Q.ANT open-sources software for its light-powered AI chips

German startup Q.ANT has released a free, open-source Q.ANT Native Computing Toolkit on GitHub for programming its photonic AI processors. Developers can write Python or C code and test it in a simulator without the chip; real hardware access will come via IONOS cloud or an on-site server in the coming months.

• Q.ANT Native Computing Toolkit is free on GitHub with a commercial-use license
• Python and C are supported; the simulator runs without Q.ANT drivers
• First version only runs pre-trained models: digits, objects, image outlines
• No PyTorch plug-in yet, so developers wire things up by hand