Analysis: Highs and lows of Intel Developer Forum

So another Intel Developer Forum (IDF) has come to an end.

IDF is Intel’s main forum for new releases and updates on the older products and as such is key to understanding the company’s direction. It’s also a chance for developers, manufacturers and academics to meet and exchange ideas that will shape the direction of computing in the next decade.

While we’ve got lots of stories and video from the show there’s a lot to process in the three days. Visitors are bombarded with new information about what’s coming down the line and blue-sky thinking about the future.

So I’ve selected my top five highest and lowest moments of the show, so you can get an idea of what’s important and what’s not.

Highs

5. USB 3.0

USB isn’t tied to any one vendor but Jeff Ravencraft, the president of the USB Implementers Forum (USB-IF), is an Intel technology strategist in the Digital Enterprise Group (DEG) and was on the show floor showing off the first USB 3.0 products.

Rather confusingly the USB-IF is calling the technology USB Superspeed (one wonders what USB 4.0 will be called) but whatever the marketing bods do to create confusion won’t do too much harm, because the speed of the technology will ensure it’s a huge seller.

The new system increases the speed of data transmission tenfold, to a cracking 4 Gbit/s, although in the real world it’ll probably only give around 75 per cent of that. It’s stunningly fast, and pretty much means game over for FireWire.

USB 3.0 requires a new USB hub, which NEC were displaying, along with a new cable and receiver, but crucially the rectangular form-factor of the ports remains exactly the same and it will be backwards compatible with USB 2.0 devices.

By keeping ports the same customers won’t have to bin their old hardware, manufacturers won’t have to redesign systems too much and everyone benefits. It’s a textbook case of how to develop and implement a new standard.

4. Harrison Schmitt

Harrison Schmitt is, to date, the only civilian ever to have walked on the moon’s surface, and the only scientist to have made the trip as well.

Despite only getting a slot late in the day (we only found out about it when we were leaving the show and the tannoy announced his talk) he held audiences spellbound with his show on the route to the moon and why we should go back.

One point was especially telling, particularly with an IDF audience. Very few new products or systems were invented in the process of taking humans out of the gravity well. However, the speed of change in things like processor design and computing had drastically increased, which Schmitt estimated had paid back the cost of the missions seven-fold.

At the end of the presentation he hung around to chat and it was a privilege to shake hands with the great man. When asked if he got irritated by those who insist we never went to the moon he was sanguine.

“It doesn’t bother me a bit; I know where I’ve been,” he said.

3. Light Peak

The demonstrations of Light Peak at the show were interesting, but it is the future technologies it will impact on that really got people thinking.

Light Peak is an experimental optical technology that will have the capability to transfer data between devices and components at 10Gbit/s, with a theoretical maximum of 100Gbit/s predicted.

It’s also flexible, with the ability to support a wide variety of standards and connections. Suddenly the mess of connectors the average power PC user has to have could be seriously reduced.

As one Intel Fellow pointed out, we’ve run about as far as we can with wire connections and optical is the next logical step forward. The biggest barrier is cost but with Light Peak Intel appears to have cracked a major part of that.

2. Core i7 Mobile processors

Mobile computing is all the rage in the market at the moment and the audience was packed out to see the latest mobile chips from Intel.

The new quad-core i7 processors look the absolute bomb. They are fast, power-efficient but also highly customisable. If you need speed then overclocking is perfectly possible, where as if long battery life is your priority then the chips can shut down individual cores to save on power use.

Manufacturers are keen to get the new chips into systems as soon as possible, for two reasons. Firstly the chips really are something that will excite the business and consumer markets and they offer original equipment manufacturers a vast range of new system designs to get sales rising again.

However the chips, like any new hardware, are a bit pricey and if demand is strong it will help people get used to the idea of spending large amounts of cash on notebooks again. With netbooks selling well several manufacturers are worried that they will be losing sales to the low-priced systems and the i7 could help with that.

1. Westmere

Without a shadow of a doubt Westmere was the star of IDF, and so it was fitting that Sean Maloney, currently Intel's golden boy, should get to show it off.

Westmere looks to have more grunt and features than almost any mainstream processor technology at the moment and will be crucial to securing Intel’s future in the server rooms and data centres of tomorrow. As the industry moves towards mass virtualisation and a bigger emphasis on cloud computing the new chip will be vital.

Intel has also packed a lot of stuff onto the die itself, including support for AES encryption, a very welcome development in today’s security-conscious times.

