Penguin Solutions Enhances OriginAI to Boost AI Factory Deployment and Optimize Performance

Penguin’s OriginAI solution features validated architectures supported by intelligent cluster management software and expert services, enabling companies to swiftly scale their AI infrastructure.

Fremont, Calif. – June 18, 2024 – Penguin Solutions™, an SGH™ brand (Nasdaq: SGH) specializing in the design, deployment, and management of AI and accelerated computing infrastructures at scale, has expanded its OriginAI® solution to include validated, pre-defined AI architectures utilizing NVIDIA technology. With Penguin's intelligent cluster management software and professional services, the OriginAI infrastructure simplifies AI implementation and management while ensuring predictable cluster performance, enhancing customer ROI from clusters ranging from hundreds to thousands of GPUs.

“Designing, deploying, and operating AI factories is inherently complex. Our OriginAI solution leverages Penguin’s extensive AI infrastructure expertise to minimize this complexity and accelerate ROI,” said Pete Manca, President of Penguin Solutions. “This solution equips CEOs and CIOs with the reliable infrastructure needed to manage demanding AI workloads at scale.”

The OriginAI solution combines proven architectures, cutting-edge hardware, advanced cluster management software, and professional expertise to deliver robust infrastructure for critical workloads. Configurations include 1-pod, 4-pod, and 16-pod setups, scalable from 256 to over 16,000 GPUs. Integrated with NVIDIA H100 GPUs, Penguin’s Scyld ClusterWare® 12.2 software, and top-tier networking and storage options, the OriginAI solution is backed by a comprehensive range of managed services.

“Enhancing GPU performance while controlling costs is pivotal for our customers. The validated, scalable OriginAI architectures from Penguin Solutions provide both,” noted Matt Eastwood, SVP of enterprise infrastructure research at IDC. “Penguin continues to empower AI workloads in the most demanding environments through its innovative high-performance solutions and services.”

Utilizing proven architectures, the OriginAI solution employs Penguin’s in-factory burn-in and integration environment to validate AI cluster performance and confirm production readiness prior to shipment. This ensures customers receive optimal performance and return on investment from deployment. By integrating these architectures with Scyld ClusterWare and managed services that proactively monitor cluster health and throughput, OriginAI achieves over 95% cluster efficiency while maximizing GPU performance compared to traditional methods.

As an NVIDIA-certified Elite OEM and DGX AI Compute Systems Solution Provider, Penguin Solutions has been delivering AI factories at scale since 2017. With over 25 years of HPC experience and more than 75,000 GPUs deployed and managed, Penguin is a trusted partner for AI and HPC solutions, serving prominent organizations such as Georgia Tech, Meta, Sandia Labs, and the U.S. Navy.

For more information about Penguin's OriginAI infrastructure solutions, visit https://www.penguinsolutions.com/solutions/ai/originai.

About Penguin Solutions

Penguin Solutions™ includes Penguin Computing™ and Penguin Edge™ products, accelerating digital transformation through emerging technologies in HPC, AI, and IoT. We empower enterprises and government institutions to build an autonomous future, drive discovery, and enhance human potential with advanced infrastructure and services.

Follow Penguin Solutions on LinkedIn, Twitter, YouTube, and Facebook.

Media Contact

SGH Corporate Communications

Maureen O’Leary

Director, Communications, Penguin Solutions/SGH

[email protected]

Penguin Solutions, OriginAI, and Scyld ClusterWare are trademarks of Penguin Computing, Inc., a subsidiary of SMART Global Holdings, Inc. All other trademarks are the property of their respective owners.