Penguin Solutions Expands Its AI Infrastructure Management Software Platform and Introduces Robust AI Optimization Service
Penguin Solutions (PENG) has announced significant expansions to its ICE ClusterWare™ software platform, introducing enhanced features for AI infrastructure management. The platform now includes multi-tenancy support, streamlined workflows, and enhanced orchestration controls to help enterprises optimize their AI ecosystems.
The company also launched the ICE ClusterWare AIM™ service, which leverages patent-pending technology for predictive automation and optimization. This service is designed to maximize performance and operational efficiency through automated remediation, prescriptive maintenance, and workload optimization.
Built on Penguin Solutions' experience managing over two billion hours of GPU runtime and 85,000 deployed GPUs, these solutions aim to address the challenge of underutilized GPU clusters, which typically operate at only 50% of their potential value. The company will showcase these technologies at NVIDIA GTC 2025 in San Jose.
Penguin Solutions (PENG) ha annunciato significative espansioni alla sua piattaforma software ICE ClusterWare™, introducendo funzionalità avanzate per la gestione delle infrastrutture AI. La piattaforma ora include supporto alla multiutenza, flussi di lavoro semplificati e controlli di orchestrazione migliorati per aiutare le imprese a ottimizzare i loro ecosistemi AI.
L'azienda ha anche lanciato il servizio ICE ClusterWare AIM™, che sfrutta una tecnologia in attesa di brevetto per l'automazione predittiva e l'ottimizzazione. Questo servizio è progettato per massimizzare le prestazioni e l'efficienza operativa attraverso la rimediazione automatizzata, la manutenzione prescrittiva e l'ottimizzazione del carico di lavoro.
Basate sull'esperienza di Penguin Solutions nella gestione di oltre due miliardi di ore di runtime GPU e 85.000 GPU distribuite, queste soluzioni mirano ad affrontare la sfida dei cluster GPU sottoutilizzati, che operano tipicamente solo al 50% del loro valore potenziale. L'azienda presenterà queste tecnologie al NVIDIA GTC 2025 a San Jose.
Penguin Solutions (PENG) ha anunciado importantes expansiones en su plataforma de software ICE ClusterWare™, introduciendo características mejoradas para la gestión de infraestructuras de IA. La plataforma ahora incluye soporte para múltiples inquilinos, flujos de trabajo simplificados y controles de orquestación mejorados para ayudar a las empresas a optimizar sus ecosistemas de IA.
La compañía también lanzó el servicio ICE ClusterWare AIM™, que aprovecha una tecnología pendiente de patente para la automatización predictiva y la optimización. Este servicio está diseñado para maximizar el rendimiento y la eficiencia operativa a través de la remediación automatizada, el mantenimiento prescriptivo y la optimización de la carga de trabajo.
Basadas en la experiencia de Penguin Solutions en la gestión de más de dos mil millones de horas de tiempo de ejecución de GPU y 85,000 GPUs desplegadas, estas soluciones tienen como objetivo abordar el desafío de los clústeres de GPU subutilizados, que operan típicamente solo al 50% de su valor potencial. La empresa presentará estas tecnologías en el NVIDIA GTC 2025 en San José.
펭귄 솔루션(PENG)은 AI 인프라 관리 기능을 향상시킨 ICE ClusterWare™ 소프트웨어 플랫폼의 중요한 확장을 발표했습니다. 이 플랫폼은 이제 다중 테넌시 지원, 간소화된 워크플로우 및 향상된 오케스트레이션 제어 기능을 포함하여 기업들이 AI 생태계를 최적화하는 데 도움을 줍니다.
회사는 또한 예측 자동화 및 최적화를 위한 특허 출원 중인 기술을 활용하는 ICE ClusterWare AIM™ 서비스를 출시했습니다. 이 서비스는 자동화된 수정, 처방적 유지보수 및 작업 부하 최적화를 통해 성능과 운영 효율성을 극대화하도록 설계되었습니다.
펭귄 솔루션이 20억 시간 이상의 GPU 런타임 및 85,000개의 배포된 GPU를 관리한 경험을 바탕으로, 이러한 솔루션은 일반적으로 잠재 가치의 50%만으로 운영되는 GPU 클러스터의 문제를 해결하는 것을 목표로 합니다. 회사는 이 기술들을 NVIDIA GTC 2025에서 샌호세에서 선보일 예정입니다.
Penguin Solutions (PENG) a annoncé d'importantes extensions de sa plateforme logicielle ICE ClusterWare™, introduisant des fonctionnalités améliorées pour la gestion des infrastructures d'IA. La plateforme comprend désormais un support multi-locataire, des flux de travail simplifiés et des contrôles d'orchestration améliorés pour aider les entreprises à optimiser leurs écosystèmes d'IA.
