While official confirmation from chip giant Intel still remains absent, it is all but certain that its upcoming processor, codenamed Prescott, will be a hot one. Sources and documents close to the company have spoken of current demands roughly 10-20% greater than earlier predicted, resulting in power dissipation levels in excess of 100 Watts for low-speed Prescotts.

More important, though, is the notion that this development seems to have caught Intel by surprise. While Intel certainly expected a process issue like this at some time in the future, it is abundantly clear that it did not foresee near-term issues with respect to its 90nm process technology.

And while Intel execs scramble over what to do with their processor-turned-furnace, executives at AMD are faced with unique problems of their own. The introduction - if you can call it that - of AMD’s Hammer series has gone about as smoothly as sandpaper. For a company that has bet its future on the success of Hammer, AMD seems alarmingly disorganized and confused, unable to adhere even to its own roadmaps for more than a quarter or two. In its current financial state, AMD is not likely to survive more than a year or two if it fails to generate some increased revenue.

Yet for both companies, the current situation may prove more familiar than strange. AMD has been on the ropes before, and, thus far, has managed to escape demise each time, though occasionally due less to its own merit than to ‘unforeseen difficulties’ at Intel. This may not be the first time Intel has looked ready to deliver the final blow, only to be thwarted by its own unfortunate circumstance.

A 10% increase in power dissipation certainly isn’t enough to stop Prescott in its tracks, but it’s more than sufficient to slow the ramp of the new processor while Intel pauses for some damage control. New data emerging from various sources, as well as simple logic, suggests that Intel may not be able to ramp Prescott as fast as it had initially planned. This buys AMD some very valuable time, as it appears to be struggling itself to increase Hammer’s clock.

In addition, a power-hungry Prescott may present AMD with an interesting opportunity to turn the marketing tables on Intel, by appealing to those who shied away from AMD processors in the recent past due to heat and noise concerns.

Whether or not AMD realizes it, Intel’s Prescott difficulties have given Hammer exactly what it needs: a second chance at a good head start. If AMD plays its cards right, the abovementioned points could slow Intel just enough for Hammer start off with the bang it needs, instead of the dull thud it has made thus far. Establishing even a small early foothold could allow AMD to snatch up a couple percentage points of market share. That may not seem substantial, but consider that just 5% of Intel’s sales is sufficient business to move AMD from the outhouse to the penthouse.

Fate has dealt AMD a bit of luck, and when you’ve posted losses for several consecutive quarters and are faced with dwindling cash reserves, you need to capitalize on any luck you get. If it wishes to avoid serious financial crisis, AMD must work to take advantage of Intel’s current predicament. That means composing itself, following through with its roadmaps, and delivering a product where Intel can’t.

The great irony is that Intel’s fiery Prescott may very well be what pulls AMD from the flames.

Dan Mepham