AI factory and accelerated computing demonstrations and expert sessions highlight how customers can dynamically manage and scale workloads
Penguin Solutions, an SGH brand that designs, builds, deploys, and manages AI and accelerated computing infrastructures at scale, announced the HPC and AI solutions it will showcase at Supercomputing 2023 (SC23) in Denver on November 12-17 in booth #843. Penguin’s booth will feature four innovative technology demonstration pods, a dedicated AI Lounge sponsored by NVIDIA, an espresso/coffee bar, and private meeting rooms to connect one-on-one with Penguin experts and executives.
Penguin and its partners will showcase several innovative capabilities demonstrating their passion for all things HPC. SC23 attendees are invited to visit booth #843 to see the depth of Penguin’s technical expertise and why for 25 years the company has been a leading provider in designing, building, deploying, and managing innovative solutions that meet business needs today and support them into the future. Demonstrations will include:
- Design — Penguin offers a tailored approach to designing new computing infrastructures that power AI. Learn how Penguin meets specific customer needs for AI workloads and clears tests to ensure an optimal computing, storage, and networking solution. This demonstration will include:
- AI/HPC cluster and high-performance networking designs
- Optimized node selection, including the latest CPU, GPU, and inference processors
- Data center power optimization to learn how power drives infrastructure selection, rack layout, and cluster density
- Build — Underscoring the high-quality solutions that Penguin experts build, a portfolio of power-efficient Penguin solutions will be showcased along with developments in immersion-cooling technology and rack-level integration and testing. This demonstration will include:
- 277V Altus® XE2342 system with the high-density 4th Gen AMD Epyc™ Processor
- 277V Relion® XE5318GTS system featuring the Intel 4th Gen Xeon® Scalable Processor with the eight-way NVIDIA HGX™ H100
- 48V DC OCP-Inspired™ Tundra® racks and servers
Also Read: Climb Channel Solutions Announces Authorized GitLab Distributor Partnership in North America
- Deploy — With a proven track record of accelerated deployment of AI factories, Penguin will provide insights into hybrid cloud solutions offering rapid availability and integrated access as well as HPC cluster installation and configuration that can be achieved with:
- Scyld ClusterWare™ software for DevOps-based cluster deployment and graphical monitoring/management
- Scyld Cloud Central™ management software, the cloud-first deployment ideal for accelerated system access timelines, hybrid system management, cloud-bursting, and cost/ROI reporting
- Professional Services for expert on-site AI/HPC cluster deployment, security policies, performance assurance, and end-to-end cluster/storage/network integration
- Manage — This demonstration of the advanced Penguin system management software suite also highlights Penguin’s managed services, including Penguin-led system management and operations, along with SLA-based processes, to ensure system availability and ongoing system and security update management. Also featured will be operational dashboards and reporting, which, along with on-site spares inventory management, minimize system downtime due to component failures.
As an NVIDIA certified DGX Managed Service Provider, Penguin will also have an NVIDIA-sponsored AI Lounge in its booth for attendees to stop by and learn more from Penguin and NVIDIA AI experts.
Penguin experts will also participate in the following speaking engagements:
- AI Factories: Exhibitor Forum
- Who: Kevin Tubbs, SVP of Strategic Solutions, Penguin Solutions
- When/Where: Tuesday, November 14 at 2:00 – 2:30 p.m. MT in rooms 503-504
- What: Learn what it takes to deliver AI to the masses – from cost to management of running AI architectures. This session will explore the best of today’s options for companies looking to scale their AI infrastructure.
- Immersion Cooling Panel
- Who: Phil Pokorny, CTO of Penguin Solutions; Cameron Cross, Sales Director of Immersion Cooling at Shell; Daniel Pope, CEO of SubMer; and Curt Wallace, Senior Solutions Architect at GRC. The panel will be moderated by Dan Olds, Chief Research Officer at analyst firm Intersect360.
- When/Where: Friday, November 17 at 8:30 – 10:00 a.m. MT in rooms 201-203
- What: The panel will discuss the need to create more sustainable deployment environments and how immersion cooling is a critical piece to the data center puzzle.
As a long-standing member of the HPC community, Penguin is proud to once again be a sponsor of the Student Cluster Competition. This year, Penguin is supporting “The Roadrunners” team from the University of New Mexico.
Penguin is a gold sponsor of the annual Beowulf Bash, recently rebranded The Y Event. Join us on Monday, November 13 at 9:00 p.m. MT as “Horns and Halos” kicks off SC23!
SOURCE: BusinessWire