Multiverse Computing, a tech startup based in Spain that builds software to make AI models run more efficiently, has raised US$570m in its third major round of outside investment.
Investors valued the company at US$1.7bn before adding the new US$570m, making it an official unicorn.
The valuation of the company has increased by five times since its previous investment round, driven in part by it proving that powerful AI models can run at full performance on a smartphone, inside a sovereign slot online data centre, or on a factory floor with no cloud connection.
What does Multiverse Computing do?
The startup developed a platform called CompactifAI, a compression technology pioneered by Co-Founder and Chief Scientific Officer Dr Román Orús, that applies tensor networks – a mathematical framework from quantum physics that finds and cuts out redundant patterns in AI – to the problem of AI model compression, which makes AI systems small enough to run on regular laptops or smartphones rather than in a data centre.
It’s all about making AI run through a facility that can do more work with less energy.
Multiverse also provides companies with an all-in-one software package to manage their AI workloads smarter.
Rather than relying entirely on expensive cloud giants, businesses can add Multiverse’s software layer on top. It shrinks AI models, manages GPUs and routes tasks efficiently, saving “hundreds of millions” in infrastructure costs, according to the firm.
“Multiverse has spent the past several years building a different path: AI that can run efficiently in the cloud, on-premise and directly on devices,” says Enrique Lizaso Olmos, Co-Founder and CEO of Multiverse Computing.
