Hewlett Packard Enterprise Crushes Earnings on AI Server Demand

What happened

Hewlett Packard Enterprise reported fiscal second-quarter revenue of $10.68 billion, up 40% year-over-year, significantly exceeding the $9.79 billion analyst estimate. Adjusted earnings per share reached $0.79, well above the $0.53 consensus forecast, marking the company's largest earnings-per-share beat since February 2018. The Cloud & AI segment generated $7.71 billion in revenue, with server revenue specifically hitting $5.45 billion versus the $4.66 billion estimate. The networking business saw revenue surge 148% to $2.7 billion. Net earnings reached $624 million, a sharp turnaround from a $1.05 billion net loss in the prior-year period. Free cash flow for the quarter totaled $915 million, an improvement of $1.8 billion from the previous year.

Why it matters

The results demonstrate HPE's ability to capitalize on enterprise and government demand for on-premises AI infrastructure, a segment where the company faces less direct competition from hyperscalers than rivals targeting cloud providers. CEO Antonio Neri noted that customers in security-focused industries are significantly increasing their investments in on-premises AI servers rather than cloud-based alternatives. The company reported record order volumes, with bookings for traditional servers up triple digits and creating the largest backlog in company history. HPE raised its fiscal 2026 adjusted EPS guidance to $3.35 to $3.45, up from $2.30 to $2.50, and increased its revenue growth outlook to 29% to 33% from 17% to 22%. The updated targets exceed financial goals originally set for fiscal 2028, effectively pulling forward long-term performance by two years.

Bigger picture

HPE's results reinforce the broad-based nature of AI infrastructure spending, extending beyond cloud hyperscalers to enterprise customers modernizing data centers and deploying on-premises AI capabilities. The company's focus on higher-margin enterprise and government customers distinguishes it from competitors like Dell Technologies, which targets neoclouds and bulk server buyers. The strong networking performance, boosted by the Juniper Networks acquisition, positions HPE to capture spending on AI-optimized data center infrastructure. The company is preparing to launch new ProLiant servers based on Nvidia's Vera CPUs, optimized for agentic AI workloads requiring real-time reasoning. However, HPE expects elevated memory costs to persist until at least 2027, reflecting ongoing supply constraints in the broader semiconductor market.

What to watch

Investors should monitor HPE's fiscal third-quarter results, for which the company guided revenue of $11.5 billion to $12.1 billion and adjusted EPS of $0.88 to $0.93, both significantly above analyst expectations of $10.9 billion and $0.58 respectively. The fall launch of ProLiant servers based on Nvidia Vera CPUs will test HPE's ability to capture demand for AI inference workloads. Progress on integrating Juniper Networks and realizing cost synergies from the Catalyst initiative will be important for sustaining margin improvement. Longer term, HPE's fiscal 2027 framework targets revenue growth of 8% to 12%, adjusted EPS growth of 12% to 16%, and free cash flow of at least $4.5 billion.