HPE targets 'self-driving' AI networking via unified Juniper and Aruba integration
Hewlett Packard Enterprise is promoting a 'self-driving network' strategy for AI infrastructure by integrating its Apstra, Morpheus, and Juniper Networks portfolios. The strategy focuses on using AI-native networking tools and high-speed, liquid-cooled switching hardware to manage the data throughput requirements of modern data centers.
Key Takeaways
- New QFX5250 liquid-cooled switch supports 64 ports with speeds reaching 1.6 terabits per second.
- HPE GreenLake Intelligence now acts as the cross-domain orchestration layer for networking, compute, and storage.
- Integrated management combines Apstra Data Center Director's graph database with the HPE Morpheus hybrid cloud platform.
- Juniper’s Mist AIOps platform has been expanded to support HPE Networking CX switches and Aruba Central.
Why It Matters
HPE is betting that networking complexity, rather than raw compute power, is the primary barrier to enterprise AI scaling. By absorbing Juniper Networks’ Mist AI and Apstra assets into its GreenLake ecosystem, HPE is positioning itself as a vertically integrated alternative to Cisco’s dominant market position. For the streaming industry, these 'self-driving' networks represent a critical shift toward infrastructure that can automatically reroute traffic or scale bandwidth to support massive data throughput for generative AI and high-bitrate video delivery without manual intervention. Success depends on whether this software-defined automation can effectively eliminate the configuration errors that currently plague high-density data centers. Watch for the adoption rate of liquid-cooled hardware as power densities exceed traditional air-cooling limits.
Additional Context
The strategic push into AI-native networking follows the formal completion of HPE’s $14 billion acquisition of Juniper Networks on July 2, 2025. Per AlphaSpread and GuruFocus (July 2025), the deal was finalized after a settlement with the U.S. Department of Justice required HPE to divest its Instant On business and auction non-exclusive licenses for Juniper’s Mist AIOps source code to maintain market competition. This acquisition effectively doubled HPE’s networking business, positioning it to challenge Cisco's estimated 76% market share in computer networking, as noted by CSI Markets and Barron's (November 2024).
Beyond software integration, HPE is aggressively pursuing hardware density and thermal efficiency. On July 23, 2026, HPE introduced the Cray Supercomputing GX5000 at AMD's Advancing AI event. Per The Futurum Group (July 2026), the system features 100% direct liquid cooling and achieves 40% higher core density per rack than its closest competitors. This focus on high-density, liquid-cooled infrastructure addresses the surging power requirements of AI workloads, which can reach up to 150 kW per rack. Early adopters like Oak Ridge National Laboratory are already deploying these systems for exascale and AI research via the 'Discovery' and 'Lux' projects.
Financial analysts suggest this pivot is essential for HPE’s long-term margin profile. During the fiscal Q3 2025 earnings call, HPE reported record revenue of $8.85 billion, and management projected that the combined networking business would eventually contribute over 50% of total company operating income. Per Illuminaire (July 2026), the integration of Juniper’s portfolio into the GreenLake 'as-a-service' model is intended to shift HPE away from traditional hardware cycles toward high-margin, recurring software revenue driven by AI operations.
Read full article at siliconangle.com
Enjoy our coverage?
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source