Neural Magic

Neural Magic is an innovative company at the forefront of artificial intelligence, specializing in high-performance inference serving solutions designed for the deployment of open-source large language models (LLMs) on standard CPU and GPU infrastructures. Founded in 2018 by MIT professor Nir Shavit and research scientist Alex Matveev, Neural Magic has its headquarters in Somerville, Massachusetts, and has grown rapidly as a leader in AI optimization. They focus on helping enterprises maximize the potential of their existing computing resources while reducing the costs associated with AI deployment.

The company's flagship product, DeepSparse, is designed to facilitate effective deployment of computer vision (CV) and natural language processing (NLP) applications. By utilizing advanced techniques like model quantization and sparsification, Neural Magic enhances performance while maintaining high accuracy. The company’s mission is underscored by its commitment to open-source practices, allowing organizations to run LLMs securely and cost-effectively within their infrastructure—whether it be in the cloud, on-premises, or at the edge. Neural Magic’s deep expertise empowers IT teams to scale AI solutions efficiently, ensuring that companies can harness the full capabilities of their AI models.