Program

• 9:00-9:05 - Introduction from CADL organizers

• 9:05-9:35 - "Supercomputing in the age of AI reasoning models", Frédéric Parienté - NVIDIA (Invited speaker)

• 9:35-10:00 - "Direct Feedback Alignment for Recurrent Neural Networks", Andrea Cossu - University of Pisa

  • 10:00-10:25 - "Optimizing edge AI models on HPC systems with the edge in the loop", Marcel Aach - Juelich Supercomputing Centre

• 10:25-11:00 - Coffee break

• 11:00-11:30 - "FP8 large-scale training: achieving speedup without compromising performance", Sergio Perez - NVIDIA (Invited speaker)

  • 11:30-11:55 - "Evaluation of Distributed Asynchronous Checkpointing in High-Performance Computing", Riccardo Scheda - CINECA

• 11:55 - 12:20 - "Assessing Tenstorrent’s RISC-V MatMul Acceleration Capabilities", Hiari Pizzini Cavagna - University of Bologna

• 12:20-12:45 - "Automatically parallelizing batch inference on deep neural networks using Fiats and Fortran 2023 'doconcurrent' ", Damian Rouson - Lawrence Berkeley National Laboratory 

• 12:45 - 13:00 - Closing remarks

Friday, June 13, 2025 9:00 AM to 1:00 PM (Hall X12 - 1st floor)