Kelvyn Taylor

Over the past few years we've all become pretty much immune to the hype surrounding the launch of a new CPU. Let's face it, they seem to pop up every other month, usually trumpeting a minor speed or technology ramp.

But this week's launch of the AMD Athlon 64 was a different beast for a number of reasons, most of them good, and a timely reminder to Intel that it can't always have everything its own way.

First, let's make it clear that I don't really care too much about the performance numbers. If the Athlon 64 is within spitting distance of the Pentium 4, that's near enough to be roundly applauded in my opinion.

Why? Well, the truly significant benchmark numbers can't be found anywhere. And they never will be, because they're the ones that measure the non-existent 64-bit performance of the Pentium 4.

The Athlon 64's 32-bit performance is a sideline, a nice bonus that gives users an excuse to buy one now, although most will happily spend much of their working lives running 32-bit x86 code (natively, of course, as opposed to the much-maligned emulation in Intel's 64-bit Itanium CPU).

It doesn't matter. When the AMD 64-compatible version of Windows XP Professional ships, unless Intel digs deep into its conjurer's hat and finds a rather large rabbit (and I've been around long enough not to totally discount that), AMD will have the show to itself.

Admittedly, the size of that show is a big unknown, but it would only take a couple of massively popular 64-bit games to tip the balance.

Whether 64-bit CPUs will have any effect on what mainstream consumers do with their PCs is debatable and, I believe, largely irrelevant.

Terabyte memory addressibility isn't a compelling feature unless you're designing an aircraft wing or interpreting oilfield seismic data. Native 64-bit applications will eventually filter into the market, but most people will just continue running their existing 32-bit stuff.

The point is that 64 bits for the price of 32 is something that will resonate with a lot of consumers: you can't ignore the 'bigger numbers are better' principle, no matter how good the technical arguments are.

But perhaps the most important point is that AMD has pulled off a useful marketing and technology coup, assuming that there aren't any hidden nasties lurking in the Athlon 64 silicon or the supporting chips.

Intel's surprise announcement at its developer forum of the unimaginatively named Pentium 4 Extreme Edition might keep it ahead on the 32-bit numbers game for a while, but the impression of a knee-jerk reaction is very strong.

It's also going to be very hard for Intel to counter the Athlon 64's market positioning: Intel's current trick of positioning the Athlon XP against the Celeron rather the Pentium 4 just won't work with the Athlon 64.

Of course, Intel isn't facing any major threat to its domination of the CPU market. But it's the small dent in its perceived technological leadership that will hurt it the most, a pain that will be aggravated if the Athlon 64 starts to fly off the shelves.

Intel's hyper-threading is an excellent technology and still a key differentiator, but it's almost impossible to market to a non-technical audience.

And the desktop market isn't Intel's only worry: in the server market, the 32-bit and multiprocessor capabilities of Athlon 64's big brother, the Opteron, are already generating a lot of business interest.

Pundits can talk until they're blue in the face about how nobody needs 64 bits on the desktop but, as ever, consumers will make that decision for themselves.

And if the outcome's the wrong one from Intel's point of view, so be it. In the meantime AMD can sit back and enjoy the limelight.

Like this story? Spread the news by clicking below:

 del.icio.us     Digg this     reddit!