Last month, Intel announced that it is rerouting its roadmap, bypassing a whole generation of x86 processors and launching dual-core versions of its desktop and server products next year. By doing so, it can deliver more computer power to customers while offsetting the heat and power dilemma that its current GHz-centric trajectory is on.
This month, arch rival AMD said it will launch dual-core versions of its AMD 64 processor, targeted at both desktops and servers, in the second half of next year. Multi-core versions of AMD 64 are also on the agenda, said AMD.
So, it looks like AMD is echoing Intel’s line. But back in February, it was Intel that seemed to be trailing when it announced plans to add 64-bit extensions to its x86 line, a strategy that AMD has followed for years, and which it finally delivered on in 2003.
Who’s really winning? As with most things in this industry, it depends what angle you’re looking from. Viewed one way, the last few months illustrate the problem that has always thrown AMD onto the back foot. AMD has often scored a short-term advantage over Intel in terms of performance or price, only to see the surprisingly nimble giant use its massive R&D resources, manufacturing muscle and marketing weight to rapidly close the gap before moving ahead again. This is presumably the way Intel likes to see the world.
Of course, there is another point of view: the way AMD likes to see the world. From this standpoint, AMD is not scrabbling for a short-term technical or cost advantage, it is delivering on a long-term strategy that will eventually position it as a serious contender to Intel rather than a low-cost alternative.
AMD argues AMD 64’s integrated memory controller and HyperTransport interconnects mean the platform will not be hostage to the diminishing returns normally associated with scaling up. So far Intel has delivered little detail on how it intends to deliver on its dual-core promises.
More importantly perhaps, AMD insists that its multi-core plans shouldn’t be a surprise to anyone. The AMD 64 platform was always designed to easily accommodate multiple cores, it says, and this was never a secret. A quick check supports this point, with the company making clear references to the platform’s scalability from early in its development. In the increasingly risk-averse world of corporate computing, this apparent consistency might actually be AMD’s strongest card.
Intel has always been a company with a long-term vision, even if it is simply to have the whole world operating on its silicon. However, some recent key announcements, such as the launch of 64-bit extensions earlier this year, or the revelation of its dual-core strategy, have looked less like the unfurling of a carefully crafted roadmap, and more like a magician pulling a rabbit out of a hat. Of course, Intel might justifiably argue that to pull a rabbit out of a hat, you first have to have the rabbit on hand, to know it’s in the hat, and to be able to efficiently execute the delivery of said rabbit from said hat on schedule. The company has a habit of stealthily including new features in its designs, only to reveal them when it is convinced the timing is just right.
Nevertheless, as both firms gallop along their roadmaps, AMD can justifiably portray itself as a serious contender in a two-horse race, rather than a perennial also-ran.
