We are the Frontier
U.S.

HPE Private Cloud AI Adds Support for Mixed GPU Clusters

The update enables management of NVIDIA H200 and RTX PRO 6000 hardware within a single, unified enterprise cluster.

Updated on Oct. 28, 2026 in Artificial Intelligence

Isometric editorial illustration of a clean, modular grid of server hardware modules, representing modern enterprise data infrastructure.
Hewlett Packard Enterprise has updated its Private Cloud AI infrastructure to support mixed clusters, allowing enterprises to manage diverse GPU hardware under a single, unified scheduling system. AI Illustration. Upload story photo >

Live Poll

Is now a good time for your organization to adopt new AI infrastructure management tools?

Hewlett Packard Enterprise has updated its Private Cloud AI infrastructure to support mixed clusters containing NVIDIA H200 and RTX PRO 6000 Blackwell Server Edition GPUs. The update also introduces the KAI Scheduler, a new tool designed to manage GPU allocation through quota- and priority-based logic.

Why it matters

This shift allows enterprise IT teams to integrate varying GPU architectures within a single cluster, addressing the efficiency challenges of managing specialized AI hardware. It marks a broader industry push to optimize GPU utilization and workload distribution in hybrid enterprise environments.

The system now supports NVIDIA H200 and RTX PRO 6000 Blackwell Server Edition GPUs side-by-side. The KAI Scheduler manages these resources by enforcing project-specific quotas and task priorities to prevent resource contention.

The players

Hewlett Packard Enterprise

An enterprise computing firm providing infrastructure, hardware, and AI cloud services to global data centers.

NVIDIA

A developer of specialized GPU hardware and software frameworks, including the Blackwell architecture.

The details

Mixed-GPU support enables infrastructure teams to pool different hardware generations into a single management domain. The KAI Scheduler functions by acting as an orchestrator that evaluates incoming workloads and assigns them to available GPU cycles based on pre-defined priority tiers. This replaces manual resource partitioning with automated, policy-driven allocation.

Timeline

  1. June 16, 2026: HPE announced secure local agent registration and centralized approval.

  2. October 21, 2026: HPE announced mixed GPU support and KAI Scheduler availability.

  3. Q4 2026: Scheduled integration of NVIDIA OpenShell and expanded developer access.

  4. 2027: Planned release of self-service agent and credential management tools.

The Tech Race

This development follows the pattern set by the NVIDIA Blackwell GPU architecture, as infrastructure providers update management software to support new, diverse hardware. It represents a significant step in the race to provide unified management layers for heterogeneous enterprise AI clusters.

Enterprise IT teams can now consolidate disparate GPU resources into a single environment, simplifying hardware management workflows. These updates are available now, with further tools for self-service agent registration planned for 2027.

The takeaway

The move toward mixed-GPU clusters signifies an industry shift toward maximizing hardware lifespan in high-performance computing environments. Watch for the 2027 release of self-service agent tools to see how HPE further automates the administrative side of its AI cloud platform.

What happens next

OpenShell integration and expanded developer access for AI Essentials are scheduled to occur in the fourth quarter of 2026.

Further reading

For broader trends in infrastructure management, explore our Artificial Intelligence section.

Source note: This article includes information reported by EWEEK.

Live Poll

Is now a good time for your organization to adopt new AI infrastructure management tools?