Multiverse Computing

Latest AI news, models and releases from Multiverse Computing. ['CompactifAI']

Hardware & Inference 🇺🇸

Multiverse Computing: CompactifAI nearly doubles Llama 3.3 performance on Intel Xeon 6

Multiverse Computing announced that its CompactifAI technology can nearly double the performance of the Llama 3.3 model when running on Intel Xeon 6 processors. The solution uses neural network compression techniques to accelerate inference without significant loss of accuracy.

Multiverse Computing's AI Model Compression Boosts Performance on Standard Hardware

Multiverse Computing has introduced a technology for compressing AI models that significantly improves performance on standard hardware. The solution enables large language models (LLMs) to run on ordinary GPUs at high speed, potentially reducing infrastructure costs and making AI more accessible.

Multiverse Computing

Meta AI (GNews)24.07 · 00:03