the very existence of a marketable K10 CPU began to enter into question. Was AMD experiencing manufacturing hitches, or was the K10 a Trojan Horse, or worse yet an outright chimera in a vain attempt to keep the stock price from complete meltdown?
AMD's jinx even extended to their once mighty ATI subsidiary. When they finally managed to get their first fully-Vista compliant DX10 videocard onto the market, they were almost half a year behing their nVidia competition (which is an eternity in the compubiz) and their new offering was welcomed with "driver issues" and collective yawns by the market.
There is nothing that corporate raiders love more than a company that they can buy then sell off the pieces for more than what they paid for it, and AMD had ineptly maneouvered itself into that category. The ATI subsidiary was still worth more or less $6 billion as a standalone, plus their various fabrication facilities and other assets made AMD worth much more than its market valuation.
What would happen if the private equity vultures were to feast on AMD's carcass? It's quite unlikely that there are any other Intel wannabe competitors on the horizon. Producing CPUs is an extremely asset-heavy process which requires billions of dollars of investment. The only realistic potential new CPU manufacturers would be Motorola or Samsung. However, Motorola has recently disentangled itself from its messy PowerPC CPU affair, and Samsung is doing so well in the chip markets which it dominates, that it likely needs to get into the CPU wars like George Bush needs another hole in his head.
Is there a PC CPU monopoly on the horizon? Perhaps. An Intel-only world presents some very interesting prospective scenarios. It is rather doubtful that the new Clinton or Obama administration would leave an Intel monopoly alone, not only on antitrust issues, but also on homeland security concerns. After all, does it make sense to equip the various threatening rogue states like Iran, North Korea, Cuba, Venezuela and Liechtenstein (!) with supercomputing power that can be purchased for a few hundred dollars? America has abandoned its traditional backbone of courage, independence and self-reliance and is now based on an internet backbone of YouTube, SEO and kiddy porn. There is little doubt that the level of vulnerability to a massive cyberattack would dwarf the economic damage wrought by 9/11.
The only businesses to truly benefit from the Bush administration have been Iraqi morticians, and the myopia demonstrated by the Commerce Department to the dynamics and critical national importance of the computer business is nothing short of shameful. A more conscious administration might decide that it is in the country's best interests to nationalize Intel, and there could be some very good reasons for such a move.
Unless AMD is able to pull its collective head out of its netherparts long enough to restore the confidence of the markets and its customers in very short order, an Intel monopoly may be inevitable. AMD's annus horribilis hasn't just shot the company in the foot, it's turned the whole corporate entity into a colander. The few legitimate executives such as the pre-buyout ATI CEO Dave Orton, tucked his little rattail between his legs and jumped the sinking ship. More were likely to follow. It is difficult to believe that if the AMD CPU executive triad of Hector Ruiz, Dirk Meyer and Henri Richard had been replaced with Curly, Moe and Larry, they would have possibly done a worse job.
Maybe when President Bush's term is up, he could get a job at AMD?
- Dr. Onicil Lah, Ph.D., Professor and Chair of the Department of Computer Science, University of Chandrapore, India. ;)
P.S. Before the AMD fanbois start their flame war, kindly note that this Hub was written on my AMD dualcore, the latest in a long line of AMD CPUs I've owned and appreciated.