By making the AI Hardware Composer accessible to students, researchers, academic institutions and commercial partners, IBM’s goal is to accelerate the process of discovering and creating new hardware systems that support AI’s ever-increasing software energy demands.
Operating from IBM Research’s open source analog libraries on the cloud, the tool is designed to help anyone from coding novice to experienced developer experiment with creating their own neural networks and tune analog devices to produce accurate models.
This launch coincides with the two-year anniversary of IBM’s AI Hardware Center, which opened in February 2019 to advance research in AI chips and hardware systems.
The AI Hardware Centre works with more than 16 organisational partners to accelerate AI performance scaling and build hardware systems from the ground up – even to the cloud.
More details on the tool can be found in this blog.