Nvidia’s annual AI developers conference (GTC) used to be a relatively modest affair, drawing about 9,000 people in its last year before the Covid outbreak. But the event now unofficially dubbed “AI Woodstock” is expected to bring more than 25,000 in-person attendees!
Nvidia’s Blackwell AI chips, the main showcase of last year’s GTC (GPU Technology Conference), have only recently started shipping in high volume following delays related to the mass production of their complicated design. Blackwell is expected to be the main anchor of Nvidia’s AI business through next year. Analysts expect Nvidia Chief Executive Jensen Huang to showcase a revved-up version of that family called Blackwell Ultra at his keynote address on Tuesday.
March 18th Update: The next Blackwell Ultra NVL72 chips, which have one-and-a-half times more memory and two times more bandwidth, will be used to accelerate building AI agents, physical AI, and reasoning models, Huang said. Blackwell Ultra will be available in the second half of this year. The Rubin AI chip, is expected to launch in late 2026. Rubin Ultra will take the stage in 2027.
Nvidia watchers are especially eager to hear more about the next generation of AI chips called Rubin, which Nvidia has only teased at in prior events. Ross Seymore of Deutsche Bank expects the Rubin family to show “very impressive performance improvements” over Blackwell. Atif Malik of Citigroup notes that Blackwell provided 30 times faster performance than the company’s previous generation on AI inferencing, which is when trained AI models generate output. “We don’t rule out Rubin seeing similar improvement,” Malik wrote in a note to clients this month.
Rubin products aren’t expected to start shipping until next year. But much is already expected of the lineup; analysts forecast Nvidia’s data-center business will hit about $237 billion in revenue for the fiscal year ending in January of 2027, more than double its current size. The same segment is expected to eclipse $300 billion in annual revenue two years later, according to consensus estimates from Visible Alpha. That would imply an average annual growth rate of 30% over the next four years, for a business that has already exploded more than sevenfold over the last two.
Nvidia has also been haunted by worries about competition with in-house chips designed by its biggest customers like Amazon and Google. Another concern has been the efficiency breakthroughs claimed by Chinese AI startup DeepSeek, which would seemingly lessen the need for the types of AI chip clusters that Nvidia sells for top dollar.
…………………………………………………………………………………………………………………………………………………………………………………………………………………….
Telecom Sessions of Interest:
Wednesday Mar 19 | 2:00 PM – 2:40 PM
Delivering Real Business Outcomes With AI in Telecom [S73438]
Chris Penrose | VP & Head of Business Development Telco | NVIDIA
Andy Markus | SVP and Chief Data and Artificial Intelligence Officer | AT&T
Kaniz Mahdi | Director Technology, AWS Industries | Amazon Web Services
Anil Kumar | VP, Head of AI Center | Verizon
Hans Bendik Jahren | VP Network and Infrastructure | Telenor
In this session, executives from three leading telcos will share their unique journeys of embedding AI into their organizations. They’ll discuss how AI is driving measurable value across critical areas such as network optimization, customer experience, operational efficiency, and revenue growth. Gain insights into the challenges and lessons learned, key strategies for successful AI implementation, and the transformative potential of AI in addressing evolving industry demands.
Thursday Mar 20 | 11:00 AM – 11:40 AM PDT
AI-RAN in Action [S72987]
Soma Velayutham | VP, AI and Telecoms | NVIDIA
Aji Ed | VP & Head of CloudRAN, Mobile Networks | Nokia
Ryuji Wakikawa | VP, Research Institute of Advanced Technology | Softbank
Freddie Södergren | VP and Head of Technology & Strategy | Ericsson
Karri Kuoppamaki | SVP, Advanced and Emerging Technologies | T-Mobile
Two AI trends are driving the need for an AI-enabled infrastructure at the 5G/6G radio access network (RAN) edge. Inferencing for generative AI and AI agents requires AI compute infrastructure to be distributed from edge to central clouds. At the same time, it’s clear that the RAN is evolving to an AI-native infrastructure. The AI-RAN brings these trends to fruition, providing an accelerated computing infrastructure that accelerates both radio signal processing and AI workloads. In the past six months, both Softbank in Japan and T-Mobile in the United States have taken leading roles toward transforming their networks to AI-RAN. Our panel will explore the motivations and practicalities of delivering a commercial grade AI-RAN network to transform commercial return on investment and increase spectral efficiency, network capacity, and network utilization.
