- Introduction
- Confirmed vs rumored at a glance
- The RTX 50 lineup you can buy today
- What the RTX 50 Super rumors say
- What we know about RTX 60
- New features are landing on current cards
- Buy now or wait: a decision guide
- Recommended
- Frequently asked questions
We are ready to help
Waiting for RTX 50 Super or RTX 60? What We Actually Know in Late 2026
As of October 30, 2026, NVIDIA has not announced an RTX 50 Super card or an RTX 60 series. Its official desktop GeForce RTX 50 lineup still lists seven models, from the RTX 5050 to the RTX 5090, with no Super variants. Every release date you have seen comes from leaks, and those dates have shifted several times. If you need a PC, plan around your own needs, not a rumored launch window.
Confirmed vs rumored at a glance
Most of NVIDIA's big announcements in 2026 were about software and data center hardware, not new gaming cards. That is where much of the confusion starts.
| Item | Status | Source |
|---|---|---|
| RTX 5090, 5080, 5070 Ti, 5070, 5060 Ti, 5060, 5050 | Confirmed | NVIDIA GeForce product pages |
| DLSS 4.5 with Multi Frame Generation 6X on RTX 50 | Confirmed, January 5, 2026 | CES 2026 |
| DLSS 5 neural rendering, debut in NBA 2K27 | Confirmed, launched September 3, 2026 | NVIDIA, GTC 2026 |
| Rubin platform | Confirmed, for AI data centers | CES 2026 |
| 24 GB RTX 5080 in Q1 2027 | Rumor | BenchLife, October 2026 |
| RTX 50 Super reveal at CES 2027 | Rumor | BenchLife |
| RTX 60 in the first half of 2027 | Rumor | Moore's Law Is Dead |
| Next gaming GPUs (GR20x) not until 2028 | Rumor | kopite7kimi, September 2026 |
The RTX 50 lineup you can buy today
The current generation uses the Blackwell architecture. Memory size is the spec to watch, because it is exactly what the rumored Super refresh would change. Figures below come from NVIDIA's own comparison page.
| GPU | VRAM | Memory bus | CUDA cores | Total graphics power |
|---|---|---|---|---|
| RTX 5090 | 32 GB GDDR7 | 512-bit | 21,760 | 575 W |
| RTX 5080 | 16 GB GDDR7 | 256-bit | 10,752 | 360 W |
| RTX 5070 Ti | 16 GB GDDR7 | 256-bit | 8,960 | 300 W |
| RTX 5070 | 12 GB GDDR7 | 192-bit | 6,144 | 250 W |
| RTX 5060 Ti | 8 GB or 16 GB GDDR7 | 128-bit | 4,608 | 180 W |
| RTX 5060 | 8 GB GDDR7 | 128-bit | 3,840 | 145 W |
| RTX 5050 | 8 GB GDDR6 | 128-bit | 2,560 | 130 W |
One practical tip: the RTX 5060 Ti ships in both 8 GB and 16 GB versions under the same name, so always check the spec sheet of a prebuilt system. If you are torn between the two upper-tier cards, our RTX 5070 Ti vs RTX 5080 comparison breaks down what the extra money buys.
What the RTX 50 Super rumors say
Nearly every leak agrees that Super would not be a new chip. It would be the same GPUs with more memory. In October 2026, Taiwanese outlet BenchLife reported that NVIDIA is preparing a 24 GB RTX 5080 for the first quarter of 2027, replacing the 16 GB model. According to the same report, board partners have only received verbal notice and no development documents. NVIDIA has made no statement.
The math behind 24 GB is simple. In October 2024, Samsung announced 24 Gb GDDR7 chips, which hold 3 GB each. The RTX 5080's 256-bit bus connects eight 32-bit memory chips. Eight 2 GB chips give today's 16 GB; eight 3 GB chips give 24 GB. The CUDA core count stays the same.
That leads to the point many headlines skip. If the leak is accurate, a 24 GB RTX 5080 would not be much faster than the current card in most games. The extra memory matters for local AI models, image generation, 3D rendering and heavily modded games, not for frame rates in a typical shooter.
