Tuesday, January 11, 2011

Dirk Meyer, great CEO. I don't know what happened

Dirk Meyer surely was a major determinant on the survival of AMD this time. I am clueless as to the reasons why, after proven himself over and over, he got booted.

I did not follow closely AMD these years, but I know this:

1) AMD had a license to the x86 instruction set that prevented it from outsourcing more than 50% of its chips, among other things. The divesture of what now is Global Foundries and the outsourcing of production violated many of the terms for the license, nevertheless, Intel timidly enforced it, perhaps due to market conditions, for Intel, it seems optimal to keep AMD in torpor, at the brink of death, but still alive; blocking the whole "Asset Light" would have been fatal.

2) Intel could have kept the legal proceedings about its abusive market practices going on ad infinitum and deny AMD closure and cash

3) The AMD products were utterly unattractive, now, unattractive but perhaps cost efficient, as traditionally.

4) Fusion may be exactly what the emerging market of hand top computers (now known as smartphones) and keyboardless laptops ("tablets") require simultaneously powerful graphics, computing power, low chip count and power efficiency, which are integrated by Fusion probably better than any other option: nVidia's Tegra has the graphics but still requires a processor such as an Intel Atom. Intel, to date, has never been able to do powerful graphics. ARM implementations always require graphics coprocessors. For further advantage, Fusion does x86 and can thus tap the large development knowhow around Windows (*1) and Linux for x86.

5) Global Foundries has been a great success, covering the flank of were the chips are going to come from.

It does not make sense to abruptly change leadership right when long term strategic plans are coalescing into market opportunities.

I strongly dispute the thesis that AMD was flatfooted entering the market for Tablets, while it's true it got rid of Geode, the line of low power processors, for all its might, not even Intel has consolidated in this space; in my opinion the opportunity is now when AMD has a compelling option which could potentially take the world by storm. This has been the official reason for booting Meyer, so, it merits a closer look

Let us recap the timeline of smartphones:
1) In the beginning, there was not enough carrier bandwidth nor processing power to squeeze a sufficiently powerful computer in hand tops.

a) While useful, hand tops such as the Palm Pilot, without wireless and always-on access to the world could not become mainstream
b) There wasn't an ecosystem for applications for what we used to call hand tops back then, PDAs, "Personal Digital Assistants".
c) The mobile carriers had an oligopoly of abuse of people's communication needs; even today they continue to pretend that it makes any sense to charge $0.25 for a few bytes of transfer if they are SMS; $1 for mediocre quality "ringtones"; etc.
2) For reasons not known to me, the guys of Palm did not enter the wireless world, but left the opportunity for RIMM and Blackberry open. Blackberry positioned at the top market of data-hungry customers willing to pay hard for reading their emails wherever they were; they contributed by things like their internal messaging system(*2), which would not ever make sense in a perfect world, but since it provided an alternative to the abuse of carriers, through things like that it built a new market, what we now call "smartphones".

3) For reasons I don't clearly understand, Blackberry did not become mainstream, which left the opportunity for Apple to do things right and take the world by storm. It helped that RIMM burned foolishly opportunity costs en masse by courting Microsoft so that their crap would run in the Blackberries, instead of creating an ecosystem of application developers that not even today they have clearly managed to do, despite the obvious successes of the very different approaches about the same subject from Apple and Google.

4) Apple did everything right, it provided a convergence path to its fanatic user base of iPods, a fashion, hip, cool way to show status to those who already had smartphones; and the most important two things of all, it dared to bitch-slap silly the carriers about their asinine policies and created an ecosystem of application developers

5) Even when Apple did everything right, it left open the opportunity for Google to go full-tilt with free and open alternatives, which they took to heart and did everything right on their own.

Please observe that Apple did not get late into the smartphone area, it gave itself the luxuries of letting Blackberry saturate the market and even sabotaged the Motorola Rokr, an early attempt at an wireless-iPod-iTunes-phone; this, in my opinion, because the conditions were not given yet. This is what leads me to opine it won't matter whether AMD has a zero footprint in today's handtops; it is only right now when you can make truly compelling products in this space, and if it dares to gamble big, like it did with Opteron, AMD64, Hypertransport; which it won, or like Apple that dared to treat the carriers like its bitches, it could take the world by storm. But, as I have said several times before, there is no substitute for real knowledge at the head of a company. I guess that only engineers would have deemed the chances of going the Opteron/AMD64/Hypertransport route worthy of being taken; without the real engineering knowledge of whether the opportunity is there or not, you can not gamble your life. I fear AMD is losing real knowledge, vital to assess whether opportunities for great gambles are there or not. (*3)

(*1): AMD, please ditch Windows, really. Do not get tangled and tied into a platform in strong decline, on the contrary, embrace wholeheartedly emerging ones, and while at that, the free/open ones. Riding Windows-based products nobody will ever again reach the world by storm; this is by design: Windows is the establishment and 100% of Microsoft objectives are to preserve the establishment, so it will make sure it will be able to control, curb and nip (castrate) any establishment threat and it is only with that intention that it associates with upcoming initiatives.
(*2): Where I come from, the blackberries continue to be very entrenched at the mainstream. The reason is that people learned to text through their Blackberry PINs, SMS were absurdly expensive, and internet chat networks came much later; full blown email is not yet practical even today.
(*3): I dread showmanship in leaders. Examples: Steven Jobs, before ratifying his genius has thrice-empire-builder (Apple, Pixar, and Apple again) ran Apple, his first empire, into the ground, and had a diet of humble pie for years. Carly Fiorina and her nonsensical acquisition of Compaq; Jeffrey Skilling of Enron, etc. Observe that showmanship is all it takes to convince the meek to take great risks; while real knowledge is what determines their success, showmanship is therefore bound inexorably towards disaster, and the sooner it happens (Steve Jobs), the better.