Thursday Mar 20 | 9:00 AM – 9:40 AM PDTHow Indonesia Delivered a Telco-led Sovereign AI Platform for 270M Users [S73440]
Anissh Pandey | Sr. Director NCP Asia Pacific | NVIDIA
Lilach Ilan | Global Head of Business Development -Telco Operations | NVIDIA
Munjal Shah | Co-Founder and CEO | Hippocratic AI
Vikram Sinha | CEO | Indosat Ooredoo Hutchison
Senthil Ramani | Global Lead – Data & AI | Accenture
Sovereign AI enables nations to create their own AI and participate in the new global economy. Nations are partnering up with telcos to enable the foundational AI platform and ecosystem required for their universities, startups, enterprises, and government agencies to create AI. Indosat, a major telecoms provider with over 100 million customers, working with Lintasarta, GoTo, Accenture, and Hippocratic, has deployed a Sovereign AI Factory and a collection of LLMs for Indonesia leveraging the NVIDIA Cloud Partner program. We’ll explore how Indosat set up the sovereign cloud and launched its own foundational large language model, Sahabat.ai, for the 277 million native Bahasa speakers with NVIDIA NEMO and NIMs. We’ll also discuss how partners like Accenture and Hippocratic AI are accelerating AI use-case deployment in Indonesia for critical sectors including banking and healthcare using NVIDIA AI Platform.
Thursday Mar 20 | 3:00 PM – 3:40 PM PDT
Driving 6G Development With Advanced Simulation Tools [S72994]
CC Chong | Senior Director, Aerial Product Management | NVIDIA
Balaji Raghothaman | Chief Technologist, 6G | Keysight Technologies
Arien Sligar | Senior Principal Product Specialist | Ansys, Inc.
Tommaso Melodia | William Lincoln Smith Professor | Northeastern University
The NVIDIA Aerial Omniverse Digital Twin (AODT) is a platform that leverages the power of NVIDIA accelerated computing, NVIDIA Omniverse, and NVIDIA Aerial to provide a comprehensive, scalable, and flexible solution for 6G research and development. The platform enables researchers and developers to customize, program, and test 6G networks in near-real time, with AI/ML in the loop, and to simulate and optimize the network performance and quality of service based on site-specific data and system-level simulation. The platform has been adopted by Keysight and Ansys with advanced simulation tools. One is leveraging our state-of-the-art 3D electromagnetic ray tracing model and radio access network (RAN) modules for advanced 5G/6G end-to-end simulations. The other has integrated their PerceiveEM Solver into the AODT to leverage NVIDIA AI/ML with full-stack RAN in the loop to accelerate and simplify 6G research and development. The NVIDIA Aerial Omniverse Digital Twin has also empowered academia to unleash the potential of cloud and AI-native networks for 6G system research, design, and development.
Thursday Mar 20 | 2:00 PM – 2:40 PM PDT
Defining AI-Native RAN for 6G [S72985]
Chris Dick | Senior Distinguished Engineer, 5G/6G | NVIDIA
Ardavan Tehrani | Director Engineering | Samsung Research
Moe Win | Professor | MIT
Jim Shea | CEO | DeepSig, Inc.
Kai Mao | Head of AI-RAN Strategy for Fujitsu Network Business | Fujitsu
The telecoms industry is working to integrate AI into the design, operation, and optimization of 6G networks to enable unprecedented levels of automation, efficiency, and adaptability. The AI-native 6G radio access network (RAN) embodies this expectation and makes AI a foundational element of the 6G stack. In this session, leading telecoms industry stakeholders will outline how they’re advancing 6G research, with AI integration, to revolutionize connectivity, performance, and societal transformation. They’ll explore how the AI-native 6G RAN can deliver extreme levels of radio access performance, integrating technologies like deep learning and edge computing, promoting sustainability, enhancing public safety, and driving economic growth through advanced applications like smart cities, digital twins, and immersive communication.
