Building PyTorch-native support for the IBM Spyre Accelerator
Technical note
Viji Srinivasan, Raghu Ganti, Mudhakar Srivatsa, Elpida Tzortzatos, Tabari Alexander, and Vaidyanathan Srinivasan
What’s Next in Semiconductors
is powering full-stack AI
The way the world computes is shifting. Over the last decade, demand for AI applications has grown exponentially. Pipelines are getting more complex, and the world needs new hardware to train and serve AI at the scale enterprise demands. At IBM Research, we’re building the hardware and the tools to serve cutting-edge AI. We’re pushing the limits of semiconductor design, packaging, and connectivity, with innovations like sub-2 nm node chips, co-packaged optics, the AIU family of chips, and 3D nanosheet transistor technology. We’re building tomorrow’s AI supercomputers from the ground up.
Our work
IBM demonstrates High NA EUV process capability on track for insertion below 2 nm nodes at SPIE 2026
Technical noteLuciana MeliLifting the cover on the IBM Spyre Accelerator
Technical noteJeffrey BurnsA glimpse at computing’s quantum-centric future
ExplainerRyan Mandelbaum and Jerry ChowFrom atoms to chips: Thermonat models heat with unprecedented accuracy
NewsPeter HessA new advance in a two-century pursuit in physics
Technical noteOki Gunawan, Doug Bishop, and Bonfilio Nainggolan- See more of our work on Semiconductors
AI Hardware Center
The IBM Research AI Hardware Center is a global research collaboration hub dedicated to creating the next generation of systems and chips for AI workloads, as well as expanding joint research efforts across technology, architecture, and algorithms.
Explore the center