A naukri.com initiative
Eu-Startups
3d
282
Spanish startup Multiverse Computing raises €189 million to scale technology that compresses LLMs by up to 95%
- San Sebastian-based Multiverse Computing secures €189 million in investment to further develop CompactifAI, a compression technology reducing LLMs by up to 95%.
- Bullhound Capital leads the Series B funding round with support from investors like HP Tech Ventures, SETT, and Forgepoint Capital International.
- Enrique Lizaso Olmos, Founder and CEO of Multiverse Computing, highlights the efficiency improvements offered by their AI model compression technology.
- Multiverse Computing, founded in 2019, specializes in quantum-inspired AI model compression, with their CompactifAI compressing LLMs by up to 95% with minimal precision loss.
- The company has over 160 patents, 100 global customers, and aims to advance efficiently compressed AI models for various industries.
- Multiverse's CompactifAI models offer significant performance improvements and reduce computing requirements, enabling new AI use cases across industries.
- CompactifAI models are highly compressed versions of open-source LLMs, remaining accurate, faster, and decreasing inference costs by 50%-80%.
- HP Inc. has praised Multiverse Computing for its innovative approach to making AI applications more accessible at the edge, potentially benefiting companies of all sizes.
- CompactifAI utilizes Tensor Networks, a quantum-inspired technique developed by Multiverse's Co-founder Román Orús, to simplify neural networks and optimize AI models.
- Compressed versions of top AI models like Llama, DeepSeek, and Mistral are currently available, with more models in development.
- Investors recognize Multiverse Computing's breakthrough in AI processing efficiency and its potential to revolutionize the global deployment and application of AI models.
Read Full Article
16 Likes
For uninterrupted reading, download the app