How High Performance Data Platforms Unlock HPC and AI Applications: Lustre, Sycomp, VAST & Weka in Google Cloud

About the Session

The demands of modern HPC and AI workloads require specialized data platforms capable of delivering ultra-low latency, massive bandwidth, and petabyte-scale capacity–-an additional features to enable collection, management, curation, replication and migration of data for a wide range of user and workload needs.

Join the Google Cloud Advanced Computing Community (GCACC) for a showcase highlighting four premier high-performance storage solutions, each engineered to eliminate I/O bottlenecks and accelerate scientific discovery. This online event will feature dedicated presentations from the experts behind four key platforms, focusing on the unique architectural strengths, integration points, and performance profiles that make them leaders in the HPC/AI space:

• Google Cloud Managed Lustre (built on DDN’s Exascaler)

• Sycomp Intelligent Data Storage Platform (built on IBM Storage Scale)

• VAST Data Platform

• WEKA NeuralMesh (formerly Data Platform)

Each vendor will present an overview into their solution's features, capabilities, and superior performance for demanding workloads. Attendees can submit questions that each vendor and/or Google storage experts will answer and post after the presentation to maximize the number of questions addressed and knowledge gained by attendees.

What You Will Learn:

• Solution Strengths & Features: Understand the core technologies and features of each platform, including POSIX compliance, multi-protocol access, and robust data management capabilities.

• Elite Performance: Discover the distinct I/O performance advantages and massive scalability each solution delivers, from ultra-low latency metadata operations for small-file workloads to petabyte-scale aggregate throughput for large-scale data-intensive tasks.

• Workload Matching: Learn the definitive criteria for selecting the optimal storage solution for your specific scientific and technical workloads. Whether you require high-throughput scratch storage for tightly coupled HPC, high-speed model training for Distributed AI/ML, or large-volume persistent data lakes, you will learn how to match the right solution to your I/O demands in Google Cloud.

Attendees will gain a clear understanding of the diverse, powerful storage ecosystem available and how to select the best platform to accelerate their computational workloads in Google Cloud and beyond.