Chinese AI and cloud companies including Alibaba, Tencent and ByteDance were expecting to receive roughly 1% of their Nvidia RTX Pro 5500 orders during a four-month period starting in early November 2026, according to people familiar with the matter who were not named.
Those people put total orders at about 6 million GPUs, with a combined estimated value of $70 billion and prices exceeding $11,000 per chip. They described anticipated deliveries of tens of thousands of units during that period. The figures represent buyers’ expectations, rather than a shipment commitment announced by Nvidia.
An Nvidia spokesperson declined to comment on shipment numbers and criticized “outdated U.S. export controls, which cover gaming products released nearly half a decade ago” alongside “China’s own limits on U.S. imports.”
Delivery estimates extend into 2027
The unnamed people said Chinese developers expected an initial batch of several thousand GPUs by the end of 2026, followed by tens of thousands around February 2027.
A longer-term estimate put potential deliveries at roughly 700,000 units by the end of 2027. That forecast came from one person citing a major buyer’s internal calculations and assumed Nvidia would increase production. Final shipment volumes could change depending on the company’s capacity adjustments.
The people also said Nvidia had warned customers of limited capacity for the foreseeable future, citing product scheduling conflicts.
Earlier, two people familiar with the matter had described orders from ByteDance, Alibaba and Tencent for more than 3 million cards, valued at over $35 billion, in October 2026. Those figures concerned the estimated value of orders. They did not establish how much hardware had been delivered.
A workstation GPU for AI workloads
Nvidia lists the RTX Pro 5500 Blackwell Workstation Edition with 84GB of GDDR7 memory and error-correcting code support. The company describes it as a GPU for rack-mounted workstations and identifies large language model inference, generative AI, simulation and graphics among its intended workloads.
