Will The US Follow Japan, Korea and the EU?

With Japan, Korea and Europe finding that Intel breached anti-trust law, who will be next? The EU sent its report on the Intel case to the anti-trust authorities in 27 countries. The US FTC, which has been investigating Intel for a year, received a copy of the report and could be the next country to bring an action.


In America, it is likely to come down to politics. Under President George W Bush, the general idea was, apparently, to let dominant companies do pretty much what they liked so long as they didn’t directly hurt consumers.


Since the basis of the complaints against Intel has been that they offered rebates to customers who restricted their use of AMD chips, it could be hard to prove that rebates hurt consumers. After all, rebates mean cheaper chips.


On the other hand, it is possible to interpret the definition of consumer harm more widely. If a company which is dominant in a market acts directly to marginalise its competitors, the reduction of competitive pressure harms consumers because it acts to keep prices up and reduces the pressure for innovation.


This does happen. I’ve heard an Intel exec say that the timing of the introduction of a new process step (in that case when Intel would replace a silicon dioxide gate with  high k material) would depend on whether and when the competition developed a competitive high k material.


Then, of course, there’s Wintel. Has Wintel hurt consumers? Without a doubt. Remember how quickly digital watches and calculators commoditised?


It happened in  a few short years under the remorseless effects of Moore’s Law which means the cost of  production of chip-based products halves every 18 months. 


The first IBM PC, using MS-DOS and an Intel 8088, was launched in 1981 for $1,565. The succeeding 28 years of  Moore’s Law cost erosion should have reduced that $1,565 to half a US cent!


Of course better processors, bigger memories, HDDs, CD drives, DVD drives, flat panels, an increasingly enormous OS, and a fabulous volumes of applications software and much else besides have combined to keep the price up at around the few hundreds of dollars mark today.


But computers have only relatively recently started to commoditise. For many years the price stayed close to that original $,565 price tag – and much of that was due to the power of Wintel.


Be that as it may, it is not going to be the effects of Wintel that determines whether a US anti-trust law-suit is initiated; it will probably not even depend on whether the FTC believes that Intel did, or did not, offer rebates to customers which were conditional on customers limiting their use of AMD chips; it will depend, it seems, on the zeitgeist in the Washington corridors of power.


On whether the new administration takes a tougher attitude to anti-trust enforcement than the last one.


The awesome power of the US lobbyists will, of course, have a part to play. But AMD can afford lobbyists as well as Intel.


It will, most probably, come down to the instincts of those who surround the new president.


TOMORROW MORNING: The Ten Best High-Tech Movies



  1. Good gracious, that’s a neat argument Tom. I would never had thought that engineers were responsible for the asset price bubble. But I suppose you’re right. Blame it on Gordon (Moore not Brown).

  2. What has been happening for the last 20 years is that Moore’s law and engineer’s ingenuity and hard work has caused the capability of PCs and technology goods to rise and their price to fall. Through the magic of ‘hedonics’ government can treat capability improvements as price falls when calculating Consumer Price Inflation.
    But central banks target 2% overall inflation when setting interest rates. Deflation in technology prices means prices of other goods can be allowed to inflate and interest rates can be lower.
    Low interest rates means more people want to borrow money and banks create more debt based money. The collateral banks understand for their loans is property so house prices rise and banks do more business and make more money. Engineer’s hard work and Moore’s law are converted by the system into profit for banks and a property price bubble.
    Personally, I think we would be better off with less competition, tech goods being more expensive and technology companies like Microsoft and Intel getting rich rather than bankers and estate agents.

  3. O, you are right if you only look at the balance sheet, but if you consider the $5 billion being pumped in by the good Sheiks of Abu Dhabi, who will doubtless have excellent contacts among the Washington lobbyists and won’t want to see their investment out-lobbied by Intel, then I think AMD won’t necessarily be out-gunned in the influence campaign.

  4. “But AMD can afford lobbyists as well as Intel.”
    That’s the only statement that strikes me as being outright wrong. Given their respective balance sheets, I’d say Intel can outspend AMD enormously on lobbyists. If it comes down to a lobbying battle then Intel will win, hands down.

  5. I don’t know, Rupert, but I assume the cost of semiconductors became an increasingly large proportion of the total PC cost as time went on

  6. Of that original cost of the PC, how much was semiconductors? I presume that was the 16K model with no display adaptor, which was basically memory, the 8088, a couple of Intel support chips (a parallel interface and the programmable interrupt controller? Something like that) and a lot of TTL.
    If you can’t do that now in the absolutely most insanely cheap programmable logic nubbins, I’d be astounded (and I’d quite like to know what that part is). The rest of the stuff – case, board, discretes, power supply, connectors, etc – will be untouched by Gordon…

Leave a Reply

Your email address will not be published. Required fields are marked *