Nvidia Reportedly Halts GeForce RTX 5090 Production to Prioritize AI and Professional
Key Takeaways
- •Leakers MEGAsizeGPU and hongxing2020 claim a factory notice dated around October 10, 2026 marks the end of RTX 5090 production, with all future GB202 silicon allocated to RTX Pro cards.
- •The RTX Pro 6000 shares the same GB202 die as the RTX 5090 and sells for approximately $15,000 with up to 96 GB of memory, compared with the gaming card's $1,999 launch price and 32 GB of memory.
- •RTX 5090 secondary market prices now exceed $5,000, with some listings reaching $9,500, fueled by growing demand for AI inference.
- •The leaked notice reportedly points to a new 24 GB RTX 5080, up from the current model's 16 GB, which would assume the top position in Nvidia's consumer lineup.
- •Nvidia had not confirmed the production halt as of October 12, 2026, despite broad coverage of the rumor, and has previously denied similar claims.

Nvidia's most powerful consumer graphics card may be on its way out. According to hardware leakers MEGAsizeGPU and hongxing2020, the company has ended production of the GeForce RTX 5090 in order to redirect its GB202 GPU die toward professional workstation and AI data center products.
Nvidia has not confirmed the report. Even so, the claim follows a familiar pattern for PC gamers: when AI demand competes with gaming hardware, AI demand tends to win. The most recent precedent was the 2020–2022 cryptocurrency mining boom, when competing demand from miners left popular GPU models scarce for long stretches.
What the leakers claim
The report stems from what is described as a factory notice dated around October 10, 2026. According to the leakers, that notice marks the end of RTX 5090 production.
Under the reported plan, all future GB202 silicon would be allocated to RTX Pro cards—Nvidia's workstation lineup, built for professional users and increasingly purchased for AI workloads. The redirect is possible at the silicon level because the RTX Pro 6000 is built on the same GB202 die as the RTX 5090: wafer capacity allocated to one product line is capacity the other cannot use.
The leaked notice also reportedly points to a new 24 GB RTX 5080—a step up from the 16 GB on the current model. That card would assume the top position in Nvidia's consumer lineup should the RTX 5090 be discontinued.
As of October 12, 2026, Nvidia had not confirmed the production halt, despite broad coverage of the rumor. Several outlets have noted that the company has previously denied similar claims.
The economics behind the move
The RTX 5090 launched in January 2025 at a retail price of $1999, pairing its GB202 die with 32 GB of memory. Secondary market prices now exceed $5,000, with some listings reaching $9,500. That surge has been fueled by growing demand for AI inference—the work of actually running trained AI models.
By comparison, the RTX Pro 6000 sells for approximately $15,000 and offers up to 96 GB of memory. RTX Pro cards such as the RTX Pro 5500 and RTX Pro 6000 can fetch prices of up to $15,000. Set against a $1,999 list price, the incentive to reroute silicon toward professional products is difficult to overlook.
What it means for gamers and buyers
If production has indeed stopped, the existing pool of RTX 5090 cards becomes finite. Supply constraints are expected given heavy demand for high-end GPUs in AI applications, and those constraints could push prices higher still.
The rumored 24 GB RTX 5080 could offer some relief, giving enthusiasts a new flagship option—though one with less memory than a card positioned for AI work would typically carry.
The key factor to watch is confirmation. Until Nvidia addresses the report directly, it remains a leak backed by a reported factory notice, not an official product decision.