GTC 2026

Microway team note- we predict this session may feature information on next-gen NVIDIA DGX Systems with NVIDIA Vera Rubin architecture. Attend and find out! The next generation of AI models is driving unprecedented compute consumption, exposing the limitations of traditional system architectures. Enterprises are now evolving data center designs to support new rack-scale architectures for modern inference, as well as large-scale training optimized for these rapidly escalating demands. Join this session to learn how NVIDIA is systemizing the new NVIDIA Vera Rubin architecture to help customers rapidly deploy gigascale AI factories.

The CUDA platform is the foundation of the GPU computing ecosystem. Every application and framework that uses the GPU does so through CUDA’s libraries, compilers, runtimes and language—which means CUDA is growing as fast as its ecosystem is evolving. Presented by one of the architects of CUDA, at this engineering-focused talk you will learn about all that’s new and what’s coming next for both CUDA and GPU computing as a whole.
Microway team note- this is where you’ll learn about new features for CUDA in 2026 and how CUDA will support new architectures.

Be among the first to hear how software-defined intelligence is aiding enterprises to scale AI operations. Hear directly from an AI innovator as they reveal key learnings from deploying NVIDIA DGX, sharing firsthand insights into architecting and deploying high-performance AI infrastructure and orchestrating complex inference and training workflows. In this session, we’ll also discuss the critical role of software for NVIDIA Rubin and how to simplify complex liquid-cooled deployments at scale, boost site reliability engineering (SRE) productivity, and enable enhanced AI-powered anomaly detection and remediation across the entire AI factory. Microway team note- we expect much information about NVIDIA Mission Control in practice from this session.

AI infrastructure must deliver predictable, high‑throughput pipelines for large‑scale training and inference while enforcing strict security, data sovereignty, and multi‑tenant isolation across cloud and on‑premises environments. These demands can overwhelm traditional host‑based networking and security, reducing utilization and fragmenting policies. This session will bring together NVIDIA technologists building NVIDIA’s own DGX Cloud platform powered by NVIDIA Networking and show how they are extending the learnings to sovereign DGX AI factory solutions. Understand design patterns such as zero‑trust for hardened multi‑tenant clusters, accelerated Kubernetes networking, intelligent data platforms, and security services that enable consistent, high‑performance AI factories everywhere.

The path to a fault-tolerant quantum computer (FTQC) is not solely a physics challenge; it is an AI and supercomputing challenge. This talk outlines the singular mission required to deliver scientifically useful accelerated quantum supercomputers by 2028: the complete convergence of AI with quantum hardware, systems, and algorithms. We will also present the blueprint for a collaborative and open quantum ecosystem. By leveraging open standards in software, CUDA-Q, interconnects, and NVQLink, we are creating a unified, hybrid infrastructure where national labs, startups, and industry partners can collaborate.

As at-scale AI workloads for training and inference make use of gigawatt-scale multi-GPU systems, massive data transfer, and memory access capabilities of Vera Rubin NVL72 systems in AI factories, we will show how these large systems get architected and deployed in the data center, leading to a robust reference architecture embraced globally for its speed of deployment and performance. This session will showcase internal deployments and the tools and software stack developed and leveraged on these platforms. You’ll also have the opportunity to have deep dive views on how multi-node NVLink, system interconnects, and storage, as well as the latest liquid-cooled designs, are shaping the future of AI infrastructure.

The fuel for AI factories is data. And enterprise data management systems are evolving to meet the large-scale data needs of AI factories. Join us to learn how enterprise storage systems are transforming into knowledge and intelligence management systems. We will discuss key technology innovations from NVIDIA that are enabling and accelerating this transition.

As organizations scale toward population-level AI adoption, running inference reliably, efficiently, and cost-effectively becomes essential. This session covers best practices for scaling inference across platform, model, and application layers, using real-world insights and NVIDIA technologies like DGXC-Lepton and NVCF. We’ll address challenges in GPU capacity, autoscaling, model optimization with NVIDIA S/W stack, concurrency, architecture, and observability. Learn how to use proven, field-tested strategies for scaling current AI workloads or launching new ones at scale.

This session delivers a technical state of the union on GPU-accelerated data processing across SQL/DataFrames, vector search, ML, and decision optimization. Learn how GPU-native engines enable interactive analytics on massive lakehouse-scale datasets, real-time semantic and vector search over billions of embeddings, and makes the hardest ML and decision science workloads tractable, cost-efficient, and energy-efficient. The talk highlights the implications for high-impact scientific and enterprise computing, then looks ahead to what’s in flight for 2026 and beyond, outlining concrete architectural patterns and practical guidance for building the next generation of GPU-accelerated data platforms and using them in your day-to-day work.

As AI transforms every layer of modern computing, security must evolve in parallel—shifting from a separate function to an embedded capability woven into every stage of innovation. Today’s CISOs and security leaders are moving beyond reactive defense, redefining their roles as proactive business enablers who accelerate safe adoption of AI, cloud, and emerging technologies. This talk explores how that shift is reshaping security architectures, operating models, and organizational influence. Finally, it highlights why meaningful collaboration across industry, academia, and government is no longer optional. Addressing global cyber threats at AI scale demands shared intelligence, coordinated standards, and partnerships that blend cutting-edge research with real-world deployment. Together, these forces define the future of secure innovation.

This session guides you through the end-to-end deployment of generative image models on NVIDIA RTX PRO workstations by leveraging ONNX Runtime and TensorRT-RTX. We’ll demonstrate how to ship a multi-stage model pipeline across a wide set of hardware without sacrificing performance. This workflow will be powered by classic graphics APIs, ONNX Runtime, and the TensorRT-RTX Execution Provider. You’ll leave with the skills to convert raw models into high-performance, offline-capable applications that run seamlessly across diverse hardware architectures.

Join NVIDIA experts to explore the field of digital biology. Discuss the latest advancements in accelerated computing and foundation models for drug discovery and genomics — from bioinformatics tooling to protein structure prediction, molecular dynamics to molecular generation, and much more. Learn how NVIDIA BioNeMo, Parabricks, and RAPIDS-singlecell, together with NVIDIA NIM, empower developers, researchers, and enterprises to quickly create generative AI solutions across chemistry, biology, and genomics. Engage directly with our product managers, developer relations, and solution architects to address your challenges and answer your questions.