[SMM Computing Power Flash] Brokerage Accounts for 30% in Computing Power Trading Groups, 40 Out of 120 Members Are Intermediaries

Published: Jul 28, 2026 11:58

Save

SMM learned that on July 28, a personnel structure survey was conducted in a computing power industry WeChat group. The group had about 120 members, among whom around 40 were intermediaries (brokers/matchmakers), accounting for roughly 30%. This means that about 3 out of every 10 members focused on deal matchmaking as their core business, highlighting the prominent layering of intermediaries in the computing power leasing market. The high density of intermediaries causes the same batch of supply to pass through multiple middlemen, widening the price spread between the end-user quotation and the actual owner's price. Consequently, the difficulty for buyers to identify first-hand supply sources increases. The 30% intermediary ratio confirms a high degree of information asymmetry in the current computing power leasing market, with transaction costs continuously rising due to the accumulation of intermediate links.

Data Source Statement: Except for publicly available information, all other data are processed by SMM based on publicly available information, market communication, and relying on SMM's internal database model. They are for reference only and do not constitute decision-making recommendations.

For any inquiries or for more information, please contact: lemonzhao@smm.cn

For more information on how to access our research reports, please contact:service.en@smm.cn

Related News

4 hours ago

[SMM Computing Power Midday Review] A800 regional divergence intensifies: Yangtze River Delta 40G surges, Beijing-Tianjin-Hebei 80G bucks the trend with concessions

The highest monthly rental price for A800 40G GPUs in the Yangtze River Delta surged to 20,000 yuan (up 21.21%), with the highest hourly rate reaching 3.47 yuan, driven by the clustering of the AI industry combined with rigid energy costs; the lowest monthly rental price for A800 80G GPUs in the Beijing-Tianjin-Hebei region dropped to 31,000 yuan (down 3.13%), as clean energy from western Inner Mongolia and supply growth from the "East Data, West Computing" project create a pricing advantage. Regional price differentials reflect a structural trade-off between "proximity to demand" and "proximity to power.

4 hours ago

22 hours ago

[SMM Computing Power Daily] 910B2 large-scale supply in the Yangtze River Delta, monthly rental 12,000, 128 units of H20 shipped outside China for 50,000

July 27: 28 units of 910B2 eight-card in the Yangtze River Delta are being released in batches, with monthly rent of 12,000 yuan and an inventory cycle of about half a month; 128 units of H20 141G are deployed outside China, with tax-inclusive monthly rent of 50,000 yuan, one-year closed contract; 5090 five-year closed contract at 10,500 yuan/month, delivery in August from Yunnan and Sichuan; RTX 50-series capacity is being squeezed by AI, 3060 production has been rescheduled, and uncertainty in 5090 futures delivery has increased; in Zhejiang, 64 units of A100 40G eight-card are being released in volume, requiring an annual contract at a monthly rent of 16,500 yuan, supporting NVLink interconnection; a certain second-hand equipment dealer has no spare H100 resources and will release in October.

22 hours ago

Jul 27, 2026 11:42

[SMM Computing Power Newsletter] Yangtze River Delta 28 Ascend 910B2 Units Re-leased in Batches, Monthly Rent 12,000, Inventory Only Half a Month.

SMM learned that the Yangtze River Delta region has recently released 28 Ascend 910B2 8-card server rental resources in batches. The first batch of 20 units was deployed initially, followed by a second batch of 8 units to replenish inventory, with a monthly rental price of 12,000 yuan and a current inventory cycle of only about half a month. Market observers noted that this price is highly competitive in the 910B2 rental market of developed regions, and the fast inventory turnover reflects a recent resurgence in demand for domestic computing power in inference and lightweight training scenarios.

Jul 27, 2026 11:42