Quantum-inspired technology halves AI model size – with no performance loss
Spanish firm Multiverse Computing has launched HyperNova60B2602, a compressed language model that halves memory requirements to 32 GB without compromising performance.
What happened?
The Spanish AI company Multiverse Computing has released a new language model, HyperNova60B2602, freely available on Hugging Face. The model is based on a development of an open model structure and has been halved in size compared to its starting point. It requires only 32 GB of RAM to operate, while maintaining accuracy and performance at near-identical levels.
Key facts
| Modellnamn | HyperNova60B2602 |
|---|---|
| Minneskrav | 32 GB |
| Teknik | CompactifAI |
| Distributionsplattform | Hugging Face |
Why it matters
Large language models consume immense amounts of energy and require hardware costing millions, creating significant barriers for smaller players. Through its CompactifAI compression technology, Multiverse Computing demonstrates that it is possible to drastically reduce hardware requirements and latency. This enables the deployment of advanced AI models on standard workstations and servers.
Who is affected?
The initiative primarily concerns companies and developers seeking to deploy large language models locally on their own hardware or in edge environments without needing to invest in expensive server clusters. AI researchers and smaller companies seeking cost-effective alternatives to cloud-based API services are also positively affected.
Impact on the EU
The technology is developed by the Spanish firm Multiverse Computing, and the model is available globally via the Hugging Face platform. This makes it directly accessible to European companies seeking to comply with EU regulations regarding local data processing and energy efficiency.
What else you should know
HyperNova60B2602 is based on a compression method inspired by quantum physics and tensor networks, which enables significantly reduced memory usage without the performance loss often associated with traditional quantisation.
Quick answers about this story
Vad har hänt?
När hände det?
Varför spelar det roll?
Påverkar det EU?
The link opens in a new window and leads to the publisher's own site.
Källan har spårats automatiskt från utgivaren via Aheadlines signalkedja.
AI-verktyg i artikeln
Topics
Get similar news straight to your inbox
The reader's room
Send in a question or an addition. The newsroom reads everything before it's published and replies when relevant. No AI-generated text – just people.
Sign in to submit a comment or question.
Read the article through your role
- Where does a window open: market, tech, distribution or pricing?
- What can you test in under two weeks without burning capital?
- Who do you need to talk to this week to validate the assumption?
Generated angle — not editorial analysis of "Quantum-inspired technology halves AI model size – with no p"