Hewlett Packard Enterprise (HPE) Q2 2026 earnings review
Juniper Integration and AI Surge Fast-Track FY28 Targets
HPE delivered a blowout quarter, hitting its FY28 long-term financial commitments two years ahead of schedule. Fueled by a faster-than-expected Juniper integration and unyielding demand for AI infrastructure, revenue surged 40% YoY to $10.7 billion, while Non-GAAP EPS skyrocketed 108% to $0.79. Order growth is vastly outpacing revenue, pushing backlog to record highs. Management's execution is pristine, raising FY26 FCF guidance to at least $3.5B (up from $2.0B). The primary constraint is no longer customer demand, but the company's ability to secure component supply in a highly inflationary memory market.
🐂 Bull Case
The integration is running ahead of schedule, contributing heavily to a 148% reported YoY jump in Networking revenue. Cross-selling is accelerating, with campus and branch orders at record highs.
AI systems backlog hit $5.9B. The modernization of data centers for AI is dragging traditional server demand upward (orders up triple-digits), providing a robust, dual-engine tailwind.
🐻 Bear Case
DRAM and NAND inflation remains a persistent macro headwind. HPE is forcing price hikes to protect margins, which inherently limits unit growth elasticity.
Despite massive revenue growth, Networking operating margins dropped sequentially from 23.7% in Q1 to 21.6% in Q2, indicating near-term integration or mix pressures.
⚖️ Verdict: 🔴🔴
Strongly Bullish. The combination of flawless M&A integration (Juniper) and structural AI tailwinds has fundamentally elevated HPE's baseline earnings power, validating the 40% EPS guidance raise for the year.
Key Themes
Juniper Synergies Reversing Historical Networking Slump
Networking is accelerating dramatically. Q2 revenue hit $2.7 billion (up 148% reported, ~10% normalized). Importantly, campus and branch orders grew in the upper 20% range. The 'self-driving network' narrative is translating to real dollars, with Wi-Fi 7 access point sales increasing more than 7x. Synergy realization is expected to exceed the annual target of $200 million by fiscal year-end.
AI Systems Momentum Remains Unbroken
Cloud & AI revenue grew 23% YoY to $7.7 billion. HPE booked $1.8 billion in new AI system orders, bringing the cumulative total to $16.4 billion. Management noted a distinct shift toward enterprise AI inferencing, moving beyond just hyperscalers. The pipeline is stated to be 'multiples' of the $5.9 billion backlog.
Traditional Server Refresh Cycle Accelerating
Traditional server orders increased triple digits. The shift toward Agentic AI workloads is forcing enterprises to modernize their standard compute infrastructure with high-memory configurations. This creates a stable, highly profitable revenue stream outside of the lumpy sovereign AI deals.
Commodity Cost Inflation
A persistent macro headwind: DRAM and NAND costs are significantly inflating. While HPE's ability to shape demand and pass through costs via price hikes is keeping gross margins stable (Non-GAAP GM at 36.9%), this inflation creates friction on unit volumes and tests customer demand elasticity.
Networking Margin Deceleration Contradicts Revenue Surge
A notable contradiction: While networking revenue surged 15% sequentially, the segment's operating margin actually decelerated from 23.7% in 26Q1 to 21.6% in 26Q2. Management attributed this to higher variable compensation and Q1 one-time benefits, but this warrants close monitoring to ensure Juniper's lower historical margin profile isn't permanently dragging down segment profitability.
Supply Chain Cap on Upside
The primary gating factor to growth is not demand, but supply availability. Management explicitly stated that the raised 2026 guidance assumes no incremental supply unlocks. If HPE cannot secure more wafers and components from silicon partners, the record backlog will simply age rather than convert to cash.
Catalyst Cost Program Execution
HPE is running lean. The workforce is down 9% since the Juniper integration and Catalyst programs began, bringing headcount to its lowest level as a combined company (~65k). Generative AI process simplification is driving 20% of the Catalyst savings, directly flowing through to the bottom line.
Other KPIs
Reversing. FCF dramatically improved by $1.8 billion YoY from an $847 million outflow in 25Q2. The cash conversion cycle improved by 2 days sequentially, driven by extended payables despite deliberately holding higher inventory ($9.0B) to secure parts for H2 AI server shipments.
Accelerating. Up 220 basis points sequentially and up from 6.6% a year ago. Driven by the pass-through of higher costs in traditional servers, improved pricing execution, and scale benefits from HPE Financial Services (which hit a record >30% ROE).
Guidance
Accelerating. Raised massively from the prior trajectory. Management shattered its previously stated FY28 goal of 'at least $3.00' two full years early, reflecting immense operating leverage and rapid synergy capture.
Accelerating. Raised from previous estimates. This factors in the full integration of Juniper Networks and resilient server pricing. The midpoint implies total revenue approaching $44 billion for the year.
Stable to slightly Decelerating on a reported basis (vs FY26's massive M&A-driven jump), but represents exceptionally strong organic momentum across both Networking and Cloud & AI segments.
Accelerating. Represents high sequential growth from Q2's $10.7B, driven by back-half weighted AI systems shipments and accelerating campus networking orders.
Key Questions
Networking Margin Dilution
Networking margins stepped down 210 bps sequentially despite massive revenue scale. Is this the bottom for the combined margin profile, or will Juniper's historically lower margin structure cap the upside despite the synergy realization?
Component Supply Elasticity
You noted that 2026 guidance assumes no incremental supply unlocks. If silicon constraints ease in late 2026 or 2027, what is the realistic upside to the 8-12% revenue growth framework for FY27?
Price Hikes vs Demand Destruction
As you pass through elevated DRAM and NAND costs to enterprise customers via server price hikes, are you seeing any lengthening of sales cycles or downsizing of cluster configurations?
AI Systems Backlog Conversion
The AI backlog sits at $5.9 billion. Given the complexities of sovereign data center readiness and power constraints, what percentage of this backlog is at risk of sliding out of the FY27 window entirely?