HPE's Q3 2025: Contradictions Emerge on AI Server Margins, Networking Margins with Juniper, and AI Demand

Generated by AI AgentAinvest Earnings Call Digest
Wednesday, Sep 3, 2025 9:52 pm ET2min read
HPE--
Aime RobotAime Summary

- HPE reported $9.1B Q3 revenue (+18% YOY), driven by AI, networking, and hybrid cloud growth post-Juniper acquisition.

- Networking revenue surged 54% YOY to $1.7B, while AI systems hit $1.6B (25% YOY growth) amid strong server demand.

- Management raised FY25 non-GAAP EPS guidance to $1.88–$1.92, with Q4 revenue projected at $9.7B–$10.1B and improved margins.

- Juniper integration faces margin pressures (networking OM ~20%), but AI strategy emphasizes networking leadership in sovereign and enterprise segments.

- Q&A highlighted AI server margin normalization (~10%), Juniper's role in expanding AI networking reach, and disciplined cost management amid product transitions.

The above is the analysis of the conflicting points in this earnings call

Date of Call: None provided

Financials Results

  • Revenue: $9.1B, up 18% YOY and up 18% sequentially; $8.7B excluding Juniper (up 11% YOY)
  • EPS: Non-GAAP $0.44 per diluted share (toward high end of $0.40–$0.45 guidance); GAAP $0.21 (below $0.24–$0.29 guidance)
  • Gross Margin: 29.9% (non-GAAP, incl. Juniper), down 190 bps YOY and up 50 bps sequentially; 28.3% excl. Juniper
  • Operating Margin: 8.5% (non-GAAP, incl. Juniper), down 150 bps YOY and up 50 bps sequentially; 8.1% excl. Juniper

Guidance:

  • FY25 revenue growth expected at 14%–16% CC (≈30 bps FX headwind).
  • Q4 non-GAAP gross margin mid-30%; FY25 above 30%.
  • FY25 non-GAAP operating margin in upper 9% range; Q4 to upper 11% range.
  • FY25 GAAP EPS $0.42–$0.46; non-GAAP EPS raised to $1.88–$1.92.
  • FY25 FCF ≈$700M (≈$1.0B excluding Juniper).
  • Q4 revenue $9.7B–$10.1B.
  • Q4 Networking revenue up >60% QoQ; Networking OM in low-20% (Q4 and FY25).
  • Q4 Server revenue down mid-to-high single digits QoQ; AI systems down >30% QoQ; Server OM ~10%.
  • Q4 Hybrid Cloud revenue roughly flat QoQ; OM mid-to-high single digits.
  • Q4 GAAP EPS $0.50–$0.54; non-GAAP EPS $0.56–$0.60.

Business Commentary:

  • Revenue Growth and Strategic Acquisitions:
  • HPE achieved record-breaking revenue of $9,100,000,000 in Q3, up 18% year over year.
  • The growth was driven by strong momentum across AI, networking, and hybrid cloud, as well as the completion of the Juniper acquisition.

  • Networking Segment Performance:

  • Networking segment revenue was $1,700,000,000, increasing 54% year over year.
  • This growth was attributed to the recovery in the networking market and the consolidation of Juniper's results for one month.

  • AI System and Server Demand:

  • AI systems revenue reached an all-time high of $1,600,000,000, with a 25% year over year increase.
  • The growth was driven by strong demand for traditional servers and the conversion of AI backlog to revenue.

  • Hybrid Cloud and Software Growth:

  • Hybrid cloud revenue was $1,500,000,000, marking the fourth consecutive quarter of year-over-year growth.
  • This was due to strong software and services revenue, particularly in GreenLake cloud, driven by innovations like GreenLake Intelligence.

Sentiment Analysis:

  • Management reported record revenue ($9.1B, +18% YOY), strong AI momentum (AI orders nearly doubled QoQ; backlog $3.7B), and raised FY25 non-GAAP EPS to $1.88–$1.92. Networking revenue rose 54% YOY; Server and Hybrid Cloud delivered growth with sequential margin improvement. Q4 guide calls for higher operating margins and sequential EPS growth, with full-quarter Juniper benefits.

Q&A:

  • Question from Aaron Rakers (Wells Fargo): How should we think about traditional vs. AI server margins and the path to ~10% total server OM?
    Response: Traditional server margins are back to historical ~10–12%; Q3 total server OM was 6.4% due to AI mix/one large deal and inventory work, but mix shifts and pricing discipline support ~10% server OM in Q4.

  • Question from Wamsi Mohan (Bank of America): Early Juniper integration progress and where AI value is larger—networking or servers?
    Response: Integration is on track (sales harmonization by year-end; channels incentivized); AI strategyMSTR-- leads with networking for model builders, integrated rack-scale (networking+servers) for sovereign, and full stack for enterprise—networking strengthens HPE’s AI position.

  • Question from Samik Chatterjee (JPMorgan): Networking margin trajectory post-Juniper and cash flow implications?
    Response: Q3 combined Networking OM was 20.8% (Edge 22.7% with sequential pressure from variable comp and product costs); guide is low-20% near term as Juniper runs a few points below Edge; cash flow focus remains high with Juniper costs/OI&E factored—more detail at SAM.

  • Question from Amit Daryanani (Evercore): Growth outlook for combined networking vs. market and Aruba–Mist product integration path?
    Response: Aim to grow above market; will thoughtfully integrate Juniper and Aruba at the AI-ops/cloud management layer while selling both portfolios; data center switching remains Juniper-led; strong security and SDN integration into Hybrid Cloud.

  • Question from David Vaught (UBS): How dependent is AI networking growth on Juniper’s traction, and what drove product costs?
    Response: AI networking opportunity spans service providers, sovereign, and enterprise; Juniper broadens reach via HPE’s footprint; product cost pressure was in Intelligent Edge due to a platform transition.

  • Question from Eric Woodring (Morgan Stanley): End-market demand outlook and HPEHPE-- share opportunities outside networking?
    Response: Demand is robust with consistent order linearity; traditional servers in refresh (Gen12 efficiency/security), disciplined volume/margins; Hybrid Cloud benefits from virtualization transitions and software growth; storage Alletra MP posted third straight quarter of triple-digit growth.

  • Question from Simon Leopold (Raymond James): Update on Juniper’s AI positioning and wins?
    Response: Juniper is becoming the standard above NVIDIANVDA-- Spectrum-X in large AI deployments, with solid backlog; strategy is to lead with networking for AI in service provider/neo-cloud segments and leverage that to pull through rack-scale solutions.

Discover what executives don't want to reveal in conference calls

Latest Articles

Stay ahead of the market.

Get curated U.S. market news, insights and key dates delivered to your inbox.

Comments



Add a public comment...
No comments

No comments yet