RTX 5080 Pulled From NVIDIA Store as Cheapest In-Stock Hits $2,199.99

NVIDIA's online store has dropped the GeForce RTX 5080's listings and its search filter, Tom's Hardware reported on Oct. 30, while the cheapest in-stock card it found was listed at $2,199.99, 2.2 times the $999 launch price.

The removal follows our Oct. 29 analysis of the card's pricing, which found the card listed at $1,999. It changes the question from what the card costs to whether NVIDIA still sells it through its own channel. Notebookcheck reported the same removal on Oct. 31. Voltgrid did not check NVIDIA's store page itself, and NVIDIA has not commented on the RTX 5080's status.

What NVIDIA's Store Still Lists

The RTX 5080 now matches the RTX 5090, which Tom's Hardware found missing from NVIDIA's marketplace by Oct. 17. Lower Blackwell cards are treated differently. The RTX 5070 Ti and RTX 5070 are still listed, marked out of stock, according to both outlets. Tom's Hardware suggests that keeping sold-out cards visible signals restocks, so pulling the two top cards breaks that pattern. Both outlets stress that a missing listing does not by itself mean the card is discontinued.

RTX 5080 Prices at US Retailers

Tom's Hardware's cheapest in-stock RTX 5080 on Oct. 30 was a PNY GeForce RTX 5080 ARGB Epic-X RGB OC at $2,199.99. Notebookcheck cites its own survey, conducted early this week, of US retailers including Best Buy, Micro Center, Newegg and Central Computers that found offers from $1,979.99, close to the RTX 5090's $1,999 launch price. The two figures come from different days and retailers, and stock is falling between them. Tom's Hardware says the card is unlikely to be available by Black Friday.

The RTX 5080 pairs NVIDIA's GB203 graphics chip with 16 GB of GDDR7, the graphics memory soldered around the chip. Notebookcheck says the rise has been attributed to limited supply, strong demand and continuing memory shortages, and notes that board partners have said little about supply. Our earlier reading still holds as an inference, not a reported fact: the same memory makers are steering DRAM capacity toward high-bandwidth memory (HBM) for AI accelerators, and AMD is reportedly preparing a second Radeon price rise on memory costs. For anyone who buys gaming cards as cheap local inference hardware, the 16 GB card at Tom's Oct. 30 price costs more than twice its launch price.

What Separates the RTX 5080 From the RTX 5090

Graphics-card memory speed comes from the width of the connection to the memory chips and the speed of each wire. NVIDIA's Blackwell whitepaper lists the RTX 5080 with 16 GB of GDDR7 on a 256-bit interface at 30 Gbps per pin, or 960 GB/s. The RTX 5090 has 32 GB on a 512-bit interface at 28 Gbps, or 1,792 GB/s.

The wider interface needs a bigger chip: the RTX 5090's GB202 die measures 750 mm2 against 378 mm2 for the RTX 5080's GB203, and carries 21,760 CUDA cores to 10,752. Tom's Hardware reports rumors, unconfirmed by NVIDIA, that GB202 silicon is being redirected to RTX Pro cards, and notes that even a rumored 24 GB RTX 5080 would not be a drop-in 5090 replacement because of its far lower core count.

What NVIDIA Has Not Confirmed

Notebookcheck says NVIDIA has yet to confirm the status of either card, and Tom's Hardware said it asked the company about the RTX 5080. On the RTX 5090, NVIDIA senior vice president Jeff Fisher reportedly said production has not been discontinued and shipments will continue, though supply is somewhat constrained, according to Taiwanese reports including 4Gamers cited by Tom's Hardware. Reports of a 24 GB RTX 5080 replacing the 16 GB model remain rumors; NVIDIA has announced no such card. Whether board partners are still building the RTX 5080, and whether NVIDIA's stores in other regions list it, has not been verified. The next dated signal is NVIDIA's third-quarter fiscal 2027 results call on Nov. 17, per the company's Oct. 28 release; whether it addresses GeForce supply is open.