HPE reports no customer pushback on rising data center equipment costs as AI workload demands drive hardware upgrades.

The generative AI wave is lifting nearly every player in the market. Although Dell has surpassed Hewlett-Packard to become the world’s largest original equipment manufacturer—a title HP once held following its 2001 acquisition of Compaq, which had previously absorbed Tandem Computers and Digital Equipment Corporation—Hewlett Packard Enterprise remains a top-five vendor and continues to be the server provider of choice for millions of enterprises globally.

Sustained market position matters more in the long term than securing a handful of multi-billion-dollar AI system contracts. Moreover, HPE’s growth rate is already constrained by its ability to secure GPU allocations from Nvidia and AMD. As Dell’s trajectory demonstrates, established leaders are scaling faster.

HPE’s third-quarter fiscal 2026 financial results confirm a turnaround in fortunes. The company is now positioned to accelerate the deployment of AI across diverse compute and networking infrastructures for enterprises, sovereign entities, service providers, and emerging neocloud operators.

For the quarter ended in July, HPE reported revenues of $12.21 billion, a 33.7% increase year-over-year and a 14.4% sequential rise. Driven by disciplined cost management, a higher-margin product mix, and the pass-through of component price increases for CPUs, GPUs, DRAM, flash storage, and networking gear, operating income surged 5.6 times to $1.39 billion. Net income followed suit, rising 5.5 times to $1.51 billion, yielding a robust 12.4% profit margin for a hardware-centric business. This marks one of the strongest quarters in HPE’s datacenter history, with profitability levels comparable only to the Unix systems boom of the late 1980s and early 1990s.

HPE closed the quarter with $6.22 billion in cash and investments, an 11.6% increase year-over-year.

These results align with HPE’s recently restructured organizational framework, which consolidates its operations into distinct business groups. The Cloud & AI group remains HPE’s largest segment, encompassing servers, storage, financial services, and technical support. The Networking group, anchored by the acquired Juniper Networks and supplemented by HPE Ethernet switches—including the Slingshot high-performance computing variant originally developed by HPE’s Cray division—is smaller but expanding at a faster pace.

Cloud & AI revenue grew 25.4% year-over-year to just over $9 billion, with a notable 17.3% sequential increase. This demonstrates HPE’s ability to convert both AI and traditional system backlogs into recognized revenue while those backlogs continue to expand. Operating income within the segment more than tripled to $1.54 billion, marking a significant profitability milestone for the division. Breakdowns within Cloud & AI show server sales jumping 35.3% to $6.77 billion, storage rising 10.2% to $1.29 billion, and financial services holding steady at $883 million. HPE does not report segment-level operating income.

Networking revenue reached $2.89 billion in Q3, surging 74.9% as HPE successfully increases networking attachment rates to its server and storage portfolios—a primary objective behind the Juniper acquisition. Campus and branch switching remains the largest contributor, generating $1.44 billion (up 31% year-over-year) and accounting for just under half of segment revenue. Routing revenue hit $788 million, a 3.7x year-over-year increase, driven largely by service providers seeking alternatives to Cisco Systems infrastructure. Contrary to prevailing narratives, hyperscalers, cloud builders, and neocloud operators continue to purchase substantial volumes of routing equipment to interconnect their datacenter fabrics.

Oracle, which operates its own public cloud, signed a supply agreement with HPE for routers and switches shortly after the close of Q3 FY2026. The majority of the order consists of Juniper QFX switches powered by Broadcom’s 102.4 Tbps Tomahawk 6 switch ASIC, which launched in June of last year and is currently shipping. The agreement also incorporates Juniper’s AIOps software for network management. HPE stated that this contract supports a potential multi-gigawatt datacenter expansion and could extend over several years. As consideration, HPE issued Oracle warrants for up to 4.2 million shares; prior to the announcement, these warrants were valued at over $200 million. Specific vesting terms were not disclosed.

Similar to other industry transactions, equity instruments are utilized to offset hardware procurement costs. The disclosure of such agreements typically enhances share value immediately, benefiting both the vendor and the customer.

Separately, HPE finalized a $3.5 billion agreement with an unnamed hyperscaler. “It is a very important use case,” Antonio Neri, HPE’s chief executive officer, explained on the call with Wall Street analysts going over the numbers. “It is a hyperscaler customer, but think about them as an enterprise customer who is going to use our AI inferencing for their own internal usage. So our strategy has not changed from selling large amount of infrastructure for them to serve, like it used to be in the past the cloud business. This is about a multi-billion-dollar AI inferencing for their own internal usage as an enterprise customer. It just happened to be they are labeled as a hyperscaler customer.”

Returning to HPE’s quarterly performance metrics:

HPE recorded just under $1.6 billion in AI server revenue for Q3 FY2026, a 7% year-over-year increase. New AI system orders totaled $2.4 billion during the quarter. High-performance computing systems appear to be categorized within this segment, given the current convergence of HPC and AI workloads. At quarter-end, the AI server backlog reached $6.8 billion. AI networking orders, tracked separately following the Juniper acquisition, amounted to $700 million, with a corresponding backlog of $800 million. While HPE did not break out AI-specific networking revenue, AI networking orders increased 3.5 times year-over-year.

The traditional, non-AI server segment experienced strong growth, with sales rising 52% to $5.17 billion. Mirroring trends observed at Dell, HPE is benefiting from a natural refresh cycle for legacy ProLiant servers in existing deployments. Replacement units are densely configured with CPUs, DRAM, flash storage, and hard drives. These high-density systems carry inherently higher bill-of-materials costs, justified by superior rack density and power efficiency relative to older architectures, alongside rising prices across virtually all underlying components.

HPE’s datacenter heritage predates both the generative AI surge and its current executive leadership. Our firm has tracked the company for four decades, maintaining a comprehensive dataset of HP and HPE’s core datacenter systems business dating back to the Great Recession, a definitive inflection point in industry history. An examination of the segment’s financial evolution over the past twenty years reveals the scale of recent acceleration.

Generative AI began driving revenue growth in fiscal 2024, a trend further amplified by the integration of Juniper. Consequently, HPE has recorded consecutive record quarters for core datacenter revenue—spanning servers, storage, networking, technical support, and equipment financing—for five straight periods. While HPE trails Dell, Supermicro, Inspur (IEIT), and Lenovo in pure server shipments, the competitive gap narrows considerably when networking and financing revenues are included.

Future revenue growth remains contingent upon GPU supply availability, which in turn hinges on customers’ ability to secure adequate power capacity and datacenter facilities.

Regardless of supply constraints, HPE’s core systems revenue totaled $10.49 billion in Q3, representing a 35.1% year-over-year increase and a 15.6% sequential gain. Based on our operational allocation model, the core systems division contributed $1.86 billion in operating income.

For fiscal Q4, HPE projects total revenue between $13.9 billion and $14.8 billion. The Networking segment is forecast to grow 11% to 13%, while the Cloud & AI segment is expected to surge 60% to 72%. These projections remain heavily dependent on the alignment of facility readiness with GPU supply chains.

According to Neri, enterprise demand remains robust, with organizations actively reallocating and expanding IT budgets at a pace unseen since the dot-com era. Notably, buyers are exhibiting minimal resistance to rising hardware costs.

“At the beginning of this hyper cycle, on the cost, obviously they were a little bit in shock, and they're trying to navigate through that timing by focusing on understanding the trends and looking