AMD: Come Back In 2011

Add Your Comments

If you’re looking for AMD to become competitive against Intel’s best, it’s pretty hard to come to any other conclusion after its financial analyst day.

The roadmap for AMD over the next couple years is pretty simple: 45nm Denebs and Shanghais and Caspians, then batten down the hatches and ride out the recession storm.  Whatever we see in the next couple months is pretty much all we’re going to get for the next 18 months or so.  No high-K until 32nm.  No new CPU design until 2011, no Fusion either. 

There was more than a bit of confusion about just when 32nm chips would actually show up. I suspect we’ll end up seeing Deneb-generation CPUs shrunk to 32nm sooner, and something like Bulldozer showing up later.

Other than that, the strategy the next couple years appears to be “cut expenses to the bone and try to make a little more money from the CPUs.  There’s nothing exciting here; I guess sheer survival will be exciting enough. 

AMD didn’t revise its financial estimates like Intel did, but don’t worry, they will, and it will be bad.  The general rule of thumb for these two companies is that, all things being equal, AMD does proportionately worse than Intel when overall conditions are bad (and proportionately better when overall conditions are good).  It’s possible AMD might gain some marketshare the next few quarters, but the price to pay for that will be low prices for the extra CPUs they do sell. 

But odds are you’re not concerned about how many CPUs AMD can sell to HP, you’re concerned about whether or not AMD can sell a world-beating CPU to you.  Unfortunately, the answer is no, and will stay no for the next couple years.  AMD will be able to sell you a number of other things you might want, or something good enough and/or cheap enough for your purposes, but not that. 

Ed

Leave a Reply

Your email address will not be published.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