Cambridge incubator spawns spinoffs aplenty
Are you allergic to VoIP?
A trip to St John's in Cambridge (England) provides an opportunity to see innovation in action. And one of the pioneers of Cambridge Technology - Dr Herman Hauser of Amadeus - was there this year to explain how it all came about - and to take modest credit for the effect.
It was ironic that in the same week as Bill Gates was being mocked for talking to the EC about innovation when he didn't even know about a deal done for the Zune music player with Universal (one dollar goes to Universal for every Zune sold) Herman Hauser was explaining how innovation came to Cambridge.
The signs were slow to be spotted. As recently as 1999, this reporter wrote a critical but hopeful piece about the Cambridge Network of high-tech startups, saying that, after many years of false dawn, perhaps it was starting to get going:
"Just maybe the time has come to start expecting real high-tech businesses to start developing in the "Silicon Fen" - the business districts around Cambridge," said a comment which has sadly been deleted from the ZDNet website. "The reason for hope: the long-missed 'magic ingredient' of professional management, is starting to appear in Cambridge."
Hauser was just another bright technologist when he founded Acorn - the company which hired wizard-level programmer Sophie Wilson who in turn gave the world the Acorn Risc Machine, or ARM - after producing the BBC Micro. But he was always inspired with the idea that Cambridge could be a rival to Silicon Valley, if enough people clustered around the University with commercial ambitions, rather than academic ones.
St John's is just one of the "incubator" centres now flourishing around the Science Park, and other industrial technology jewels which make the city of Cambridge one of the world's biggest innovation springs.
It's not just computer technology, either. These days, chemistry (scent detectors like Tourist) biochemistry (Food Detective) and medicine (the Model Gut) line up in St John's alongside computer software like Undo (a debugger which runs backwards) and Margo (filling in missing radio bursts to provide higher fidelity audio) looking for a chance to turn a good idea into a growing corporation.
"You know most of your startups won't do it," said Hauser candidly, introducing this year's crop of hopefuls. "But as a venture finance investor, you know that you only need one or two stars to create enough money to start another dozen new ideas."
Here's a look at the pick of the list on show at this year's Open Day in Cambridge.
Sometimes, innovators persuade you they've done more than they have; years later, you feel an idiot for giving them publicity. But when OptiSynx touted a new high-resolution clock as "enabling faster wireless broadband" many reporters retreated into scepticism.
"Replacing the existing clocks in mobile phone base-stations with Optisynx, not only gives improved user experience with fewer dropped calls," says the company blurb, "but also allows an operator to fit more calls into each base-station..." - a claim which few have felt confident enough to repeat.
Dominic Mikulin, CEO of the St John's sponsored startup, impressed observers with the claim that his low-cost silicon chip would more than just rival caesium-based clocks. He even got traction for saying that his technology was 1,000 times more accurate than rubidium based timers (one second in 30,000 years, compared with 30 years for the radioactive decay machine).
"If you want accuracy, in the past you had to use caesium - but that's old tech and not safe, said Mikulin. "Using rubidium is safer, but not accurate enough - so people have used rubidium, and then re-calibrated the rubidium clock frequently with a GPS-provided signal, which is caesium-generated."
His theory is that GPS clocks can be hacked, while his can't. "GPS signals can be disrupted by weather or environmental conditions. But also they're easy to tamper with - using very inexpensive kit - leaving you with a system that can't be audited - so if you rely on GPS, you're no longer in control."
In fact, Optisynx's predictions for the value of the new timer technology in mobile comms are quite conservative. When Mikulin says that future wireless broadband will run at 70 megabits and then 150 megabits, few in the mobile area would disagree.
But quite how having a better clock will make this possible isn't clear - hence the reluctance, perhaps, for the story to hit the headlines.
Undo...debugging in reverse
Any programmer will know how you debug. You "step through" the code with a machine emulator, which shows you what is happening after every instruction.
"The trouble with standard debug," says Greg Law at Undo-Software, "is that you can see that a problem has occurred all right - but you often don't get to see when the problem was caused. It can be thousands of steps before it shows its effects."
So his software runs both ways.
First, it runs through the software, and throws up bugs in the normal way. But then, it runs backwards. "It's like a black box flight recorder. If a problem occurs, you trace it back to the source, you go back in time to the point where things originally went wrong."
The first version, now shipping, is for Linux developers on Intel X86 family computers, but there's no reason to fear that other versions can't be produced in due course, says Law. "That's why we are here; we need the money to do the next versions."
Sponsored: Benefits from the lessons learned in HPC