Alibaba, Tencent, ByteDance and other Chinese AI and cloud firms have ordered about 6 million of Nvidia’s RTX Pro 5500 chips, Bloomberg reported on 4 November, citing people familiar with the matter. The companies expect to get “only some tens of thousands, or roughly 1%, within the next four months”, according to the report.
Bloomberg’s headline values the order at $70 billion. One per cent of that is about $0.7 billion of chips, by our arithmetic: 5.6 per cent of the $12.4 billion Nvidia booked from China-headquartered customers in the first half of its fiscal year.
Business Pill · Backlog
An evergreen Business Engineer explainer on backlog: orders signed but not yet delivered or earned. It teaches the general idea only and does not discuss Nvidia, the Bloomberg report or any company in this story.
The key insight: Bloomberg reports that Chinese AI and cloud firms have ordered about 6 million Nvidia RTX Pro 5500 chips and expect roughly 1% within four months. At the $70 billion in Bloomberg’s headline, that is about $0.7 billion, by our arithmetic: 5.6 per cent of Nvidia’s $12.4 billion of first-half China revenue.
The report was published at 03:14 UTC under the headline “Nvidia China Clients to Get 1% of $70 Billion RTX Chips Ordered”. Its opening says the firms “have in total ordered about 6 million of Nvidia’s RTX Pro 5500 chips”.
The page was blocked to us, so we read only the headline and the opening two sentences, as indexed by a search engine. Crypto Briefing, relaying the report at 03:21 UTC, wrote that “Nvidia’s Chinese clients will receive 1% of a $70 billion order for RTX chips”. Wall St Engine posted the headline on X at 03:15 UTC.
The value of RTX Pro 5500 chips Chinese firms have ordered ($70 billion, the figure in Bloomberg’s headline, for about 6 million cards) against Nvidia’s revenue from customers headquartered in China, including Hong Kong ($12.4 billion in the first half of fiscal 2027; $7.9 billion in the quarter to July), and the roughly 1% the buyers expect within four months. The $0.7 billion and the 5.6% are our arithmetic; a list-price chip value is not Nvidia revenue and the periods differ. Data: Bloomberg report, 4 Nov 2026, citing people familiar with the matter; NVIDIA Form 10-Q for the quarter ended 26 July 2026.
What Reuters Reported in October
On 20 October Reuters reported that ByteDance, Alibaba and Tencent had ordered more than 3 million RTX PRO 5500 cards, with a collective sticker price of over $35 billion, at about $11,400 each. Two sources told Reuters that total orders from customers in China had exceeded 6 million.
One source told Reuters that Nvidia said partners “might receive only five per cent to 10 per cent of the quantities requested”. Another said Nvidia had indicated it could produce no more than 3 million of the cards globally in 2027.
Reuters described the RTX PRO 5500 as “a workstation chip, designed for powerful professional computers rather than the large data-center clusters that use Nvidia’s most advanced AI processors”. The sources said Chinese companies want the cards mainly for AI inference.
At Reuters’ per-card price, 6 million cards come to about $68.4 billion, by our arithmetic, close to Bloomberg’s $70 billion. We could not read how Bloomberg reached its figure.
Against Nvidia’s China Revenue
Nvidia’s Form 10-Q for the quarter to 26 July 2026 counts revenue by where direct customers are headquartered. China, including Hong Kong, was $7,880 million in that quarter and $12,430 million in the first six months of fiscal 2027, down from $13,644 million a year earlier.
The $0.7 billion at Bloomberg’s headline value would be 5.6 per cent of that first-half figure and 8.9 per cent of the July quarter, by our arithmetic. The comparison has limits: a list-price value of chips is not Nvidia’s revenue, and four months is not six.
As we read the two reports, the October orders and the November expectation describe the same queue from two ends: Reuters reported what was ordered, Bloomberg reports what buyers expect to receive soon.
As we read Reuters, supply is the stated constraint: one source said Nvidia indicated it could make no more than 3 million of the cards globally in 2027, against more than 6 million ordered from China.
Reuters, 20 October 2026, citing three people familiar with the matter
“Nvidia has told customers and partners it cannot guarantee that it can fill the orders because demand from Chinese technology companies far exceeds planned production”
Three Implications
FOR NVIDIA’S CHINA LINE The 10-Q puts first-half revenue from China-headquartered customers at $12,430 million, down from $13,644 million a year earlier. The roughly 1% Bloomberg reports would be about $0.7 billion at list value, by our arithmetic.
FOR CHINESE BUYERS Reuters reported the cards are wanted mainly for AI inference. Bloomberg reports buyers expect some tens of thousands of about 6 million ordered within four months.
FOR THIS WEEK’S CHINA THREAD The FT’s 51 per cent rule concerns cloud providers’ clients; Bloomberg’s report concerns chips ordered by Chinese firms. The readable text does not link the two.
The framework puts it this way: “That is why the AI economy is not a list of companies racing each other. It is a layered industrial stack — and at any given moment, the binding constraint sits at a different layer than it did six months ago.”
As we read it, the reports place the constraint for these buyers at supply of one export-compliant card, as Reuters and Bloomberg describe it. The framework was published in May 2026 and does not discuss these reports.
The part of the Bloomberg report we could read does not say why deliveries are so small, whether export licences play a part, or which firms ordered how many chips. It does not give a comment from Nvidia or the buyers. The roughly 1 per cent is what the buyers expect, as people familiar with the matter told Bloomberg; it is not a statement by Nvidia in the part we read.
Reuters’ 5 to 10 per cent and Bloomberg’s roughly 1 per cent are different measures from different sources: one is what partners might receive, with no time frame given, the other is what buyers expect within four months.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
Bloomberg reports that Chinese firms have ordered about 6 million RTX Pro 5500 chips and expect roughly 1 per cent within four months. At its $70 billion headline value, that is about $0.7 billion, by our arithmetic, against $12.4 billion of first-half China revenue in Nvidia’s 10-Q.
A note on sourcing. The Bloomberg report is paywalled and its page was blocked to us; we read only its headline and opening two sentences, as indexed by a search engine, and Crypto Briefing’s relay. Order and allocation figures are as people familiar with the matter told Bloomberg and Reuters; we have not verified them independently. China revenue is from Nvidia’s Form 10-Q for the quarter ended 26 July 2026. The $0.7 billion, $68.4 billion and all ratios are our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
