What was reported
According to Notebookcheck and Mezha.net, a post on the Chinese hardware forum Board Channels, reproduced by Japanese outlet Gazlog, claims Nvidia told its add-in board (AIB) partners it had abandoned the launch, without providing a revised release window. VideoCardz carries the same claim, saying Nvidia reportedly informed board partners that the RTX 5080 24GB will not launch in January. China's IT Home also repeated it on November 11, attributing it to a Board Channels moderator.
Be careful with the word "canceled." Notebookcheck's headline uses it, but the reporting itself describes a dropped China launch with no new date. That is a narrower claim than a global cancellation. One retailer blog covering the story notes that Nvidia never announced a 24GB RTX 5080 and hasn't commented publicly on the reports.
The reason: pricey 3 GB memory
The forum post blames the cost of denser memory. The 3 GB GDDR7 modules reportedly cost around three times as much as the 2 GB modules used in the existing RTX 5080. Notebookcheck says that makes a 24 GB card difficult to sell at a reasonable price, so Nvidia would keep buying 2 GB modules and keep the 16 GB card on top in China.
The memory layout is simple. The reported 24 GB card would keep the same GPU and 256-bit bus. It would use eight 3GB modules instead of the eight 2GB chips found on the existing model. The capacity jump therefore comes from chip density alone, not from a wider design.
The three-to-one ratio is the forum's figure. None of the coverage gives a supplier quote, a bill of materials or a target retail price, so it should not be converted into dollars.
A rumor that has changed direction
This report contradicts the earlier ones:
- In late October, Board Channels suggested Nvidia was accelerating the RTX 5080 24 GB launch to December 2026 or January 2027 while winding down production of the 16GB variant.
- Gazlog reported that Nvidia's development materials for AIBs, due in October, had not arrived by the end of the month. That put the January date in doubt before the latest post.
- TechTimes also reported that board partners were still missing design kits, and it tied the delay to a wider GDDR7 shortage driven by AI data center demand. That is TechTimes' own analysis, not something confirmed by Nvidia.
Some social-media posts go further and say a "5080 SUPER" is dead worldwide. The search results do not support that. Gazlog says only that a 24 GB RTX 5080 SUPER sample that reached AIBs in July has had shipments held for the same cost reason regardless of region, and that this makes a domestic Japanese launch unlikely. That is Gazlog's assessment, not an Nvidia statement.
What it means for buyers
The most direct effect is in China. If the report is right, shoppers there who were waiting for 24 GB will not get it in January. The 16 GB card stays the top GeForce option.
Outside China, the picture is unclear. Gazlog argues that Nvidia supplies GPU and memory to AIBs as a set, so 16 GB production would likely continue for other markets too, including Japan. That is an inference, not a confirmed commitment. Gazlog also notes that Japanese retail prices for the 16 GB card had started easing before this report, falling from about ¥359,800 on October 31 to about ¥329,800 on November 10. It expects prices to stay well above early-October levels while memory costs stay high.
For anyone weighing a high-end GPU purchase:
- Don't plan around a 24 GB RTX 5080 arriving soon. Even sympathetic reports put it well outside January.
- Treat the "no new date" detail as uncertainty, not as proof that the card is gone for good.
- If 16 GB of VRAM is a limit for your work (large local AI models, heavy texture packs, or creative workloads), the report offers no relief in the near term.
Why the evidence is thin
Mezha and Notebookcheck are not independent confirmations. Mezha credits Notebookcheck, and both trace back to one Board Channels post, relayed by Gazlog. VideoCardz and IT Home repeat the same source. Several outlets repeating one forum claim is still one claim.
The forum has been wrong about direction before, as the October reversal shows. The reasoning is plausible, because AI demand has squeezed memory supply and denser GDDR7 is the obvious pressure point. But the story stays a supply-chain rumor until Nvidia or its board partners say something.
Bottom line
The China launch of an RTX 5080 with 24 GB of GDDR7 is reportedly off, and the existing 16 GB card is reportedly staying. The stated reason is the high cost of 3 GB memory modules. No new date has been given, Nvidia has not commented, and nothing yet shows a global cancellation.