The Switch--Part Two

Add Your Comments

We explain why AMD may still go down, and why they’re going to have to be lie low for the next few years, but we still like the new captain a lot better.

Yesterday, we pointed out that Derrick Meyer was likely to run the nuts-and-
bolts operations at AMD better than Hector Ruiz.

Will that be enough to keep AMD alive? Make it a success again? That really
depends on what you mean by “success.”

I think for many reading this, “success” means “making chips as fast
or faster than Intel’s, and soon.” If that’s your criteria, Meyer will not
succeed. He’s not even going to try for that kind of success, because it’s
just not doable. If you think an AMD exec can give you a superior chip by
next year, you don’t want Derrick Meyer running the company, you want Jesus
Christ.

The CPU industry (and to a lesser extent the GPU industry) is like a huge
ocean liner, only a lot worse. You can’t turn the thing around fast. It takes years to create
and test a truly new CPU design. It takes at least a year to take that idea
and turn it into silicon (and AMD has never been very good at doing that).
It takes two-three years to build a fab get the equipment for it and get it
to full production. It takes at least a year for such a fab to go through a
process shrink, or a chemistry change like high-K.

Making the fastest CPUs in the world is a very slow process!

And for most of these criteria, if what you’re looking for is another Hammer

to put AMD on top, AMD is at or near starting from scratch. The cupboard is

bare, and even if Meyer started serious work on all this yesterday, it would

take years for them to become reality.

I’ll put it this way, if some ruler of an oil-rich Arab country/national

investment firm showed up at Meyer’s office, pulled out his sword, plunged it

into his desk and announced, “By Allah, I swear, I declare jihad on

Intel! Your company will build a killer CPU to destroy them!! All our money

is yours, here’s ten billion to start!” Mr. Meyer could not possibly begin to

make an Intel-beater, much less destroyer, until at least 2011, and building

the fab capacity to replace Intel’s would take even longer than that.

If AMD couldn’t do that until 2011 with infinite resources, do you think they

can do better with inadequate ones? No, AMD is stuck with essentially

tinkering some more with the old Hammer design, shrinking the process,

changing the fab chemicals, for the next two-three years, at best.

People repeat the mantra, “AMD needs to stay alive so Intel won’t rape us.”

I have news for you. For the high performance part of the market, AMD is dead,

and has been for some time. Intel isn’t going to rape us, they

are raping us; prices on the higher-end models are a good deal higher

than they used to be. They’re going to do it even more with Nehalem, with

its expensive “overclocker socket” with huge price increases for slight speed bumps.

That will be the only game in town for about a year, followed by

nonoverclockable models.

(Now if you think this raping isn’t as
bad as some have predicted, I’d agree, and never thought it would

be. We’ll talk much more about this in the weeks and months ahead, but for now,
understand that AMD is no longer “protecting” us).

So AMD is out of the high-performance ballgame for at least the next two-

three years. Bad news for us, but that still leaves the other 95% of the

market, and there AMD looks to have a reasonably decent fighting chance. If

they can move most of their production from bankrupting X2s over to reasonably well performing and somewhat better priced 45nm CPUs

quickly in 2009 and start making some money on the video card and mobo

chipset end, the losses should stop and AMD should recover enough in the

minds of investors to have a decent chance to somehow finagle its way into

financing that New York fab.

Make no mistake about it, this is retrenchment. Intel and AMD fought, and

Intel won.

The “do or die” reach for glory is over, outside of the antitrust

bets AMD has made. “Success” for AMD for the rest of this decade is going to

be written down to “Stop losing money and make enough to finance the next

fab. Try again next decade.”

It’s humbling; it’s not glorious; it’s just

necessary, and doable. Even these modest goals will be tough enough to

reach, and if the world economy gets whatever the US is catching, AMD could

still go down. In the longer-term, AMD may still end up losing out due to

problems caused by lack of resources.

Nonetheless, if keeping the AMD ship afloat is doable, if it’s within the

power of any CEO to do it, it is far more likely this guy is going to do it

than his predecessor.

A lot of people think I’ve been pathologically harsh on AMD the last few

years, to the point where they think there must be some other deep dark secret reason for it, like bribes from Intel, or at least
losing all my money and/or family members because of AMD. The truth is much

simpler than that:

I ODd on the steady diet of deceptions, on the relentless dumps of BS piled higher and deeper on my computer screen, on load after load of dishonest language implying one thing but saying another, on statements hinting but never quite saying things I knew weren’t true, couldn’t be true, statements that would eventually have to blow up in their faces.

I saw, read and heard a captain who seemed to be drifting further and further from reality, someone who seemed greatly annoyed at the distraction coming from anyone who interrupted his pontifications to point out that his ship was sinking. It seemed like crazy people were running the place, and that got me more than a bit crazed myself. But this has gotten considerably better recently, and I think it will get better still.

Some may say, “We agree about the captain, but what makes you think the first mate is going to be so much better?” It’s a reasonable point, but you read interviews like this one, the guy at least seems to realize how much deep doodoo AMD is in, and what he realistically can do, now, to keep the ship from sinking.

That may not prove to be enough, and the ship may end up sinking anyway, but at least you think this guy knows he can’t BS his way out of it and will go down fighting. That’s a whole lot better than having a captain who seemed to live in la-la land having “corporate visions,” fixing holes in the hull with clever wording in press releases and if the ship inexplicably went down anyway, to go down lying.

I’ll leave you with an example which indicates why I think things are going to improve. You’ve heard about “asset light.” For eighteen months, Ruiz talked about its existence, but acted like AMD would explode if he gave even the slightest clue as to what it was. After a while, many concluded that this was just a new fecal chunk from the corporate diarhhea machine, including the stock analysts who pretty much said so while they were downgrading the stock. In the interview linked above, Meyer simply said what it was:

“. . . Meyer says AMD is just months away from a major restructuring that will spin off its manufacturing operations into a separate company, with new ownership.”

Let’s put aside the question of whether or not this is a good idea. Was that so hard to do? What was the big deal? Did the company collapse as a result? Would it have been so disastrous if Hector had said the same thing a year ago?

Meyer didn’t say much, just enough. A little more openness and honesty will go a long way, and I think we’re going to get it.

Ed


Leave a Reply

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