Introducing

Get started
Benefits
No-Hardware AI
GPU-class performance on commodity CPUs. No special hardware required.
Run Anywhere
Run in containers in the data center, in the cloud or at the edge.
Reduce Costs
Deploy and scale models on cheaper and more commonly available CPU servers.
How it works
Neural Magic Machine Learning Lead, Mark Kurtz, shows how it's possible to achieve GPU-class performance on commodity CPUs for deep learning by using the Neural Magic inference engine.
Watch Video
Nir Shavit
Co-founder
We believe software-led machine learning will empower data scientists to answer bigger questions, in a broader setting, and with greater ease.

Use Neural
Magic For

Image Recognition
Image recognition systems powered by Neural Magic’s Inference Engine can be used to create sophisticated computer vision applications.
Object Detection
We are driving sophisticated computer vision with machine learning.

Neural Magic Careers

Want to build the engine that powers the next machine learning breakthrough?
Learn More