Welcome to Downing Ventures' new series, Why We Invested In, that gives the team's insight into the exciting deal closes they have been working on and how they got there in the first place.
High-performance computing (HPC) is a sector that a tiny percentage of the world’s population know anything about. Even smaller is the percentage of brains that understand it so well they know how to solve critical problems that the industry face - which is exactly what Cornelis Networks is doing.
I have been fascinated by computing since I was a kid growing up in Iowa and my dad brought my first Apple computer home. From that moment I have watched the computing industry transform and more specifically, the growth of supercomputing and how it is enabling so many of society’s major developments – quantum mechanics and molecular modelling, climate research, space exploration, and chess! But of course, whenever technology progresses, supercomputers must up their game too.
Phil Murphy and Gunnar Gunnarson, the co-founders of Cornelis Networks are industry veterans and long-time employees of Intel. They understand the industry better than most and were visionaries in Intel’s Omni-Path Architecture business. Essentially, Omni-Path Architecture enables quicker processing and output with minimal lag and power consumption – a critical element with the increasing data requirements in next-gen technology that is driving an unbelievable demand in the scale-out interconnect.
I believe in this product because Phil, Gunnar and their team have the experience and expertise and have proven that their product works. There are very few players in this market and with partnerships with the US Department of Energy’s National Nuclear Security Administration and original equipment manufactures such as Dell and Lenovo already in the place, this is an incredible vantage to scale up from.
AI, data analytics and simulation techniques are used by almost every scientific, commercial and government organisation in the world, each relying on enormous computational power. I believe that the future of this depends on the innovation of the scale-out interconnect with its value as a market segment being in the multi-billion-dollar realm. Furthermore, the unstoppable growth of the cloud is a powerful propellant for additional use cases. As AI and other deep learning technologies make their way to the cloud, the entire stack will have to transform, which directly benefits Cornelis’ core products and the teams' core competencies at rapidly scaling in an industry known to be in continual evolution. Watch this space.