Distributed cloud Hive generates buzz as data center power demand grows

  • Hive is the latest distributed/decentralized cloud provider to garner investor attention

  • Cloud startups like Hive are aiming to capture unused compute resources to improve sustainability and efficiency as data center power demand grows

  • Other players in this space include Akash Network, Flux, Golem and more

Buzz, buzz, buzz. That’s the sound of a happy Hive, thanks to a fresh influx of €12 million (nearly $13 million) from investors looking to give its distributed cloud offering a boost.

Founded in 2022, Hive’s business model is fairly simple – and, if you’ve been paying attention, familiar.

In short, its hiveNet platform weaves together unused compute and storage power from user devices and then allocates those resources based on demand. Users are rewarded for contributing their idle resources with a discount on their Hive subscription.

The company currently has more than 25,000 users and contributors around the globe who both serve up their compute power, store files and run workloads (including AI inferencing, video processing and more) are using Hive.

If it sounds a lot like what Akash Network, Golem, Mutable, Hydra and others are doing, that’s because it is. The difference is Hive’s focus seems to be more about gleaning unused compute power from individual users than from data center servers (FluidStack, for instance, does the latter). Hitherto, Hive’s focus has also been less enterprise-oriented than its competitors.

Hive previously raised €7 million (roughly $7.6 million) in seed funding from Global Ventures and OneRagtime.

With the influx of fresh cash from SC Ventures and OneRagtime, Hive plans to boost marketing efforts to get startups and enterprise clients to join its community, invest in R&D for feature development and strengthen its technical capabilities to better meet the needs of artificial intelligence workloads.

All about sustainability

While almost all the distributed and decentralized cloud providers tout efficiency as a key reason to use their services, Hive explicitly frames its offering as a “greener” alternative to data centers.

Among other things, it noted that its user-supplied compute doesn’t require additional cooling – which accounts for a significant portion of the energy used in data centers.

AvidThink founder Roy Chua said the decision to market Hive based on its sustainability features is a “smart move.”

Why’s that? Well, data center power consumption continues to make headlines and not in a good way.

According to the International Energy Agency (IEA), data centers across the globe consumed an estimated 460 terawatt-hours (TWh) of electricity in 2022. That figure could more than double to over 1,000 TWh by 2026, it warned.

“Updated regulations and technological improvements, including on efficiency, will be crucial to moderate the surge in energy consumption from data centers,” IEA said in a report.

McKinsey recently estimated data center power consumption in the U.S. alone will jump from 17 gigawatts (GW) in 2022 to 35 GW in 2030.

Just this week National Grid CEO John Pettigrew predicted energy demand from commercial data centers “will increase six-fold in the next 10 years.” National Grid provides electricity to over 20 million customers each in the U.K. and U.S.

That increase is a big deal given data centers are already estimated to account for up to 3% of global energy consumption.

There’s another reason, though, why Hive’s marketing choice is smart: namely the fact that distributed computing still faces an uphill battle for adoption.

Chua explained that while the push for distributed computing has continued despite the crypto crash, there have yet to be any real breakthroughs.

“The vast majority of what runs in data centers is owned by businesses,” he said. “So I think the reality is that you probably need to appeal to businesses in order to grow the business and expand it. To convince businesses to buy into that…the governance model around that needs to be sufficiently validated, the encryption and confidentiality model needs to be there.”

Chua said some companies could dip their toe into the distributed computing waters by moving unique workloads that have less stringent privacy requirements. But he added that many companies are unlikely to take the leap into running all their workloads on computers and software they don’t control and haven’t vetted simply for compliance reasons.

“It’s a really interesting concept but there’s a lot of work left to be done,” he concluded.