Peter Hopton was an electronics engineer at the University of Sheffield who, in the mid-2000s, studied the history of liquid cooling in supercomputers. The way those early systems - like the Cray 2 which was nicknamed "Bubbles" - worked, was by pumping a coolant. The liquid used was also less environmentally-friendly. But Hopton thought he could go back to liquid cooling and do one better.
Green Revolution uses mineral oil as the immersion fluid, which is considerably cheaper than the 3M Novec fluid, but Hopton says the costs even out over an extended period and the Iceotope system is safer because the fluid used is non-flammable. But Iceotope has Hong-Kong based Allied Control and US-based LiquidCool in the fray which also use dielectric fluids.
Hopton's big lesson: Realising how important it is to know one's market. In 2009, Iceotope tried to gain traction in the financial services market and even secured a case study with Credit Suisse. But it soon realised it was the wrong market to pursue: buyers were conservative and used to dealing with big brands, not a brash start-up. "If you don't know your market, it can end up driving the business in the wrong direction," says Hopton. Lesson No.2: Study the IP landscape. Spend three times as much time studying other people's patents as you do writing your own. The common mistake is to go to the patent office the very next day after inventing something, pay money and have the patent pending. Hopton says that is probably the worst way to do it.
India is one country on his list, says Hopton, acknowledging there are several data centres here, and it is a high growth market. Schneider Electric, one of its major investors, and already present in India, could become Iceotope's channel to market.