L'entreprise a également lancé le service ICE ClusterWare AIM™, qui exploite une technologie en instance de brevet pour l'automatisation prédictive et l'optimisation. Ce service est conçu pour maximiser les performances et l'efficacité opérationnelle grâce à la remédiation automatisée, à la maintenance prescriptive et à l'optimisation des charges de travail.
Reposant sur l'expérience de Penguin Solutions dans la gestion de plus de deux milliards d'heures de temps d'exécution GPU et de 85 000 GPU déployés, ces solutions visent à relever le défi des clusters GPU sous-utilisés, qui fonctionnent généralement à seulement 50 % de leur valeur potentielle. L'entreprise présentera ces technologies au NVIDIA GTC 2025 à San José.
Penguin Solutions (PENG) hat bedeutende Erweiterungen seiner ICE ClusterWare™-Softwareplattform angekündigt und dabei verbesserte Funktionen für das Management von KI-Infrastrukturen eingeführt. Die Plattform umfasst jetzt Multi-Tenant-Unterstützung, optimierte Arbeitsabläufe und verbesserte Orchestrierungssteuerungen, um Unternehmen dabei zu helfen, ihre KI-Ökosysteme zu optimieren.
Das Unternehmen hat auch den ICE ClusterWare AIM™-Dienst gestartet, der eine patentierte Technologie für prädiktive Automatisierung und Optimierung nutzt. Dieser Dienst zielt darauf ab, die Leistung und die Betriebseffizienz durch automatisierte Behebung, vorschreibende Wartung und Arbeitslastoptimierung zu maximieren.
Basierend auf der Erfahrung von Penguin Solutions, die über zwei Milliarden Stunden GPU-Laufzeit und 85.000 bereitgestellte GPUs verwaltet haben, zielen diese Lösungen darauf ab, die Herausforderung unterausgelasteter GPU-Cluster zu bewältigen, die typischerweise nur 50 % ihres potenziellen Wertes nutzen. Das Unternehmen wird diese Technologien auf der NVIDIA GTC 2025 in San Jose präsentieren.
- Patent-pending technology for AI infrastructure optimization
- Management experience of 2+ billion GPU runtime hours
- Large-scale deployment of 85,000 GPUs
- Hardware-agnostic platform enabling flexible implementation
- Current industry GPU clusters operating at only 50% utilization
Insights
Penguin Solutions' expansion of its ICE ClusterWare platform represents a strategic enhancement of its AI infrastructure management capabilities. The addition of multi-tenancy support, enhanced orchestration controls, and streamlined workflows directly addresses critical pain points in enterprise AI deployment. These features solve the fundamental challenge of resource isolation while simplifying the management of complex AI environments.
The new ICE ClusterWare AIM service leverages patent-pending technology and the company's extensive operational data from managing over two billion hours of GPU runtime. This positions Penguin Solutions to capitalize on a significant market inefficiency highlighted by IDC - the fact that traditional approaches often result in GPU clusters operating at just 50% of their potential value.
What's particularly noteworthy is how Penguin Solutions is tackling both deployment complexity and operational optimization simultaneously. Their automated remediation and prescriptive maintenance capabilities address the substantial hidden costs of AI infrastructure - downtime and underutilization - which often exceed the initial hardware investment.
While competitors focus primarily on either deployment or optimization, Penguin's integrated approach creates a more compelling value proposition for enterprises struggling with scaling AI initiatives. The hardware-agnostic nature of their platform also provides flexibility that proprietary solutions lack, potentially expanding their addressable market across diverse enterprise environments.
Penguin Solutions' product expansion represents an incremental advance in their AI infrastructure management portfolio rather than a transformative offering. While the new capabilities address real pain points in enterprise AI deployment, the announcement lacks specific performance metrics or customer success stories that would validate their effectiveness.
The market for AI infrastructure optimization is becoming increasingly crowded with both established players and startups offering similar capabilities. Penguin's claim of managing over 85,000 GPUs demonstrates scale but doesn't necessarily translate to competitive advantage without evidence of superior outcomes.
The multi-tenancy support is particularly interesting as it enables more efficient resource sharing across departments, potentially improving the economics of AI infrastructure investments. However, the announcement characterizes this as "foundational support" with "further multi-tenancy capabilities" to be added later, suggesting the functionality may not be fully mature.
From an investment perspective, this incremental product evolution likely represents business-as-usual product development rather than a step-change in capability or market position. The IDC quote highlighting that traditional approaches lead to GPU clusters operating at only 50% of potential value underscores the opportunity, but Penguin must still prove their solution can significantly outperform this baseline to drive meaningful adoption.
