2026-01-16
HPE Server Shipments Continue to Bottom Out, Forecast to Drop to
Core Overview: Shipments Fall Below the 260,000 Mark, Hitting a Recent Low
According to the latest data, HPE (Hewlett Packard Enterprise) server shipments in the third quarter of 2026 are estimated to be only 255,000 units (255.0 K pcs), a decline of approximately 3% from the previous quarter's 263,000 units. Compared to the 270,000 units in the same period of 2025, the drop is even more significant at 5.5%. This continues the trend of sliding from a quarterly level of over 300,000 units since 2024, indicating that in terms of shipment "volume," HPE is facing structural contraction pressure, forming a sharp contrast with revenue performance hitting new highs due to the AI wave.
Key Details: Quarterly Decline in 2026, Weak Demand for General Purpose Servers
Observing the forecast for the full year of 2026, shipment volumes show a clear trend of quarterly decline: Q1 is estimated at 272,000 units, Q2 drops to 263,000 units, and Q3 further slides to 255,000 units. Compared to the first half of 2024, which maintained a level of 310,000~320,000 units, current forecasts suggest that the refresh cycle for traditional General Purpose Servers may be slowing down. Additionally, data shows a brief rebound to 281,000 units in the fourth quarter of 2025, but the momentum was not sustained, implying that market demand returned to a wait-and-see stance after a short-lived restocking.
In-Depth Attribution: AI Crowding-Out Effect and ODM Competition
Analysts point out that the main reason for the decline in HPE shipments lies in a drastic shift in market structure. First, Hyperscalers prefer to purchase directly from ODM manufacturers such as Quanta and Wiwynn, leading to a loss of market share for branded manufacturers in the cloud infrastructure market where volume is highest. Second, although HPE has made significant gains in the AI server field (such as those equipped with NVIDIA Blackwell/Hopper chips), boosting revenue and gross profit, the unit price of AI racks is extremely high and deployment density is great. With limited IT budgets, enterprises often choose to "spend heavily on AI while cutting General Purpose Server quantities," resulting in a divergence phenomenon of "hot revenue, cold shipments."
Outlook and Risks: Supply Chain Constraints and Transformation Pains
Short-term (1-2 months): Constrained by tight supplies of high-end AI chips, although HPE holds a large Backlog, the actual speed of shipment ramp-up may be limited by supply chain bottlenecks, making it difficult for short-term shipment numbers to rebound significantly.
Medium-term (3-6 months): As enterprises accelerate the adoption of GreenLake hybrid cloud services, HPE will continue to optimize its profit structure, shifting from "selling hardware" to "selling services." While this helps profitability, it means the downward trend in pure hardware shipment volumes may be difficult to reverse. Investors should pay closer attention to whether the revenue share of AI servers can effectively offset the decline in traditional computing units.
Web Search References
https://www.techmonitor.ai/hardware/servers/hpe-surpasses-forecasts-in-q4-fy24-driven-by-ai-powered-server-demand
https://www.investing.com/news/stock-market-news/hpe-flags-impact-of-delayed-ai-server-revenues-sending-shares-lower-3759363
https://www.ciodive.com/news/dell-hpe-ai-server-revenue-earnings/726117/