NVIDIA Workshop
Power Networks for AI Clouds
An advanced technical session showcasing the technologies needed to power networks for AI Factories and Clouds.
Offer
The network is the backbone of the modern data center, and ultimately responsible for ushering in the era of AI. The network’s ability to scale and handle an increasing number of nodes is essential for training large AI models and handling high data volumes.
Join NVIDIA for a deep dive on the role of networking technologies - NVIDIA Quantum InfiniBand and NVIDIA Spectrum-X Ethernet - in shaping the future of data centers and facilitating AI.
- How advances in AI model architecture impact data center infrastructure requirements and design
- Why AI training parallelism requires particular collective operations with different traffic patterns and network sensitivity
- How techniques such as GPU Direct RDMA, adaptive routing, congestion control, and link rerouting are crucial for high bandwidth and performance isolation
- And watch a demo in NVIDIA Air to see how Spectrum-X networks are configured and provisioned to support multi-tenant AI Clouds
Additional Resources
- NVIDIA Networking for AI White Paper
- NVIDIA Spectrum-X Technical White Paper
- NVIDIA BlueField-3 SuperNIC White Paper
Prerequisites
Attendees should have a technical background in ethernet networking technology, with an understanding of AI and how it is impacted by the network. Attendees should also have basic knowledge of generative AI, large language models, RDMA/RoCE, and AI Collective operations (NCCL).
It is highly recommended that attendees watch the on-demand Networking for the Era of AI webinar, before attending this workshop.
Register Now
THANK YOU FOR REGISTERING FOR THE WEBINAR
You will receive an email with instructions on how to join the webinar shortly.
Main Content
maincontent goes here
Content
Content goes here
Content
content goes here
Content
Content goes here
Content
DGX Station Whitepaper
Dive deeper into the DGX Station and learn more about the architecture, NVLink, frameworks, tools and more.
Dive deeper into the DGX Station and learn more about the architecture, NVLink, frameworks, tools and more.
DGX Station Whitepaper
Dive deeper into the DGX Station and learn more about the architecture, NVLink, frameworks, tools and more.
Dive deeper into the DGX Station and learn more about the architecture, NVLink, frameworks, tools and more.
DGX Station Whitepaper
Dive deeper into the DGX Station and learn more about the architecture, NVLink, frameworks, tools and more.
Dive deeper into the DGX Station and learn more about the architecture, NVLink, frameworks, tools and more.
Content
Content goes here
Speakers
Gadi Godanyan
Senior AI Networking Lead, EMEA, NVIDIAGadi Godanyan is leading AI Networking for NVIDIA in Europe, Middle East and Africa. In his capacity as Senior AI Networking Lead, Gadi is responsible for AI Networking sales—from defining sales and GTM strategies to working closely with key accounts to drive technology adoption and sales. For the past decade, Gadi has worked at NVIDIA in various roles, most recently as Senior Ethernet Switch Manager, focusing on Ethernet Switch sales.
Jeff Tantsura
Distinguished Architect, NVIDIAJeff Tantsura has been in the networking space for 25+ years and has authored/contributed to many RFC's and patents, and worked in – hyperscale, SP and vendor environments. He is co-chair of IETF Routing Working Group, chartered to work on New Network Architectures and Technologies, and Next Gen Routing Protocols as well as co-chair of RIFT (Routing in Fat Trees) Working Group chartered to work on the new routing protocol that specifically addresses Fat Tree topologies typically seen in the Data Center environment.
Prior to joining NVIDIA, Jeff worked as Sr. Principal Network Architect Azure Networking at Microsoft, working on the Next Gen DC architecture and AI/ML networking.
Prior to joining NVIDIA, Jeff worked as Sr. Principal Network Architect Azure Networking at Microsoft, working on the Next Gen DC architecture and AI/ML networking.
Igor Miroshnichenko
Senior Technical Marketing Engineer, Ethernet Networking Platforms , NVIDIAIgor Miroshnichenko is a senior technical marketing engineer for Ethernet networking platforms at NVIDIA, focusing on modern data center networking. He is a subject matter expert on switch networking, monitoring, and telemetry tools. He has extensive experience designing, deploying, and managing real-world, cloud-scale data center networks for AI/ML, cloud, and storage infrastructures.
Main CTA for lightbox form use class="lightbox" CTA
Content Title
Content here
Register
Webinar: Description here
Date & Time: Wednesday, April 22, 2018