If you haven't yet, you really owe it to yourself to take the time and read AMD's antitrust complaint against Intel. The document is largely written in plain language and makes for great reading, believe it or not. The complaint lays out in juicy detail AMD's allegations against Intel, explaining the structure of the PC market and how Intel's web of incentives, rebates, quotas, and intimidation tactics allegedly serves to box AMD's market share into the 15-20% range. Whether or not you buy the argument and whether or not it has legal merit, the complaint presents a compelling case. Intel's purportedly anticompetitive measures are explained in great detail, with multiple and specific examples cited for each practice. Names of individual OEMs, distributors, and retailers are named and connected with those examples, too. More dirty laundry is aired in the 48 pages of this complaint than I've seen pass through my inbox since the founding of this site nearly six years ago. What fun!
I must admit, as I read it, that a number of the examples made things fall into place, from small specifics like whispers I'd heard about large OEMs pulling out of AMD product launches at the last minute to larger issues, like the bizarre and persistent incongruity between AMD's apparently strong competence at strategic planning, design, engineering, and manufacturing amazingly complex microprocessors on the one hand and AMD's seeming ineptitude at sales and marketing on the other. If even a portion of what AMD alleges in the complaint is true, it explains a lot. They really had little chance to push their market share higher in certain sectors.
I'm no legal expert by a long shot, but I will venture a comment or two on some of the larger issues as I see them.
There will be a big discovery process that must happen before this case goes forward in court. In fact, AMD told me its lawyers have already filed to have relevant documentation preserved by the parties involved (such as big PC OEMs). They're saying this case could potentially have the largest discovery process that's occurred to date. AMD's legal types will use this process to seek evidence of Intel practices that violate antitrust law.
Whatever that process turns up, we probably know enough now about some of Intel's tactics to say that those tactics violate the spirit of antitrust law in that they are anticompetitive. We can also probably assume that Intel's size, influence, and market share are sufficient to qualify it as a potentially monopolistic enterprise.
But can AMD establish in court that Intel has technically violated antitrust law? One practice the lawyers are likely to zoom in on is Intel's system of "first dollar" rebates. The complaint alleges that these retroactive discounts are triggered when OEMs purchase a certain target number of Intel CPUs in a quarter, at which point the firm is eligible to receive rebates on all CPUs purchased that quarter. If the OEM doesn't meet this targetwhich the complaint claims Intel sets individually per OEM per quarter based on sales projectionsno rebate check is forthcoming, and the OEM's profit margins are toast. AMD alleges that by setting these targets, Intel is able effectively to dictate the OEM's product mix, by extension dictating AMD's share of the PC processor market.
The complaint explains how similar mechanisms are supposedly imposed on distributors and retailers, as well as PC makers. These practices could well be judged as running afoul of antitrust regulations.
Antitrust law also asks questions about how such practices harm consumers, and here AMD may have rougher going. Sure, anticompetitive practices by Intel might hamper competition. But Intel pours huge amounts of money into R&D, develops new standards for the industry to follow, and regularly passes along the happy benefits of Moore's Law to consumers in the form of faster, better, and cheaper microprocessors. No to be too trite, but it's hard to find the harm in a Pentium D 820 processor for $241. I suppose if there were justice in the world, AMD's market share would have risen perceptibly since the introduction of the Pentium 4 "Prescott," after which Intel's processors gained some small ground in clock speed, but offered few performance gains to go with their additional heat and power consumption. Instead, Intel turned in record numbers last year. But such technology stumbles have been rare for Intel over the long run, mere blips in a trajectory of clear, consistent progress on many fronts.
|Phanteks Eclipse P400 gets a tempered glass option||0|
|GeForce driver 375.57 is prepared for Titanfall 2||1|
|Radeon 16.10.2 drivers add support for October's big games||10|
|Strong revenue doesn't stem red ink in AMD's fiscal third quarter||26|
|Razer unsheathes the Blade Pro gaming laptop||17|
|Acer XB241YU G-Sync display stalks the FreeSync competition||19|
|PowerColor Devil Box cages high-performance graphics cards||24|
|Samsung builds 8GB LPDDR4 packages on its 10-nm process||8|
|Latest Nintendo console can Switch form factors on the fly||129|
|A real "console monitor" would be 720p @ 30 Hz ;P||+55|