The timing of the announcement ahead of NVIDIA GTC 2025 appears strategic, positioning Penguin Solutions to leverage industry attention during the conference to generate interest in their expanded capabilities.
New multi-tenancy support, simplified workflows and controls, and predictive automation enable enterprises to rapidly and effectively manage and optimize AI infrastructure at scale
Penguin Solutions' ICE ClusterWare software platform is a hardware-agnostic, intelligent software platform for managing, scaling, and optimizing AI, HPC, and data infrastructure. It combines open-source, industry-standard, and third-party tools with Penguin Solutions’ innovative software to create the foundation of an Intelligent Compute Environment. (Graphic: Business Wire)
Additionally, Penguin Solutions unveiled its ICE ClusterWare AIM™ service, an advanced optimization service designed to maximize performance, availability, and operational efficiency of AI infrastructure through predictive automation.
“AI computing demands an integrated, intelligent approach that goes beyond traditional IT infrastructure,” said Trey Layton, vice president of software and product management, Penguin Solutions. “With ICE ClusterWare software and ICE ClusterWare AIM service, we enable enterprises to build intelligent compute environments that optimize efficiency, scalability, and cost-effectiveness—ensuring AI workloads run at peak performance while minimizing operational complexity.”
The ICE ClusterWare software platform is a hardware-agnostic, intelligent software platform for managing, scaling, and optimizing AI, HPC, and data infrastructure. It combines open-source, industry-standard, and third-party tools with Penguin Solutions’ innovative software to create the foundation of an Intelligent Compute Environment, enabling IT leaders to confidently deploy and scale complex infrastructure that meets the specific business needs of their organizations.
With the latest updates, the ICE ClusterWare software platform now includes:
- Multi-tenancy Foundational Support – Enables secure, logical resource isolation and dynamic workload partitioning for multiple users and departments from which further multi-tenancy capabilities will be added.
- Enhanced Orchestration Controls – Introduces a unified control plane for managing sub-clusters and associated infrastructure.
- Streamlined Workflows – Simplifies AI infrastructure deployment and scaling through automated policy enforcement and intelligent job scheduling.
The new ICE ClusterWare AIM service applies Penguin Solutions’ patent-pending software, automation, and operational intelligence, derived from managing more than two billion hours of GPU runtime. Designed for IT and data center operations teams managing large-scale AI and HPC clusters, ICE ClusterWare AIM service delivers:
- Automated Remediation – Detects and resolves system inefficiencies before they impact performance.
- Prescriptive Maintenance – Uses patent pending technology to identify and mitigate failures before they occur, reducing downtime.
- Operational Efficiency at Scale – Working with common scheduling software, optimizes workload distribution to maximize GPU utilization and improve ROI on AI infrastructure investments.
“Navigating AI infrastructure at scale is a challenge—traditional IT approaches often lead to underutilized GPU clusters, operating at only
Penguin Solutions’ ICE ClusterWare software platform and ICE ClusterWare AIM service are included in Penguin Solutions OriginAI® Infrastructure Solution. All are built leveraging the company’s 25+ years of experience designing, building, deploying and managing advanced computing, including more than 85,000 GPUs currently deployed and managed.
Penguin Solutions will showcase its ICE ClusterWare software platform and OriginAI® infrastructure solution at NVIDIA GTC 2025, booth #332, at the San Jose McEnery Convention Center, the week of March 17, 2025.
Penguin Solutions, ICE ClusterWare, and OriginAI are trademarks or registered trademarks owned within the family of companies owned by Penguin Solutions, Inc. All other trademarks are the property of their respective owners.
About Penguin Solutions
The most exciting technological advancements are also the most challenging for companies to adopt. At Penguin Solutions, we support our customers in achieving their ambitions across our computing, memory, and LED lines of business. With our expert skills, experience, and partnerships, we turn our customers’ most complex challenges into compelling opportunities.
For more information, visit https://www.penguinsolutions.com.
View source version on businesswire.com: https://www.businesswire.com/news/home/20250304897051/en/
Penguin Solutions Contact
Maureen O’Leary
Corporate Communications
602-330-6846
pr@penguinsolutions.com
Source: Penguin Solutions, Inc.
FAQ
What new features does Penguin Solutions (PENG) add to ICE ClusterWare platform in 2025?
How many GPUs does Penguin Solutions (PENG) currently manage?
What is the ICE ClusterWare AIM service launched by PENG?
Where will Penguin Solutions (PENG) showcase its new AI infrastructure solutions?
What problem does PENG's new AI infrastructure solution address?
