What happened
IDC reported that global server vendor revenue reached an all-time high of $166.3 billion in Q2, marking a 52% year-over-year increase. Despite rising memory costs and supply issues that pushed average selling prices up significantly, server unit shipments grew by 15.4%. The report highlights a shift in demand drivers, noting that AI infrastructure investment is broadening beyond major hyperscalers to include specialized cloud providers, sovereign AI programs, and enterprises adopting agentic and inferencing workloads.
According to market intelligence firm IDC, the second quarter was a record-breaking period for the server sector, with vendor revenue reaching an all-time high of $166.3 billion. This represents a 52 percent increase from the same period last year. The growth occurred despite average selling prices being pushed up by elevated memory pricing and continued supply issues with other components.
Server shipments increased by 15.4 percent year-on-year in Q2. Average selling prices for GPU-accelerated servers rose by nearly 44 percent to $170,200, even as GPU unit shipments fell 10.8 percent year-on-year. For non-accelerated systems, average pricing was up by more than 33 percent to nearly $13,000. IDC noted that GPU-accelerated servers for the AI market made up nearly 53 percent of total revenue during Q2.
IDC research vice president Kuba Stolarski stated that the notable shift in the server market this quarter is in who is now buying. He explained that demand is broadening beyond the largest hyperscalers toward specialized cloud providers, sovereign AI programs backed by public capital, and enterprises beginning to adopt agentic and inferencing workloads. This policy and capex-driven layer of demand is largely insulated from near-term commercial budget cycles.
The report also highlighted changes in market structure. Non-x86 servers now account for 44.8 percent of all server market revenue, a share that has fallen from the first quarter despite actual revenue rising from $58.7 billion to $74.4 billion. Additionally, big brands are eating into the share of original design manufacturers (ODMs). ODMs' share fell from over 60 percent last year to 53.9 percent in Q2, with Dell Technologies leading the charge as its share rose from 7.7 percent to 13.4 percent.
Source details: theregister.com ↗
Why it matters
This data confirms that AI infrastructure demand is no longer limited to a few tech giants but is becoming a broader economic force involving government and enterprise sectors. The resilience of server sales despite higher component costs indicates strong underlying demand for AI capabilities. This shift suggests that AI adoption is becoming a standard operational requirement across various industries and national strategies, rather than a niche experiment, which has significant implications for long-term hardware supply chains and market competition.
The expansion of AI server demand to enterprises and governments indicates that AI infrastructure is becoming a critical component of national and corporate strategy, not just a competitive advantage for tech giants. This diversification of buyers reduces the market's dependence on the capex cycles of a few hyperscalers.
The significant rise in average selling prices, particularly for GPU-accelerated systems, reflects the high cost of AI hardware and the supply constraints in the memory and component markets. This suggests that AI infrastructure remains a capital-intensive sector with high barriers to entry for smaller players.
The shift in market share from ODMs to established brands like Dell and Supermicro indicates a potential consolidation or re-evaluation of supply chain relationships. As AI workloads become more complex and integrated into enterprise IT, buyers may be prioritizing the support and integration capabilities of major vendors over the cost-efficiency of white-box solutions.
What to watch next
Monitor whether enterprise and government AI spending continues to outpace hyperscaler growth in subsequent quarters. Watch for further shifts in market share between traditional server brands and ODMs, as well as the impact of memory pricing on smaller vendors. Additionally, track the specific adoption rates of agentic AI workloads in enterprise environments to gauge the practical deployment of these technologies.
Future IDC reports will be crucial to see if the broadening of demand to sovereign AI programs and enterprises sustains the 52% revenue growth rate or if it moderates as initial deployment phases complete.
The impact of memory pricing on PC and server markets will continue to be a key factor. If memory shortages persist, it could further drive up costs for AI infrastructure, potentially slowing adoption among smaller enterprises.
The competitive dynamics between ODMs and traditional server brands will be a key indicator of how the AI hardware market matures. If ODMs can regain share, it may indicate a return to cost-focused procurement; if brands maintain gains, it suggests a premium on integrated AI solutions.