NVIDIA Denies RTX 5090 Discontinuation Rumors and Admits Supply Is Tight
NVIDIA has come out to deny rumors that the RTX 5090 was at the end of its lifecycle. Jeff Fisher, senior vice president of the company’s personal computing business, stated on October 16 in Taipei, following the launch party for RTX Spark architecture systems, that the card has not been discontinued and will continue shipping. The statements, reported by Taiwanese outlet 4Gamers and Central News Agency (CNA) and documented by TechPowerUp, come after a week of speculation about the end of production of NVIDIA’s flagship card.
What Jeff Fisher Said to Taiwanese Media
Fisher was questioned about the end-of-life status of the RTX 5090 at the Taipei event. In the response collected by 4Gamers, the senior vice president acknowledges the underlying problem and simultaneously denies the cutoff:
The supply chain (capacity) is indeed somewhat limited, but NVIDIA is pushing hard to get products to market. Although there are other variants of the RTX 5090 chip, this one has not been discontinued. NVIDIA will continue shipping it, and that is very important.
To CNA he added another argument: supply tension affects not just NVIDIA but the technology industry as a whole, and without the GeForce business the company wouldn’t be where it is today. 4Gamers also reports his connection between both businesses: without GeForce there would be no CUDA, and without CUDA there would be no AI market as we know it.
What the Denial Leaves Unresolved
The statement confirms two things — there is no discontinuation and the company continues shipping the card — but leaves other questions open. Fisher did not clarify whether assemblers are receiving GB202 dies again, which was at the heart of last week’s leaks, nor did he specify what he means by “other variants” of the chip. GB202 is the silicon that also powers RTX PRO 6000 and RTX PRO 5500 workstations, so the phrase admits more than one reading.
There is also no explanation for the facts that fueled the rumor: the RTX 5090 disappeared from NVIDIA’s official marketplace and Micro Center’s catalog, and it had previously been reported that the company stopped assigning GB202 silicon to the card. The denial neutralizes the headline about production ending, not the evidence of availability already visible.
Why the RTX 5090 Remains Far From $1,999
Here is the data that matters to buyers. The recommended price for the RTX 5090 in the U.S. is $1,999, but the card remains very hard to find near that price: according to TechPowerUp, there are listings above $8,000 on Amazon and Newegg. Acknowledging that capacity is limited effectively amounts to admitting that this gap won’t close overnight.
We already reported how first-hand inventory vanished in the United States and how resellers went as high as $9,500 for a unit. Since then the market has not normalized, and with NVIDIA declaring that the bottleneck is real, it’s reasonable to expect scarcity will accompany the card for some time.
What Changes for Buyers
The denial eliminates, for now, the hypothesis of an imminent cutoff of the RTX 5090: NVIDIA assures that the model remains on its shipping line. What doesn’t change is access: continuing to ship does not mean arriving at shelves, and certainly not at catalog price.
For anyone who needs to buy now and doesn’t want to absorb such a pronounced markup, the high-end reference remains the RX 9070 XT, the alternative we’ve been pointing to as the most sensible per-frame price option against an RTX 5090 that is already nearly impossible to buy.
