Penguin Solutions Expands AI Infrastructure Offerings with Latest GPU and CPU Technologies

Penguin Solutions Boosts AI Infrastructure with Latest GPU and CPU Technologies

Penguin Solutions, Inc. (Nasdaq: PENG) has announced significant upgrades to its AI and high-performance computing (HPC) infrastructure offerings, incorporating the latest GPU and CPU technologies from industry leaders NVIDIA and AMD 1

1

2

OriginAI Infrastructure Expansion

The company's OriginAI infrastructure solution has been enhanced with the addition of NVIDIA H200 and AMD Instinct MI300X GPU-powered servers. This expansion includes support for Penguin's own Relion and Altus servers, as well as Dell PowerEdge XE9680 servers. Penguin Solutions plans to enable NVIDIA Blackwell B200 and AMD Instinct MI325X within OriginAI solutions in early 2025, further broadening the range of options for organizations looking to deploy AI infrastructure at scale 1

1

Enhanced Cluster Management Capabilities

Accompanying the hardware upgrades, Penguin Solutions has updated its Scyld ClusterWare software platform. The enhancements include:

1. Improved infrastructure security and encrypted communications
2. Expanded node management support for NVIDIA Cumulus Linux and Edge-core's SONiC switches
3. Enhanced node status polling
4. Increased support for Ubuntu within the ClusterWare platform 1

   1

These improvements aim to drive higher cluster performance and ease of management in diverse computing environments.

New Altus Servers with AMD EPYC 9005 Series Processors

In a separate announcement, Penguin Solutions revealed the availability of its new Altus servers powered by AMD EPYC 9005 Series processors. These processors boast a leading 192 core count and 12 channels of DDR5 memory, making them particularly suitable for compute-intensive HPC and mixed-use AI workloads 2

2

Key Features of the New Altus Servers

• Air-cooled models Altus XE1312v2 and Altus XE2312v2 are currently available
• Liquid-cooled Altus XE2342v2 servers are targeted for availability in the first half of 2025
• The servers offer best-in-class price-to-performance value
• They are designed to architect HPC and AI infrastructure, including AI clusters that support mixed-use workloads 2

  2

Related Stories

Dell and AMD expand on-premises AI platform with next-generation MI350P GPU support

AMD Helios AI rack system takes direct aim at Nvidia with Microsoft, OpenAI backing

AMD Unveils MI325X AI Chip to Challenge Nvidia's Dominance

Supermicro Unveils Advanced AI and HPC Solutions with NVIDIA Blackwell Platform

Industry Perspective

Matt Eastwood, senior vice president worldwide research at IDC, commented on the importance of well-designed AI infrastructure: "AI requires carefully laid out infrastructure that is integrated by expert hands and managed by specialized tools in order for the full benefits to be realized" 1

1

Availability and Showcase

Penguin Solutions is showcasing these new offerings at SC24 Supercomputing, booth #1523 at the Georgia World Congress Center in Atlanta from November 18-22. Attendees can learn more about the OriginAI solution, ClusterWare software platform, and Penguin's collaborations with NVIDIA, AMD, Dell, and other partners 1

1

2

These advancements by Penguin Solutions demonstrate the company's commitment to providing cutting-edge AI and HPC infrastructure solutions, enabling organizations to rapidly deploy and manage complex computing environments for advanced workloads and research applications.