Why the date keeps moving
The rumor timeline tells the story. In mid-2025, BenchLife expected a refresh in early 2026, right after CES. In early 2026, reports said the series was delayed indefinitely because of memory costs. By June 2026 the expected window had moved to CES 2027, and in October a Q1 2027 date surfaced for a single model.
The reason given is always the same: 3 GB GDDR7 chips are needed by both gaming cards and AI accelerators, and AI demand is absorbing supply. In January 2026, NVIDIA told reporters that GeForce demand is strong, memory supply is constrained, and it continues to ship all GeForce models. That confirms the shortage, but says nothing about Super.
What we know about RTX 60
RTX 60 is just the name people use for the next gaming generation. NVIDIA has not announced it, and there are no official specs. In September 2026, two well-known insiders gave conflicting timelines.
- Moore's Law Is Dead claimed NVIDIA would start launching RTX 60 cards based on the Rubin architecture in the first half of 2027.
- Days later, kopite7kimi said the GR20x Rubin gaming GPUs are a 2028 product and that the launch had slipped again.
Rubin itself is real. NVIDIA unveiled it at CES on January 5, 2026, as a six-chip platform for AI supercomputers, including the Vera Rubin NVL72 rack system. That announcement tells us nothing about when gaming cards will ship.
Also watch the naming. The RTX PRO 6000 Blackwell is an existing workstation card, unrelated to an RTX 60 gaming generation, and "RTX 6090" exists only in forum discussions for now.
New features are landing on current cards
"I will wait for the new features" is a weaker argument in 2026 than in past cycles, because NVIDIA's main graphics updates arrived on RTX 50.
- DLSS 4.5, shown at CES 2026, brought an improved upscaling model and Multi Frame Generation 6X, which creates up to five AI frames for every rendered frame. The 6X mode runs on RTX 50 cards.
- DLSS 5, unveiled at GTC on March 16, 2026, adds neural rendering for lighting and materials. It debuted in NBA 2K27 on September 3, 2026, and at launch works only on RTX 50. NVIDIA says RTX 40 support will follow.
Keep in mind that frame generation makes motion look smoother but does not reduce input latency. Compare cards on native performance first.
Buy now or wait: a decision guide
Waiting without a date has a cost: you keep playing on hardware that may already hold you back. Here is how we would decide.
| Your situation | Our take |
|---|---|
| Your current PC runs your games smoothly | Waiting for official announcements and independent reviews is fine |
| Your PC struggles or cannot run the games you want | Waiting for an unannounced product makes little sense; pick from RTX 50 |
| You need a PC by a set date for work, study or a game launch | Plan around your deadline, not a leak |
| You run local AI and hit the 16 GB VRAM limit | Consider the 32 GB RTX 5090 now, or watch for an official 24 GB announcement |
| You expect older cards to drop in price after Super | Do not build that into your budget; with memory this tight, nobody has promised it |
Choosing a GPU from today's lineup
- Start with your monitor. For 1080p, an RTX 5060 or RTX 5060 Ti covers most games. For 1440p, look at the RTX 5070 and RTX 5070 Ti. For 4K, the RTX 5070 Ti, RTX 5080 or RTX 5090, depending on your frame rate and ray tracing goals.
- Check VRAM. New games with high resolution textures benefit from 12 to 16 GB at 1440p and above. For AI work, memory size matters more than clock speed.
- Prioritize the GPU over the CPU. At high resolutions, the graphics card usually sets the frame rate.
- Size the power supply. The RTX 5080 is rated at 360 W and the RTX 5090 at 575 W.
You can browse ready-to-ship systems in our gaming PC range or spec your own in the HYPERPC configurator.
Recommended
Frequently asked questions We have the answers.
- When is the RTX 5080 Super coming out? There is no official date. BenchLife reports a 24 GB RTX 5080 could arrive in Q1 2027, but NVIDIA has not confirmed it.
- Will the RTX 5080 Super be faster in games? If only the memory changes, as leaks suggest, the gaming uplift should be small. Workloads that run out of 16 GB would benefit most.
- Should I wait for RTX 60? Only if your current PC already meets your needs. Leaked timelines range from early 2027 to 2028, and NVIDIA has confirmed none of them.