NVIDIA Event at KAUST
KAUST will host NVIDIA Days from September 15 to 17, 2026, a three-day program supporting the community's ongoing work in AI and High Performance Computing (HPC).
The program has three components:
- Sessions: Morning and afternoon sessions presenting current tools, methods, and practices from NVIDIA specialists. Open to all attendees; registration required.
- Clinics: 30-minute one-on-one consultations with an NVIDIA expert for guidance on a specific project or question. Prerequisites shared upon registration.
- DLI Course: A structured, hands-on training course delivered by NVIDIA instructors. Enrollment is limited; prerequisites shared with confirmed participants ahead of time.
| Register |
Agenda
Day 1: Talks
Date: September 15
Venue: Building 3, Auditorium 0215 (between buildings 2 & 3)
AI Session
08:45–09:30 — Nemotron Agentic AI Systems
AI is evolving from conversational assistants into agents that can reason, use tools, and execute complex workflows. This shift creates new requirements for the models powering these systems, including strong reasoning, reliable tool use, specialization, and high efficiency. In this session, we'll explore how NVIDIA Nemotron supports agentic AI through open reasoning, multimodal, retrieval, speech, and safety models. We'll also discuss the techniques behind Nemotron's efficiency and agentic capabilities, along with their relevance to use cases such as deep research, RAG, and enterprise agent workflows.
09:40–10:25 — NVIDIA Warp and PhysicsNeMo: Accelerated Frameworks for Differentiable Physics and Physics AI
Physics developers often face a trade-off between the performance of CUDA and the productivity of Python. NVIDIA Warp bridges this gap by compiling Python functions to native x86, ARM, or CUDA code while supporting automatic differentiation.
Building on this foundation, NVIDIA PhysicsNeMo provides an open-source framework for developing physics AI models at scale. It combines simulation, observed data, and physics-driven constraints to enable fast, physically consistent predictions. In this session, we'll provide an overview of both frameworks, covering differentiable physics with Warp and the PhysicsNeMo development stack, including neural operators, graph networks, transformers, generative models, PDE-constrained training, distributed scaling, and guardrails for physical validity.
10:35–11:25 — Physical AI: Digital Twins, Simulation, and the Path from Pixels to Robots
Large language models taught machines to reason about text. The next frontier is machines that perceive, reason, and act under the constraints of physics; robots that manipulate, inspect, and navigate real environments.
11:35–12:05 — Weather and Climate (TBC)
Speaker TBC
Weather prediction and modeling (Earth-2).
HPC Session
These three talks will broadly cover NVIDIA’s technologies in technical computing. The first talk will cover HPC & AI hardware and GPU concepts, the second talk will provide an overview of HPC software including CUDA-X and Slurm/Slinky, and the third talk will cover NVIDIA’s approach to accelerated quantum supercomputing.
13:30–14:15 — GPU for Developers
GPU computing fundamentals; architectural foundations of modern GPU (GH200, GB and beyond); hardware features (tensor cores, NVLink-C2C, EGM, coherent memory); ARM CPU benefits; low-precision computing and Transformer Engine (what it is and use cases).
14:25–15:50 — Developer Ecosystem
Libraries; zero-code path and abstractions; StdPar; CuPyNumeric; Speed of Light quest; Roofline with Nsight; CUTLASS; cuTile; nvCOMP; communication libraries including NVSHMEM and NCCL.
16:00–16:45 — Quantum Computing Ecosystem on GPUs
Quantum computing ecosystem on GPUs, including GH200 and beyond.
Day 2: Clinics
Date: September 16
Venue: TBD
09:00–12:00 - Clinic sessions
30-minute 1:1 consultations with NVIDIA experts. Slots by registration; prerequisites shared upon sign-up.
Register
Day 3: DLI Course
Date: September 16
Venue: Building 1 (desert side), Level 2, MPR Room
09:00–17:00 — Building Agentic AI Applications with LLMs
Hands-on, onsite training. Please click here for the course details.
Prerequisites and access details shared with confirmed participants.
Speakers
Meshari Alaeena
Solutions Architect, NVIDIA
Meshari is a Solutions Architect at NVIDIA with experience in deep learning, agentic AI solutions, and data processing pipelines for structured and unstructured data. His work focuses on designing scalable AI architectures, building reliable data workflows, and developing practical AI solutions for complex, real-world use cases.
Rema Algunaibet
Solutions Architect, Physical AI, NVIDIA
Rema Algunaibet is a Physical AI Solutions Architect at NVIDIA, holding an MSc in Data Science and Advanced Computing from the University of Reading and currently completing an MBA at Imperial College Business School. She works with organizations and partners across EMEA to accelerate development of digital twins, robotics, and physical AI workloads. Her work spans OpenUSD and Omniverse-based simulation, robot policy training and sim-to-real transfer in Isaac Sim, and agentic AI for operations, with recent projects across energy, sustainability, manufacturing, and giga-project infrastructure. She is an ambassador for Saudi Arabia in the Women in AI program by IPSoft in the US, a Chartered Statistician (CStat) and Accredited Advanced Data Science Professional (ADSP) with the Royal Statistical Society in the UK, and a published researcher in Procedia Computer Science.
Shuaib Akhtar
Senior Solutions Architect, NVIDIA
Shuaib Akhtar is a Senior Solutions Architect at NVIDIA. He has spent 15 years in HPC working at Bull, IBM, and Platform Computing, and prior to that worked at AWE, a UK Government research lab. He is experienced in large-scale HPC heterogeneous system architectures and HPC software and datacentre integrations, from initial design through deployment and system upgrades.
DLI Instructor
Zahra Al-Kaf
FSI Solution Architect -GenAI
Zahra is a Senior Solutions Architect at NVIDIA, where she focuses on bringing generative AI to life for the Financial Services Industry. With over a decade of experience at leading technology companies — including Groq, Google, Oracle, and Cisco — she brings deep technical expertise and a sharp understanding of what it takes to move AI from concept to production in one of the world's most demanding industries.
At NVIDIA, she works with financial institutions to design and deploy generative AI use cases that drive real business impact, helping FSI leaders move beyond the pilot stage and into meaningful, scalable AI adoption
© 2026 King Abdullah University of Science and Technology. All rights reserved.