Hewlett Packard Enterprise is turning the AI server frenzy into another earnings win, and Wall Street is noticing. The company reported fiscal third-quarter results Wednesday afternoon, Sept. 2, with revenue rising 34% from a year earlier to $12.2 billion, topping the FactSet consensus of $12 billion. Adjusted earnings came in at $1.11 a share, also ahead of the 93-cent estimate. In after-hours trading, the stock slipped about 1%, but that came after a huge run: HPE shares were up 115.8% year-to-date through the close on Wednesday.
The numbers show a company still benefiting from the same data-center spending surge that has powered peers, but with a different mix. HPE’s cloud and AI revenue rose 25% to $9 billion, while server revenue climbed 35% to $6.8 billion. Networking was the standout, jumping 75% to $2.9 billion, with data-center networking alone up 112.2% to $382 million. That combination helped HPE post another quarter of growth that was broad enough to support the company’s most upbeat messaging yet.
For HPE, the AI trade is no longer just about selling more hardware. It is about a portfolio that now looks unusually well timed for enterprise customers that want to build out AI capacity without taking on the same level of risk tied to consumer-facing hype cycles. CEO Antonio Neri framed the quarter as proof that the model is working. “Our results demonstrate the durability of our profitable growth momentum,” he said.
That phrasing matters because investors have spent much of the year trying to separate real AI beneficiaries from companies merely attaching the label to older business lines. HPE’s report gives bulls fresh evidence that demand is still flowing into servers, networking gear and adjacent infrastructure at a pace that can move the entire revenue base. The company also announced an expanded Oracle partnership to deploy Juniper networking in Oracle’s AI data centers, adding another layer to the AI infrastructure narrative.
One reason HPE’s story has sounded different from some of the other AI hardware winners is its customer base. According to Deutsche Bank analyst Gianmarco Conti, who initiated coverage with a Buy, HPE is “the deliberate contrarian of the AI server cohort.” He pointed to the fact that more than two-thirds of HPE’s $6.3 billion AI backlog is tied to enterprise and sovereign customers, which he said have a better profit profile.
That distinction helps explain why the stock has become such a market favorite even after a large run. HPE is not just selling into the hottest part of the AI market; it is selling into a part that appears to be more durable and, at least in the analyst’s view, more attractive on margins. For investors, that can matter more than raw growth alone. It suggests the company may be able to turn demand into earnings without relying on the most speculative parts of the spending cycle.
The $6.3 billion AI backlog has become one of the most important figures in the HPE debate. It gives investors a visible pipeline beyond the current quarter and helps support the case that HPE’s AI momentum is not just a one-time burst. The company’s latest results, combined with the new Oracle partnership, suggest that backlog is still converting into actual revenue. HPE’s raise to its fiscal 2026 and fiscal 2027 outlooks reinforces that view.
Management lifted fiscal 2026 revenue growth guidance to 34% to 37%, and fiscal 2027 growth to 13% to 17%. Those are aggressive numbers for a mature enterprise-tech company, and they signal that HPE expects the current wave of demand to remain powerful well beyond the next quarter. Marie Myers, the CFO, kept the message simple: “We’ve just got the right portfolio at the right time.”
HPE’s year-to-date move is striking even by AI-stock standards. A gain of 115.8% through Wednesday’s close puts the company in rare territory for an enterprise hardware name, especially one that for years was often treated as a slow-moving legacy tech story. The market has clearly re-rated the company as a direct beneficiary of AI infrastructure spending, not a bystander.
That re-rating has been helped by the fact that HPE is showing both growth and scale. Revenue of $12.2 billion is not a small-company number. Nor is the $9 billion in cloud and AI revenue. The market likes AI growth stories most when they come with evidence of repeat demand and operational breadth, and HPE’s latest quarter supplies both. The business mix also suggests the company has more than one lever to pull, with networking adding another growth engine on top of servers and AI infrastructure.
The comparison with Dell is hard to miss, even if HPE’s customer focus differs. The market has been rewarding companies that can translate AI buildout into near-term sales, and HPE now sits squarely in that group. The difference is that HPE appears to be pitching itself as the steadier, more enterprise-heavy version of the trade, not the flashiest one.
The next checkpoint is already visible. HPE’s October-quarter, which is fiscal Q4 2026, is expected to bring revenue of $13.9 billion to $14.8 billion and adjusted EPS of $1.20 to $1.30. Both figures are above the FactSet consensus, giving management another chance to show that the recent strength is still accelerating rather than fading.
That guidance matters because the market has become less forgiving with AI winners that miss the pace of expectations. HPE has avoided that trap so far. It beat revenue and earnings, raised its outlook, and kept the AI backlog story intact. The stock’s slight decline in the extended session suggests some investors may be taking profits after the strong run, but the broader message from the report is still bullish.
For now, HPE has done what the market wants from an AI infrastructure name: it has turned demand into numbers, raised guidance, and given analysts a story that extends beyond the quarter. The question is no longer whether HPE belongs in the AI trade. It is whether the company can keep converting enterprise and sovereign demand into another stretch of outperformance.
