On Demand: SC23

SC23 | Denver, CO

On Demand: Step onto the SC23 show floor and into the DDN booth for many exciting talks!

On-Demand

Optimizing LLMs with Shared Parallel Storage

Louis Douriez | DDN

Data-Intensive Research at Purdue University

Preston Smith | Purdue

Maximize Your Deep Learning Investment Using DDN (Part 1)

David Hall | Lambda

Lambda has built a few of the largest and highest-performance GPU clusters for deep learning training. This session discuss some of the key design principles to minimize all potential bottlenecks and how we have partnered with DDN specifically to address I/O performance.

Maximize Your Deep Learning Investment Using DDN (Part 2)

David Hall | Lambda

An AI Journey with DDN & Intel Gaudi2

Binay Ackalloor | Habana Labs

Brining AI and HPC together can be complex and expensive, learn how DDN partners with Intel Gaudi2 to offer high-performance AI solutions.

DDN DataFlow: Data Management at Scale

Mike Oakes | Atempo

Learn how to manage data at scale utilizing DDN data flow data management with analytics.

Providing Best in Class Storage Speed Across the AI Lifecycle

Robert Mango | Run:ai

See how you can leverage the power of Run:ai’s scheduling and orchestration capabilities along with DDN’s best in class storage speeds to facilitate the full AI lifecycle. Topics covered will include data loading as it pertains to data pre-processing, model training and inference serving.

Accelerating Storage10 w/NIVIDIA Networking

Rob Davis | NVIDIA

The AI and HPC data surge has led to advanced storage technologies like NVMe SSDs and distributed file systems. However, connecting them across a network can easily overwhelm it. NVIDIA InfiniBand and High Performance Ethernet networking solutions provide cost effective solutions to this dilemma.

DGX SuperPOD: Turnkey AI Datacenter Solution

Mike Sabotta | NVIDIA

Employing On-Perm AI for Enterprise Storage (AMAX, DDN, NVIDIA – LLMs)

Andrew Lekashman | AMAX

In the enterprise sphere, advanced server solutions for Large Language Models (LLMs), with an emphasis on Retrieval-Augmented Generation (RAG), provide businesses with powerful tools to pull and combine information from their own large datasets.

DDN Overview

Jenn Autry | DDN

Ranked #1 for High-Performance Storage for many years, our advanced data storage solutions ensure maximum GPU utilization for the best performance and return on investment possible. If you need a fast, secure, scalable data environment, there is no comparison to DDN.

EXAScaler/Lustre at Scale for AI at NVIDIA

Adam Deconinck | NVIDIA

Two of our DGX SuperPODs, Selene and Eos, run AI at scale with EXAScaler/Lustre. We’ll review how these systems leverage some of the latest features of Lustre to provide performance, observability, and security when running deep learning workloads.