[SMM Computing Power Flash] 32 units of H100 spot in Inner Mongolia quoted at 80,000 yuan; 64 units of futures in Xinjiang to go online in August

SMM learned that on July 8, Tencent WorkBuddy’s inference computing resource consumption peaked due to the launch of the Hy3 large model, with the afternoon queuing rate once exceeding 50%, and the team completed emergency capacity expansion that day. This computing resource crunch directly stemmed from the exponential growth in user scale — WorkBuddy’s monthly active users had reached 20 million and daily active users had exceeded 13 million. Unlike the “one-time heavy computation consumption” on the training side, inference computing consumption is linearly linked to the number of users: each interaction by every user generates continuous token consumption, and the daily office scenarios (document processing, code generation, agent tasks) of 13 million users create high-frequency, massive inference computing demands. Data from the OpenRouter platform shows that in May 2026, Chinese models accounted for six of the top ten spots in token calls, and domestic model inference traffic is rapidly catching up with mainstream North American models.

July 8: 32 units of spot H100 in Inner Mongolia quoted at 80,000 yuan per month, 64 units in Xinjiang to go online in August; two large orders of 128 H100 units surfaced, suppliers said "no one in the market can supply"; operators bid 65,000–68,000 yuan per month for H100, a price spread of over 10,000 yuan from the market; RTX 4090 24G traded at 7,000 yuan, near the low end

SMM has learned that two large orders for leasing 128 H100 eight-GPU servers have appeared in the market recently. A buyer in Hangzhou, besides H100, also inquired about higher computing power, with the budget undisclosed; another buyer is negotiating with a supplier in Inner Mongolia. The supplier holds 64 H100 servers spot and offered 80,000 yuan, which the buyer considers high. Both demand orders are for 128 H100 servers, but a certain intelligent computing service provider reported that no single supplier in the current market can deliver 128 H100 servers in a cluster. Meanwhile, industry insiders pointed out that 128 H100 servers far exceed the needs of the commonly used large model DeepSeek V4pro full-blooded version. Based on the above information, SMM believes that there is a significant discrepancy between the reported demand and actual demand in the current market.