Cheryl Hung

Joining Helio as an advisor to drive globally sustainable computing

I am delighted to join the advisory board of Helio, a Zürich-based startup with a compelling vision for global sustainable computing.

The need to reduce IT’s climate impact

I first encountered Helio in Stockholm in 2019, where I spoke about how data centers emit as much carbon dioxide as the airline industry, and how better resource utilization could reduce IT’s contributions to climate change.

Typically 80-90% of global server capacity sits idle, wasting energy and money. This can be improved with the right architecture and tools. For example, Spotify found that moving to Kubernetes improved CPU utilization by 2-3x, which is not only great from a cost and engineering perspective, but also reduces the energy required and emissions generated to run those workloads.

A model for sustainable computing

Helio’s CTO and co-founder Christoph Buchli presented a highly complementary platform, loosely described as “Airbnb for data centers”. It matches sellers with spare compute capacity with buyers, who can run workloads at prices they claim to be on average 30% cheaper than other cloud providers.

Utilizing spare capacity reduces wasted energy and creates a new revenue stream for companies with existing data centers. It empowers small players to join the cloud computing market and, in Helio’s words, “democratises compute”.

Such a platform will need a thriving ecosystem of developers, which is why the co-founders approached me. I will be advising Helio on strategies to grow developer adoption, build technical communities, and be good open source citizens. The co-founders are eager to follow the principles of social good, including donating spare compute resources towards Covid-19 research.

If you are a developer or operator with a real world use case, do get in touch with Kevin and Christoph. The ideal workload is compute-heavy and not latency critical, and furthermore offers some flexibility in either geography or time of execution. The first deployments were 3D rendering workloads from CGI companies, but training machine learning models would be another ideal use case.

Our responsibility in IT

Those of us who work in cloud infrastructure have a responsibility. Significant impact rides upon our choices, as millions of servers sitting idle wastes our planet’s resources. Using renewable energy sources is an important first step towards reducing carbon emissions. But intelligent workload deployments leading to better utilization of existing data centers could result in an order of magnitude improvement.

Enabling carbon neutral compute benefits everyone for the next hundred years or more. This is why I find Helio’s vision compelling and I am delighted to join as advisor.