Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Qualcomm Moderated Thread - please read rules before posting -- Ignore unavailable to you. Want to Upgrade?

To: waitwatchwander who wrote (196260)	10/28/2025 9:47:32 AM
From: kech	1 Recommendation Recommended By Dr. John Read Replies (1) \| Respond to of 196287

WWWander - I asked Chatgpt your question and here is its answer: How does the order from Humain for Qualcomm chips get translated to a 1 billion dollar order? ChatGPT said: Here’s a breakdown of how the deal between Humain (a Saudi-Arabia–backed AI infrastructure company) and Qualcomm Incorporated can be translated into “a $1 billion-plus order” (or multi-billion dollar revenue opportunity) and why analysts are treating a “200 MW deployment” as effectively a large-scale commitment. ? What we know Qualcomm announced that Humain will be its first customer for Qualcomm’s new AI-inference chips/racks (AI200 and AI250) and will deploy up to 200 megawatts of compute infrastructure. Reuters+2Qualcomm+2 Analysts estimate that this deal could translate into $2 billion in revenue for Qualcomm from this customer relationship. TipRanks+2Barchart.com+2 Qualcomm is positioning this as a high-volume “rack-scale AI inference” business, meaning many racks, cards, components over many years. South China Morning Post+1 ?? How a “200 MW deployment” becomes ~$1 billion+ Here’s the logic behind taking a “200 MW deployment” and arriving at a large dollar figure: 200 megawatts of data-center capacity implies a very large number of server racks or AI accelerator systems. It’s a scale metric commonly used in hyperscale / cloud data-centre buildouts. Each rack-scale system contains multiple AI inference cards or modules, plus supporting infrastructure (chassis, cooling, networking, power supplies, integration services) which all add cost. If you assume each rack might cost hundreds of thousands to over a million dollars, then across hundreds or thousands of racks the total can escalate quickly. Component cost mix: The order is not just “chips” but includes the accelerator cards, full rack systems, integration, cabling, power & cooling deployment, installation, and probably service/maintenance contracts. Multi-year deployment: The order may not be “one payment today” but spread over years (2026, 2027 etc). That means the “200 MW” is a commitment that unlocks revenue across the lifecycle. Analysts often estimate future revenue accordingly. Margin & recurring revenue: Beyond equipment, Qualcomm may derive recurring revenue (support, upgrades, new releases) which magnifies the total contract value further. Analyst estimates: As noted, one estimate sets ~$2 billion in revenue from this deal. TipRanks So the “$1 billion” figure is a conservative shorthand or early-stage valuation of the contract’s size (or portion thereof). ?? Summary The headline “200 MW deployment” is a large scale commitment which, when translated into equipment cost + integration + services over time, easily becomes a $1 billion+ order. The actual dollar figure depends on assumptions about rack cost, number of racks, cost per watt of infrastructure, deployment timeline, etc. Analysts believe the value is closer to $2 billion for Qualcomm from this one customer. It’s not simply “chips × price” — it’s an ecosystem deal (racks, cards, infrastructure, services, deployment) at scale. If you like, I can pull together publicly-reported assumptions (e.g., cost per MW of rack-scale AI deployment) and produce a back-of-the-envelope model showing how 200 MW ? ~$X billion for this type of deal.