Intel's hope for a boost from Xeon

Intel has high hopes for its new chip

When Pat Gelsinger took the stage in Santa Clara this week to launch the new Xeon 5500 he told the audience that this was the biggest chip launch for Intel since the Pentium Pro 10 years earlier.

"You won't get this level of attention paid to other processor launches," he said. And, judging from the amount of effort Intel put in, he may well be right. But is the new Xeon the revolutionary product that the company claims?

Well, on the face of it, the new Xeon is certainly a major step forward. The Turbo power management features are going to prove very handy in managing power, and the early benchmarks look very good indeed for dual-socket servers. The integrated memory controller is also a positive move, and improved virtualisation support is what the market wants.

The launch will also have given AMD something to think about. AMD's Shanghai processors are now looking at some very stiff competition, and Intel has also stolen some of its clothes. Intel was keen to stress the upgrade possibilities of the new platform, something AMD has ruled the roost on in recent years.

Meanwhile, the rest of the industry players in the server field have been rapturous in their response to Xeon. HP, IBM, Sun, Cisco and Dell all came out in force to praise the new chip as the 'next big thing' in the server market.

As the launch day wore on, however, one couldn't help feeling that it was all a bit much. Yes, it was a spiffy new chip and it's nice that everyone was so keen, but wasn't there just a little too much hype going on? Gelsinger's an honest sort who obviously believes in what he is doing, but to call the Xeon " the greatest leap in performance in the history of data processing" seems a tad over the top.

The mood wasn't helped by the minor earthquake that happened just before the launch. While only measuring 4.9 on the Richter scale (Californians don't get out of bed for anything below a six it seems) it did lead to plenty of 'jokes' from Intel executives about the ground shaking new qualities of the processor.

The fact of the matter is that the Xeon is a major step forward, but there's more to it than that. These are tough times in the semiconductor industry and Intel is hurting more than most. To add to the pain, the old arguments that companies always need faster chips are patently no longer true, since processor technology has advanced to such a level that many companies have all the run speed they need with second-generation hardware.

Server manufacturers are hurting too. While some sections of the market are holding steady, and the decline in sales is nowhere near as severe as for the PC market, sales are still down and servers are generally seen as one of the most profitable sectors of the market. Not only is there revenue in selling the hardware itself, but in selling high margin services and support to customers.

Thus we saw everyone getting behind Xeon in a way that mimicked Obamamania last November. Intel and its customers don't just want to show how wonderful Xeon is, they need to. The companies involved are hoping that Xeon will kick-start the server market back into the kind of sales it is used to. The crucial question is, will it?

Sad to say on the face of it the answer is no. All the feedback we are getting from companies is that there is no way that cash-strapped businesses are going to rush out and buy the new processor in bulk. At a time when IT departments are struggling for every penny, there's little chance of being able to sell the board on a major new hardware revamp.

Intel was pushing the cost saving aspects of the chip hard. It would pay for itself in eight months, was the common refrain. That may well be so, even if such estimates are notorious in the IT industry for needing not so much a pinch of salt as a tanker-load of Saxa. But, once you add in the cost of junking your old systems and the downtime to set up the new ones, it adds up to a pretty penny.

It's going to take much more than the Xeon's excellent new features to get IT spending up to the levels that vendors are used to. It's looking likely that, from a business growth standpoint, we can write off this year and maybe 2010 too.

Given that kind of outlook companies aren't going to spend to upgrade hardware when they might not be in business in 12 months' time. Memories are still fresh of the last downturn at the turn of the century, when top technology was sold at fire-sale prices by bust firms that neglected the fundamentals of business in favour of fancy kit.