Qualcomm Technologies and Spain-based Multiverse Computing announced on August 5, 2026 a partnership to deploy compressed AI models on Qualcomm’s data center accelerators using the CompactifAI platform. Compressed models running on the Dragonfly AI200 and AI250 chips achieved response times up to 93% faster, 44% performance improvement, 45% memory usage reduction, and 21% lower energy consumption with no accuracy loss. Multiverse Computing raised 215 million dollars in Series B funding in June 2025 and is currently pursuing a Series C targeting up to 570 million dollars at a pre-money valuation of 1.7 billion dollars.
Source: Read the original article