Thursday Mar 20 | 4:00 PM – 4:40 PM PDT
Pushing Spectral Efficiency Limits on CUDA-accelerated 5G/6G RAN [S72990]
Vikrama Ditya | Senior Director of Software, Aerial | NVIDIA
Tommaso Balercia | Principal Engineer, Aerial | NVIDIA
Yuan Gao | Senior Software Engineer, 5G – Algorithms | NVIDIA
The demand for mobile broadband continues to grow, and correspondingly the radio access networks (RAN) delivering it are asked to support an extraordinary rate of innovation. RAN-based on software-defined platforms are of key importance in enabling such a rate of innovation sustainably. Among all software-defined platforms, the one based on the GPU allows for the introduction of complex algorithms to pursue new levels of spectral efficiency. This is both to the ease with which the GPU can be programmed and their computational efficiency. Moreover, GPU-based 5G and 6G RAN stacks can run coherently with the code of a new class of simulation tools: digital twins. The efficiency with which the GPU can run both physics simulation chains as well as the 5G/6G RAN stack enable simulations to reach new scales and levels of accuracy. Such simulation environments, in turn, enable the creation and the detailed characterization of new algorithms. In this talk, we will demonstrate how a series of key algorithms that make use of the GPU computational power allow us to push the boundary of system spectral efficiency. The characterization is performed in a digital twin environment, where the radio environment is closely resembling the field.
Thursday Mar 20 | 4:00 PM – 4:40 PM PDT
Enable AI-Native Networking for Telcos with Kubernetes [S72993]
Erwan Gallen | Senior Principal Product Manager | Red Hat
Elad Blatt | Global Head Business Development Telco Networking | NVIDIA
Ahmed Guetari | VP and GM, Service Providers | F5
As telcos transform their networks for the age of AI, utilizing the Bluefield 3 (BF3) DPU engines has been challenging for developers and clients alike. Yet, BF3 is a critical part in deploying and securing an accelerated compute cluster to enable this networking-for-AI infrastructure. The DOCA Platform Framework (DPF) simplifies this task, providing a framework for LCM (life cycle management) and provisioning of both the BF3, as a platform, and the services running on it as K8 containers. DPF is deployed via two network operators in the K8 environment that allow you to deploy and service chain NVIDIA and third-party services. With this, for the first time, independent software vendors (ISVs), operating systems (OS) vendors, and developers can deploy and orchestrate services with ease, as well as onboarding these tools to their environment. In this session, OS and ISV partner companies who’ve adopted the DPF will share their experience, what they are able to achieve, and what comes next.
Monday Mar 17 | 3:00 PM – 4:45 PM PDT
Automate 5G Network Configurations With NVIDIA AI LLM Agents and Kinetica Accelerated Database [DLIT72350]
Maria Amparo Canaveras Galdon | Senior Solutions Architect Generative AI | NVIDIA
Swastika Dutta | Solutions Architect Generative AI | NVIDIA
Shibani Likhite | Solutions Architect | NVIDIA
Nick Reamaroon | Solutions Architect | NVIDIA
Anna Daccache | Solutions Architect Generative AI | NVIDIA
Learn how to create AI agents using LangGraph and NVIDIA NIM to automate 5G network configurations. You’ll deploy LLM agents to monitor real-time network quality of service (QoS) and dynamically respond to congestion by creating new network slices. LLM agents will process logs to detect when QoS falls below a threshold, then automatically trigger a new slice for the affected user equipment. Using graph-based models, the agents understand the network configuration, identifying impacted elements. This ensures efficient, AI-driven adjustments that consider the overall network architecture.
We’ll use the Open Air Interface 5G lab to simulate the 5G network, demonstrating how AI can be integrated into real-world telecom environments. You’ll also gain practical knowledge on using Python with LangGraph and NVIDIA AI endpoints to develop and deploy LLM agents that automate complex network tasks.
Prerequisite: Python programming.
………………………………………………………………………………………………………………………………………………………………………………………………………..
References:
