Air-cooled Blackwell acceleration for enterprise AI, inference, and reasoning at scale.
Organizations are moving beyond AI experimentation to production deployments that support increasingly sophisticated workloads, including AI reasoning, agentic AI, large-context inference, and model adaptation. As AI models continue to grow in scale and complexity, organizations need infrastructure that delivers predictable performance, operational efficiency, and the confidence to deploy and scale AI in production.
Penguin Solutions ComputeAI IEA8-B300 is a pre-validated, air-cooled NVIDIA HGX™ B300-based platform designed to help organizations deploy enterprise AI infrastructure with greater speed and predictability. By integrating validated compute, memory, networking, storage, power, and cooling technologies into a production-ready architecture, the platform helps reduce integration complexity and accelerate deployment.
Designed for enterprise, neocloud, and sovereign AI environments, ComputeAI IEA8-B300 enables organizations to deploy next-generation NVIDIA Blackwell acceleration while leveraging existing air-cooled data center infrastructure.
ComputeAI IEA8-B300 is designed to help organizations operationalize and scale AI by supporting advanced AI workloads that drive measurable business outcomes.
Organizations are increasingly using AI reasoning and large-context inference to analyze larger volumes of information and maintain context across complex interactions. ComputeAI IEA8-B300 combines high-performance NVIDIA Blackwell acceleration, large-scale memory, and high-bandwidth networking to support these workloads, helping organizations generate more accurate insights and make better-informed decisions.
Agentic AI applications orchestrate tasks, interact with enterprise systems, and execute multi-step business processes. ComputeAI IEA8-B300 provides the performance, scalability, and infrastructure foundation required to support these advanced workloads in production environments, helping organizations improve efficiency and automate increasingly complex operations.
As AI adoption expands, organizations must support growing numbers of users, applications, and AI requests while maintaining performance and responsiveness. ComputeAI IEA8-B300 is designed for high-concurrency inference environments that enable organizations to scale AI services across business functions while maintaining a consistent user experience.
Organizations are increasingly adapting and fine-tuning models using proprietary data, industry expertise, and business-specific knowledge. ComputeAI IEA8-B300 provides the compute, memory, and networking resources required for model adaptation and fine-tuning, helping organizations create differentiated AI capabilities and maximize the value of their AI investments.
Full specifications for ComputeAI IEA8-B300.
| Component | Validated Configuration |
|---|---|
| GPU | 8x NVIDIA B300 GPUs |
| GPU Memory | 2,304 GB HBM3e |
| CPU | 2x Intel Xeon 6767P |
| Host Memory | 3 TB / 4 TB DDR5 RDIMM (32 x 96 GB / 128 GB) |
| GPU Networking | 8x NVIDIA ConnectX-8 800 Gb/s |
| DPU | 2x NVIDIA BlueField-3 B3220 |
| Local Storage | 8x 3.84 TB E1.S NVMe SSDs |
| OS Storage | 2x 1.92 TB M.2 NVMe SSDs |
| Power | 12x 3200W Titanium PSUs |
| Supported Operating System | Ubuntu, RHEL, Rocky, NVIDIA Base Command (Optional) |
| Warranty | 3-Year Standard Warranty* |
Download product documentation and resources
Our team of experts is ready to help you find the perfect solution for your business needs. Get personalized advice and competitive quotes.
We're here to help with any questions