The RTX 5080 Super and the rest of the RTX 50 Super models are reportedly fully developed and, in some cases, already with board partners—yet Nvidia is postponing the launch indefinitely. The new memory is to blame: The higher-density 3-GB GDDR7 chips currently cost about three times as much as the previous 2-GB modules, driving up production costs per card by several hundred euros. An official release date therefore remains up in the air.
Key Points at a Glance
- According to leaks, Nvidia’s RTX 50 Super series (including the RTX 5080 Super, RTX 5070 Ti Super, and RTX 5070 Super) is ready and has already been shipped to some board partners.
- Nevertheless, the release has been put on hold due to the sharp rise in prices for 3-GB GDDR7 memory.
- A 3-GB chip currently costs around $60 to $70, while a 2-GB module costs only about $20—meaning the 3-GB chip costs roughly three times as much for just 50 percent more capacity.
- The scarce chips are primarily being allocated to NVIDIA’s highly profitable AI hardware, such as the RTX PRO 6000 Blackwell.
- In addition, significantly higher power consumption is expected: The RTX 5080 Super is said to draw up to 415 watts.
- All information is based on leaks and unofficial sources. Nvidia has not yet made an official statement.
RTX 5080 Super: Hardware Ready, Sales Halted
The story of the Super models has taken several twists and turns recently. At the beginning of the year, it was reported that there would be no new offshoots of the Blackwell generation at all. Then, in June, it was leaked that the variants with more memory would be coming after all. Now the next twist: According to VideoCardz, the cards are ready, and at least one board partner is said to already have the hardware in hand—but sales have been put on hold.
The reason is purely economic. The new models rely on higher-density GDDR7 memory chips with three gigabytes per chip instead of the previous two. This allows for more video memory without having to change the number of modules or the width of the memory bus. For gamers, this means more VRAM for memory-hungry games and applications. For Nvidia and its partners, however, it currently means one thing above all else—significantly higher purchase prices.
Why GDDR7 Memory Is Becoming a Problem
The numbers explain the hesitation quite clearly. A single 3-GB GDDR7 chip currently costs between $60 (about 52 euros) and $70 (about 61 euros). A standard 2-GB module costs only about $20 (around 17 euros). The bottom line is that the new chips offer only 50 percent more capacity but are about three times as expensive to procure.
A graphics card requires six to eight of these modules. As a result, the pure production costs increase by several hundred euros, depending on the model—even before factoring in the cooler, circuit board, and profit margin. Nvidia has therefore instructed its board partners to postpone sales for the time being until the market situation stabilizes.
Added to this is an availability issue. The 3-GB chips are in short supply because they’re in high demand in the professional sector. Nvidia uses the same memory in highly profitable products for AI infrastructure, such as the RTX PRO 6000 Blackwell or the Rubin CPX. When allocating these scarce chips, these cards take priority—gaming GPUs are at the back of the line. For gamers, this simply means a longer wait.
These models are affected—the expected specs
The “Super” update affects several cards. The focus is on the RTX 5080 Super, along with the RTX 5070 Ti Super and the RTX 5070 Super. An RTX 5050 with 9 GB is also reportedly on hold for the same reason. The following overview summarizes the leaked information currently circulating—none of it has been officially confirmed yet.
| Model | Memory (expected) | Memory Clock | Power Consumption |
|---|---|---|---|
| RTX 5080 Super | 24 GB GDDR7 (instead of 16 GB) | up to 32 Gbps | up to 415 W |
| RTX 5070 Ti Super | 24 GB GDDR7 (instead of 16 GB) | 28 Gbps, 256-bit | TBD |
| RTX 5070 Super | 18 GB GDDR7 (instead of 12 GB) | TBD | TBD |
| RTX 5050 (9 GB) | 9 GB GDDR7 (3 × 3-GB chips) | TBD | TBD |
The RTX 5080 Super, in particular, stands out with a massive leap: With memory increasing from 16 to 24 GB plus a faster clock speed (32 Gbps instead of 30 Gbps), the bandwidth climbs toward 1 TB/s. For now, however, anyone looking for a current graphics card for their gaming PC will have to continue shopping within the existing RTX 50 lineup.
More VRAM, more watts: higher power consumption expected
The new memory modules not only cost more, but they also apparently require more power. Early entries in Seasonic’s power supply calculator suggest that power consumption is increasing noticeably. For the RTX 5080 Super, up to 415 watts are listed there—an increase of 55 watts, or about 15 percent, compared to the standard model’s 360 watts.
What’s interesting is the reason behind this: According to current information, the number of processing cores remains unchanged. The increased power consumption is therefore primarily due to the new memory design. It’s also possible that Nvidia is slightly raising the clock speeds to justify the higher power consumption in the first place. Anyone already considering an upgrade should keep an eye on their power supply—and while they’re at it, maybe check out how to optimally set up an Nvidia system with G-Sync.
When is the RTX 5080 Super coming out?
That’s the big unknown. There’s no official release date, and Nvidia hasn’t commented on either the specs or the delay so far. Some reports are now speculating on a launch window in late 2026 or around CES 2027—but nothing is certain. As long as memory prices remain at current levels, Nvidia is unlikely to have much interest in rushing the cards to market with slim margins or at inflated prices.
FAQ on the RTX 5080 Super and the RTX 50 Super Series
Why is Nvidia delaying the RTX 50 Super?
Because the required 3-GB GDDR7 memory chips are currently about three times as expensive as the previous 2-GB modules and are also in short supply. This drives up production costs per card by several hundred euros.
Which models are part of the RTX 50 Super series?
So far, the RTX 5080 Super, RTX 5070 Ti Super, and RTX 5070 Super have been announced. An RTX 5050 with 9 GB is also reportedly included.
How much memory will the RTX 5080 Super have?
According to leaks, 24 GB of GDDR7 instead of the previous 16 GB, combined with a faster memory clock speed of up to 32 Gbps.
What is the power consumption of the RTX 5080 Super?
Entries in the Seasonic power supply calculator list up to 415 watts—about 55 watts more than the standard model.
Is there a release date for the RTX 5080 Super?
No. There is no official release date. Some reports speculate on late 2026 or CES 2027, but nothing has been confirmed.
Conclusion: Great Cards at the Wrong Time
To be honest, the delay is understandable. An RTX 5080 Super with 24 GB of VRAM would be a significant upgrade on paper—but it’s of no use to anyone if the memory makes the card so expensive that, in the end, neither Nvidia nor the buyers will be happy. The fact that the scarce 3-GB chips are being diverted to high-margin AI hardware fits the picture: As long as a memory chip can earn more in a data center than in a gaming PC, gamers will get the short end of the stick.
For you, this means one thing: If you need a new graphics card right now, don’t hold out for the “Super” models—instead, go for the current RTX 50 lineup. If you can wait, wait—and keep an eye on memory prices. Are you waiting for the Super variants, or do you not need a new GPU right now anyway? Feel free to share your thoughts in the comments.
Sources: VideoCardz, WinFuture, Tom’s Hardware, TweakTown (Seasonic)
