Multiverse Computing SL raises $570M in Series C

Multiverse Computing SL, a startup specializing in AI model compression technology, has secured $570 million in a Series C funding round, bringing its valuation to $1.7 billion. The financing was led by Forgepoint Capital International, BNPP SIVF, and Bullhound Capital, with participation from investors including Santander Alternative Investments and Tikehau Capital. This substantial capital injection follows the company's previous Series B funding, marking a significant valuation increase from earlier rounds.

Multiverse's innovative technology focuses on reducing the size of artificial intelligence models using tensor networks, which originate from quantum physics. This compression allows AI models to retain efficiency while significantly lowering the hardware requirements, reportedly cutting the hardware footprint by 80% to 95% without substantial accuracy loss. Multiverse CEO Enrique Lizaso emphasized that the traditional need for expensive infrastructure for powerful AI models is obsolete, suggesting that their technology permits running high-performance models on more accessible and less costly devices like smartphones and personal computers.

The implications for AI model deployment are notable. Smaller models can not only operate faster and with less power consumption but also enable the deployment of robust AI models on devices that previously supported only basic functions. This capability has the potential to shift workloads from cloud environments to local processing, thereby reducing latency and keeping data securely on the originating device. Through its platform, Multiverse is also addressing AI deployment across varying scales, from local devices to cloud-hosted solutions, aiming for enhanced AI deployment efficiency and security.

The broader AI market is poised for transformation as Multiverse's compression technology could dismantle existing infrastructure constraints that favor larger data centers. The prospect of running complex AI on local and edge devices, as demonstrated with Meta’s Llama 3.3 70B model on an Intel Xeon processor, positions Multiverse as a potential disruptor in sectors such as finance, manufacturing, and defense, where heavy computational tasks can now be localized. Multiverse’s high-profile partners, including Allianz and the Bank of Canada, underscore the broad applicability of its innovations.

Looking ahead, the freshly secured funding will enable Multiverse to expand its range of AI model compression offerings and further invest in proprietary R&D. As the company continues to refine its technology, integrating operations within enterprise infrastructure without reliance on hyperscale cloud services remains a focal objective. Progress here will ensure that enterprises can maintain data sovereignty while harnessing the power of advanced AI models internally.