Le projet THINK

Less Gold Rush and more Boring Factory – The evolving AI mindset.

From community curiosity to real-world inference – showing how local language models run with day-zero Intel hardware support.

In this post, we’ll dicuss how to run responsive, CPU-only applications using a quantized SLM in the GPT-Generated Unified Format (GGUF). 

The latest Intel® Xeon® 6 processors deliver performance advantages across key enterprise workloads, enabling companies to deploy fewer servers and still deliver a similar aggregate performance level compared to AMD EPYC solutions

Intel® Xeon® processors can deliver a CPU-first platform built for modern AI workloads without added complexity or overhead.

Transform enterprise documents into insights with Document Summarization, optimized for Intel® Xeon® and Intel® Gaudi® with automated NUMA-aware scheduling.

We are thrilled to announce an official collaboration between SGLang and AutoRound, enabling low-bit quantization for efficient LLM inference.

In this guide, you’ll learn multiple aspects of optimizing the Search and Recommendation model deployed in Production using Intel Xeon CPU servers. 

Using three U.S. Department of Energy (DOE) supercomputers, researchers from the University of Southern California (USC) and DOE’s Lawrence Berkeley National Laboratory developed new ways to model these complex systems with greater precision than ever before.