Stefan Ideler, co-founder of i3D.net, on ECOFED: “Radical innovation is inherently risky”

Imagine selecting the cloud services you want from a marketplace, while behind the scenes those services are provided by different providers who may not even know each other, let alone have worked together before. That is what five organisations are working towards in the ECOFED project. It is exactly the sort of challenge that appeals to Stefan Ideler, co-founder of i3D.net.

His father worked at ASML and Philips NatLab. From an early age, Stefan Ideler learnt that radical innovation is a process of trial and error, and that most innovations fail. Sometimes that is because the right technology cannot be developed; other times, the market ultimately opts for something else. And sometimes it is simply down to sheer bad luck.

Huge potential

Those childhood lessons are proving particularly useful in the ECOFED project. The outlook on life Ideler developed is one reason why i3D.net is keen to take part. “Investing time in ECOFED is a gamble, but if it succeeds, it could become something really big,” he predicts. And i3D.net, which has a presence in data centres across Europe, would certainly benefit.

The company has another, more immediate interest. “For our engineers, it’s great fun because they’re working on something genuinely challenging. They’re building something completely new, bringing together lots of different technologies and ways of thinking. Our project partners are constantly challenging them. That’s hugely valuable, both for the engineers themselves and for us as an employer,” he says.

i3D.net is also involved because of the project’s wider societal importance. “ECOFED started with the ambition of giving customers more choice beyond the hyperscalers when it comes to cloud services. Since Trump has been concentrating so much power in his own hands in the US, the importance of a sovereign cloud has become part of that too. We’re not a charity, but it’s good to know that, through this project, we can help address these issues.”

Minimising latency

Ideler was not quite so focused on wider societal issues when he founded i3D.net with his business partner Stijn Koster. “We were students together and both keen gamers. Online gaming was taking off while we were at university, and you soon realised that latency was an issue. So, in early 2004, we installed a few servers in a data centre ourselves and created an online space where people could play against one another. At the time, there was one big, dominant company in the Netherlands offering the same service. But its connections weren’t up to scratch and things kept crashing, so we thought: we can do better than that.”

And so i3D.net began, with one overriding aim: to deliver the best possible gaming experience by minimising latency. As online gaming became more successful, game developers started appointing companies to host their games exclusively. i3D.net was one of them and soon grew to become the largest in Europe. The time had come for a data centre of its own in the Netherlands.

Until then, i3D.net had focused entirely on the consumer market. That changed when its first gaming company wanted to buy server capacity. Ideler remembers it well. “Suddenly, we found ourselves in the B2B world. We were involved in tough contract negotiations, with large legal teams on the other side of the table. We’d entered a completely different world.” It was a world i3D.net quickly got to grips with, and one that meant renting data centre space in other countries. Those data centres were linked by a fibre-optic network. Along the way, i3D.net became something of a network specialist too.

The economics of hyperscalers

i3D.net was approached about joining the ECOFED project through the Cloud Infrastructure Coalition (CiC), an initiative led by TNO. Ideler saw it as a great opportunity. “We were facing growing competition from hyperscalers. Demand in the gaming industry rises and falls dramatically. Far more people play games in the evenings and at weekends than during the day on weekdays. And when a new game has just been released, demand goes through the roof. Hyperscalers are ideal for handling short-lived spikes in demand. But some customers were moving all their workloads to them.” Many of those companies are now coming back, however, having realised that a hybrid cloud is a better long-term option.

i3D.net rents space in 61 data centres across Europe, but even that is not enough at peak times. “That’s why we also use hyperscaler cloud services to handle spikes in demand. It’s ridiculous, really, that we have to turn to American companies for that. We asked ourselves: why can’t you handle peaks in demand just as easily within Europe as you can in America? In Europe, you might have to deal with ten or twenty different data centres and manually integrate their services yourself. That simply doesn’t scale, and it’s difficult to manage.”

Learning from one another

So there are plenty of reasons to take part in the ECOFED project. Beyond the business case and the societal benefits, there is another reason to join the consortium. “The mix of companies is really interesting. In terms of our DNA, we’re most like BIT: another company full of techies. Alongside cloud services, InfoSupport develops bespoke software and does a lot of work for local authorities. You can see that reflected in how the company is run, and we learn a lot from that. AMS-IX is a fantastic partner to bounce ideas off when it comes to networking. And TNO brings proper researchers to the table, which is an area we don’t know nearly as much about. The exchange of ideas between these very different organisations generates so much new knowledge and insight.”

In that respect, taking part in the project is already a success for i3D.net, Ideler says with a laugh. “One thing I learnt from my father is that, with radical innovation, you should be pleased if 10 per cent of projects succeed. That might sound like a low success rate, but if that holds true for all the IPCEI projects (Important Projects of Common European Interest, of which ECOFED is one), just think how much we’ll achieve together.” Given the strong demand for a sovereign European cloud, one innovation risk has already been reduced considerably: the risk of developing a technological marvel that is easy to use but that nobody actually wants. “In that respect, we’ve got the wind in our sails.” All the more reason to make the new protocol and marketplace a success too!