Events · DTELM 2025

KAUST Workshop on Distributed Training in the Era of Large Models

24–26 November 2025 · KAUST, Thuwal, Saudi Arabia
Building 4 & 5, Level 0, Auditorium 0215
Original KAUST page (may be archived)

A three-day in-person workshop on scaling learning algorithms to the computational and data demands of modern AI models: distributed optimization, communication-efficient algorithms, model-parallel architectures, and hardware–software co-design. Invited talks, lightning talks, a poster session, and informal discussion.

Organizers

• Peter Richtárik (KAUST)
• Marco Canini (KAUST)
• Laurent Condat (KAUST)
• Kaja Gruntkowska (KAUST)
• Egor Shulgin (KAUST)

Speakers

• Thalaiyasingam Ajanthan (Pluralis Research)
  Communication Efficient Model Parallel Training of Language Models
• Marco Canini (KAUST)
• Volkan Cevher (EPFL)
  Training neural networks at any scale
• Heming Cui (University of Hong Kong)
  Building Multi-dimensional Parallel Training Systems for Large AI Models
• Hongchang Gao (Temple University)
  On Provable Benefits of Muon in Federated Learning
• Kaja Gruntkowska (KAUST)
• Frederik Kunstner (Inria)
  The Benefits of Adam and Sign-like Methods in Training Large Language Models
• Zhouchen Lin (Peking University)
  Conda: Column-Normalized Adam for Training Large Language Models Faster
• Antonio Orvieto (MPI-IS, Tübingen)
  Training LLMs: Do We Understand Our Optimizers?
• Edouard Oyallon (CNRS)
  Hiding Communication Cost via Overlapping Methods
• Youngsuk Park (Amazon AWS AI)
  Toward Compute- and Memory-Efficient LLM Training: Optimizers and Low-Precision Recipes
• Peter Richtárik (KAUST)
• Frank Schneider (University of Tübingen)
  Accelerating Neural Network Training: An Analysis of the AlgoPerf Competition
• Egor Shulgin (KAUST)
• Sebastian Stich (CISPA)
• Kun Yuan (Peking University)
  A Memory Efficient Randomized Subspace Optimization Method for Training Large Language Models

Program

24 November 2025

• 09:00–09:35 Opening
• 09:35–10:05 Kun Yuan
• 10:10–10:40 Youngsuk Park
• 10:45–11:15 Coffee & networking
• 11:15–11:45 Peter Richtárik
• 11:50–12:20 Hongchang Gao
• 12:25–13:45 Lunch
• 13:45–14:15 Zhouchen Lin
• 14:20–14:50 Edouard Oyallon
• 14:50–15:30 Lightning talks
• 15:30–17:00 Poster session
• 18:00 Gala dinner (by invitation)

25 November 2025

• 09:00–09:30 Volkan Cevher
• 09:35–10:05 Egor Shulgin
• 10:10–10:40 Heming Cui
• 10:45–11:15 Coffee & networking
• 11:15–11:45 Antonio Orvieto
• 11:50–12:25 Group photo
• 12:25–13:45 Lunch
• 13:45–14:15 Informal research discussions
• 15:00 Desert tour (by invitation)

26 November 2025

• 09:00–09:30 Thalaiyasingam Ajanthan
• 09:35–10:05 Frank Schneider
• 10:10–10:40 Marco Canini
• 10:45–11:15 Coffee & networking
• 11:15–11:45 Frederik Kunstner
• 11:50–12:20 Sebastian Stich
• 12:25–13:45 Lunch
• 13:45–14:15 Kaja Gruntkowska
• 14:20–14:50 Closing remarks
• 16:00 Al Balad tour (by invitation)