AI server and network demand "explodes" and struggles to resolve supply chain issues Hewlett Packard Enterprise Co. technology (HPE.US) Q3 earnings exceeded expectations but still faced post-market shock.

After the market closed on Wednesday Eastern Time, Hewlett Packard Enterprise Co. (HPE.US) announced its third-quarter results for the period ending July 31. Benefiting from strong demand for AI-related servers and networking equipment, the companys revenue and profits both exceeded market expectations, leading to a significant upward revision of its performance outlook for the current and next fiscal years. However, as management warned that supply constraints continue to limit growth, Hewlett Packard Enterprise Co.'s stock price fell more than 8% in after-hours trading. The financial report showed that Hewlett Packard Enterprise Co. achieved revenue of $12.21 billion in the third quarter, a year-on-year increase of 33.7%, surpassing analysts' previous expectations of $11.9 billion. The adjusted earnings per share were $1.11, well above the market forecast of $0.93. Hewlett Packard Enterprise Co. stated that continuous investments from cloud service providers and enterprise clients in AI, server upgrades, and data center capacity have driven strong demand for the companys high-performance servers and networking equipment. CEO Antonio Neri noted in a statement, Our performance demonstrates the durability of our profitability growth momentum. He mentioned that AI is becoming the growth driver for the company for years to come, and Hewlett Packard Enterprise Co. is well-prepared to seize this opportunity at scale. From a business segment perspective, the networking business performed particularly well. In the third quarter, networking revenue reached $2.89 billion, a year-on-year increase of 75%. Of this, data center networking revenue was $382 million, up 112.2%; routing sales skyrocketed by 270%. The significant growth in the networking business was partly due to Hewlett Packard Enterprise Co.s acquisition of networking equipment vendor Juniper Networks, Inc. (Juniper Networks), which was completed in July 2025. Revenue from cloud and AI businesses grew by 25% year-on-year, reaching $9 billion. Of this, server revenue increased by 35% to $6.8 billion; storage revenue rose by 10% to $1.3 billion. It is understood that Hewlett Packard Enterprise Co. produces traditional data center servers as well as optimized servers equipped with AI chips like NVIDIA Corporation (NVDA.US) GPUs, both of which benefit from the trend of enterprises shifting to AI workloads. CFO Marie Myers stated that customers are upgrading data centers to meet new workload demands, and the adoption of enterprise-grade AI is accelerating. She noted that the companys operating profit for the third quarter exceeded $2 billion, contributing to strong cash flow. We have a timely product portfolio, Myers said, and the positive impact of AI is becoming apparent across the company. She added, With demand continuing to grow, now may be the best time to engage in this business. Supply Chain Bottlenecks: Memory, NAND, CPUs, and Hard Drives in Short Supply Despite strong demand, supply constraints remain a major challenge for Hewlett Packard Enterprise Co. In an interview, Myers stated that supply is still limited, with memory being the primary bottleneck, followed by NAND flash, CPUs, and hard drives. To address this issue, Hewlett Packard Enterprise Co. has signed longer-term supply agreements to improve component availability. Demand far exceeds supply, Myers candidly stated. As of the end of July, Hewlett Packard Enterprise Co.'s inventory was $11.82 billion, a significant increase from $7.16 billion in the same period last year. Myers explained that the rise in inventory reflects increased parts costs, as well as targeted procurement to support a backlog of orders. Given the sustained strong demand for AI, HPE has significantly revised its performance expectations upwards. For the current fiscal year ending October 2026, the company expects revenue growth to reach 34% to 37%, up from the previous estimate of 29% to 33%; the adjusted earnings per share are expected to be between $3.75 and $3.85, compared to the previous forecast of $3.35 to $3.45. Analysts had previously estimated an average revenue growth of around 31% for the fiscal year, with adjusted earnings per share of about $3.45. For the fiscal year 2027 (starting November 1, 2026), HPE anticipates revenue growth of 13% to 17%, an increase from the previous expectation of 8% to 12%. The midpoint of the revenue growth expectation of 15% is above the average analyst expectation of 12%; adjusted earnings per share are projected to grow by 16% to 20%, compared to previous expectations of 12% to 16%. For the upcoming fourth quarter (ending October), HPE expects revenue to be between $13.9 billion and $14.8 billion, with adjusted earnings per share between $1.20 and $1.30, both significantly higher than the analysts expectations of $13 billion and $1.07. Additionally, Hewlett Packard Enterprise Co. announced an expansion of its collaboration with Oracle Corporation (ORCL.US) to help the software giant expand its global AI infrastructure. According to the agreement, Hewlett Packard Enterprise Co. will deploy HPE Juniper networking equipment in Oracle Corporation's AI data centers. At the same time, Hewlett Packard Enterprise Co. granted Oracle Corporation warrants to purchase HPE common stock. Hewlett Packard Enterprise Co. CEO Neri also revealed that after the end of the third quarter, the company signed a $3.5 billion server supply agreement with a large cloud computing company to support the client's internal AI model operations. Neri did not disclose the name of the specific customer. Analysts are optimistic about enterprise-grade AI differentiation. Prior to the financial report release, Deutsche Bank Aktiengesellschaft analyst Gianmarco Conti initiated coverage of Hewlett Packard Enterprise Co. stock with a Buy rating. He described Hewlett Packard Enterprise Co. as a unique countercurrent in the AI server camp. Conti pointed out that unlike competitors such as Dell Technologies, Inc. Class C (DELL.US) and Super Micro Computer, Inc. (SMCI.US), which focus on providing low-margin products for hyperscale data centers, more than two-thirds of HPE's $6.3 billion AI backlog is concentrated in the enterprise and sovereign AI sectors, where its networking and software businesses truly have monetization capabilities, resulting in a more favorable profit structure. He also stated that Hewlett Packard Enterprise Co.'s networking products are the company's profit engine, with its Juniper and Aruba products holding the second position in the campus networking market. Myers also indicated in the interview that enterprise and sovereign clients make up the bulk of Hewlett Packard Enterprise Co.s AI server demand, which is also the area where the company sees the greatest profit opportunities. As of Wednesdays close, Hewlett Packard Enterprise Co. stock was priced at $51.82, having increased approximately 116% year-to-date. In after-hours trading, the stock initially dropped over 3%, but the decline later narrowed to about 1%, essentially flat. Investor concerns over supply constraints somewhat offset the positive impacts of strong performance and optimistic guidance.