Westmere is also Intel’s first 32nm mainstream processor and as such is key to the company’s move to scale down processor technology in the future. The company is refreshing its manufacturing every two years at present and is pushing down to 22nm by 2011. Looking even further ahead the company is predicting getting down to 4nm by the 2020s.

Smaller is usually better when it comes to transistors, in terms of power use and productivity of the finished product Intel claims (although AMD disagrees) and Westmere is an important step on the road.

Read on for the low points of this year's event.

Lows

5. Numbers

This has been my ninth IDF and I've never seen the numbers of attendees so low. Compared to the boom years IDF felt like it was deathly quiet and sections of seating were taped off in the keynotes to pack everyone in at the front of the hall.

The situation was even worse in the exhibition halls, with the number of stands from suppliers well down and Intel setting up booths of its own to keep the density of the show the same. There was some interesting stuff there to be sure, but the crowd was mainly big-name companies and there were very few of the wacky start-ups that usually make IDF such fun.

Of course Intel was upbeat about the situation, stressing that we shouldn't so much look at the size but feel the quality. They have a point in a way, but it's a worrying sign that the industry still has a long way to go before it reaches the levels of interest seen a few short years ago.

4. Atom and 3D TV

Intel's new Atom system on a chip is actually a rather clever piece of engineering, but the uses it has been designed for were a little uninspiring.

The company was keen to stress that it was not looking to reinvent the PC in TV's image but that's what it looked like to some delegates and I've got my doubts that people are quite ready to swap out a familiar technology for something new. After all, Intel's Media PC system was a horrible bit of kit when it first came out and hasn't got much traction since.

But what was really mystifying was the company's insistence that 3D TV was going to be the wave of the future. We've had people trying to flog us 3D TV systems for decades now and the technology stubbornly refuses to catch on.

People don't want to have to wear special glasses to view something, particularly in a public setting like a sports bar. And I doubt cinemas are going to rush to buy every customer a set of glasses to watch their films.

The demos delegates saw were all very impressive, once the glasses had been handed out. I have to say that seeing Bono in 3D was annoyingly tantalising; he looked almost close enough to punch.

3. Whiteboards

This year's IDF had a large number of whiteboards scattered around the venue, where delegates were supposed to write their view of what the future of computing could be and what industries it would spawn.

I'm sure this idea may have sounded good in the planning meetings but the end results were less than impressive. If someone has a good business idea then they are hardly going to write it down for all the world to see and some of the delegates recognised this by suggesting sharks armed with lasers or hybrid zombie-helicopters.

Whiteboards have a place in meetings for brainstorming, but on a conference they looked a bit desperate. By the end of the show they were also covered with graffiti by people trying to use them for self promotion, which was a bit off-putting.

2. Larrabee

For a technology that was supposed to be out this year Intel is remaining very quiet about Larrabee, and everyone noticed.

Intel's forthcoming graphics processor has been talked up by the company so much in the past that you'd have expected a major push this time around to prime the market. Instead we got a grudging mention from Otellini, a short demo and virtually no technical details at all.

What we did get was a talk on programming for the chip itself, but graphics software is only part of the challenge. With AMD a clear leader in the graphics market and Nvidia firmly in second place Intel has a mountain to climb before it can come up with a product that is going to win widespread support.

It was all very disappointing, and from private chats with a couple of PC manufacturing representatives the industry isn't holding its breath for anything spectacular from Intel in the next 12 months on this front.

1. Where's Pat?

On the aforementioned whiteboards one message kept getting put up, and almost immediately wiped off by Intel staff - "Where's Pat." One delegate managed to keep it up by writing it in Hindi but almost everyone was missing Pat Gelsinger.

Since his surprise departure from the company just before the show the rest of the industry has been coming to terms with the departure of one of the men who typified the engineering culture of Intel. Gelsinger was a true enthusiast, his keynotes among the best attended at the show and while he was sometimes wrong about technology he was a feature of IDF.

With Gelsinger going the last remains of Intel as an engineering-led company have disappeared and it's clear the accountants and marketers are in charge now.

Of course, there are still skilled engineers in the company and will continue to be so but one effect of the bloodletting over the last few months has been to keep them firmly in their place.

Intel has undergone a sea-change in attitude and direction and it is going to be interesting to see how it develops as a company.

Some might say that the loss of engineering focus was inevitable as the market matured but on the other hand a company at the apex of the computing industry needs to have an engineering vision, and that seems to be less in focus than before.