Nvidia Scraps RTX 5080 Super as 24GB Memory Costs Bite

What Nvidia reportedly told its partners

The RTX 5080 Super is not delayed this time. It's gone.

The leak comes from Board Channels, a Chinese forum that has been a reliable early source for GPU launch schedules, and was picked up by VideoCardz. The claim is that Nvidia informed graphics card manufacturers in China that the January 2027 launch is cancelled, and gave them no clear follow-up schedule. That wording matters: a week earlier the same rumor mill was talking about a delay, not a cancellation.

Nvidia has not commented publicly, so treat this as an unconfirmed report rather than a press release. But the direction of travel is consistent with everything else happening in the GPU market right now.

RTX 5080 Super: what was rumored vs. the RTX 5080

Why 24GB of VRAM became the problem

The Super was supposed to use 3GB GDDR7 memory modules instead of the 2GB chips in the standard RTX 5080. Those bigger modules are now dramatically more expensive.

That single component swap is what turns a 16GB card into a 24GB card, and it's also what killed the project. Memory prices have climbed hard over the past year as AI data center demand soaks up supply, and the premium on 3GB GDDR7 modules apparently makes a mainstream-priced Super card impossible to build at a sensible margin.

The irony is sharp. The RTX 5080 Super existed to answer the most common complaint about the base card: 16GB of VRAM feels thin for a GPU in this price bracket, especially with games and AI workloads getting hungrier. Nvidia's fix was to add memory. The cost of that memory is now the reason the fix isn't happening.

Gamers are running out of patience

The reaction online has been less about one cancelled SKU and more about a pattern. The RTX 50 series launched with criticism over pricing and VRAM tiers, and 2026 has brought a steady drip of bad news: cards disappearing from prebuilt systems, retail prices climbing well above MSRP, and now a refresh that was supposed to be the good-news story getting axed before it shipped.

The mood is exhaustion, not outrage. As one common refrain puts it, "I wish they'd just do something." That's the sentiment: fatigue at a consumer GPU market where the only movement is upward on price and downward on choice. The RTX 5090 shortage has already shown how ugly that can get.

What this means for the RTX 5080 and the rest of the lineup

In China, the plain 16GB RTX 5080 reportedly stays the flagship, since the RTX 5090 isn't sold there at all.

For everyone else, the practical result is a lineup that doesn't change. If you were holding off on a GPU purchase waiting for a 24GB RTX 5080 Super in early 2027, that plan is now dead, and the alternatives are the existing 16GB RTX 5080 or stepping up to much more expensive cards.

The broader worry is what this says about the next generation. If memory costs are high enough to cancel a mid-cycle refresh, they're high enough to shape pricing on whatever comes after the RTX 50 series. Nvidia's consumer division is increasingly competing with its own data center business for the same memory supply, and the consumer side keeps losing.

Availability and pricing

Nothing changes on store shelves in the US, UK or EU, because the RTX 5080 Super never got a price or a release date. The card that exists, the 16GB RTX 5080, remains on sale at whatever your local retailer is charging, which in recent months has often been well above the official MSRP.

The cancellation is a rumor at this stage. Nvidia has not confirmed it, and there's no official statement on whether a 24GB consumer card is planned for a later date. If you're building a PC in the next six months, plan around the hardware that's actually available.

A cancelled refresh is a small story on its own. Taken with rising GPU prices and shrinking retail availability, it's a clear signal that the consumer graphics market is being squeezed by forces gamers can't control. The RTX 5080 you can buy today may be the RTX 5080 you're stuck with for a while.