GeForce RTX 5080 SUPER: 24 GB of Memory Instead of 16 GB Amid Record Prices

The GeForce RTX 5080 SUPER remains an unreleased NVIDIA graphics card, although information about it first appeared back in 2025. According to leaks, the model will receive 24 GB of GDDR7 instead of the 16 GB found in the RTX 5080, while retaining the same 10,752 CUDA cores. At the same time, the regular RTX 5080 had risen to $2,200 at US retailers by November 2026, and there are still no independent tests of the SUPER or an official release date.

The main question is no longer how much faster the SUPER will be. If NVIDIA actually releases a 24 GB version, its price could become a serious obstacle for buyers. The regular RTX 5080 already costs almost twice its original recommended price, while the new higher-capacity GDDR7 chips are considerably more expensive than standard ones.

RTX 5080 SUPER Specifications

The rumored RTX 5080 SUPER uses a GB203 graphics processor based on the Blackwell architecture. The number of compute units is expected to remain the same as on the RTX 5080: 84 SMs, 10,752 CUDA cores, and 84 RT cores.

The main difference is therefore expected to be in the memory subsystem.

The RTX 5080 SUPER specifications are based on preliminary reports rather than official NVIDIA documentation. The RTX 5080 price reflects the US market as of the end of October 2026.

The card is expected to use eight 3 GB GDDR7 memory chips instead of eight 2 GB chips. This would increase the video memory capacity by 50% without changing the 256-bit memory bus.

Memory bandwidth, however, would increase by only 6.7%. The higher 415 W power limit could allow GPU clocks to be raised, although specific figures are not yet available.

Thus, the RTX 5080 SUPER is expected to be primarily a higher-memory version rather than a graphics card with a substantially larger number of compute units.

Benchmarks: What Performance to Expect

There are still no real RTX 5080 SUPER results in 3DMark or games. For comparison, tests of existing graphics cards based on the Blackwell, Ada Lovelace, and RDNA 4 architectures can be used.

The Wccftech tests used specific models, including the MSI GeForce RTX 5080 EXPERT OC, MSI RTX 5090 SUPRIM SOC, Colorful RTX 5070 Ti Battle NX, Palit RTX 4080 SUPER GamingPro OC, and ASRock RX 9070 XT Taichi OC. The test system used an Intel Core i9-13900K.

The gaming results are given at a resolution of 3840 x 2160. Alan Wake 2 was tested at maximum settings without ray tracing.

The RTX 5080 scores approximately 19% higher in Time Spy Graphics than the RTX 5070 Ti. However, the gap in games is considerably smaller: around 4% in Alan Wake 2 and less than 1% in Forza Horizon 5.

These are results from a specific set of tests, not an average advantage across all games. Nevertheless, they clearly illustrate the RTX 5080’s problem: high compute performance does not always translate into an equally noticeable increase in FPS.

The RTX 5080 SUPER could be faster thanks to higher GPU and memory clocks. But with the same number of CUDA cores, there is no reason to expect a leap comparable to the transition to the RTX 5090.

Where the Additional 8 GB Will Actually Help

Video memory and compute performance address different needs. If a game uses less than 16 GB of VRAM, moving to 24 GB will have virtually no effect by itself.

Additional memory is important in other scenarios:

• 4K gaming with demanding textures and mods. The larger memory capacity can reduce data streaming and performance drops when 16 GB is no longer enough.
• Local language models. The 24 GB capacity can accommodate a larger quantized model or leave more memory available for context. This is a significant advantage for running neural networks locally.
• Image and video generation. The larger VRAM capacity makes it possible to work at higher resolutions and with some demanding models without moving part of the processing to system memory.
• Blender and 3D graphics. Additional memory is useful when rendering complex scenes with large amounts of geometry and textures.

For these workloads, 24 GB could prove more important than even 10-15% additional compute performance. However, the RTX 5080 SUPER itself would still trail the RTX 5090, which has 21,760 CUDA cores and 32 GB of memory.

Why NVIDIA Is Delaying the RTX 5080 SUPER

In July 2026, VideoCardz reported that at least one graphics card manufacturer had already received RTX 50 SUPER-series units, but NVIDIA had suspended their release.

The reported reason was the high cost of 3 GB GDDR7 memory chips. According to the source, one such module cost approximately $60-70, compared with around $20 for a standard 2 GB module.

For a graphics card with eight memory chips, this means an increase of hundreds of dollars in component costs.

The situation in the graphics card market is creating additional pressure. By November 2026, the RTX 5080 and RTX 5090 had become significantly more expensive, while supplies of high-end models had declined.

There have also been reports of a possible 24 GB RTX 5080 release in early 2027 instead of the original 16 GB model. However, it is unknown whether such a card will be called the SUPER and when it will go on sale.

NVIDIA has officially confirmed neither these plans nor the final specifications of the updated model.

How Much Could the RTX 5080 SUPER Cost?

The original recommended price of the RTX 5080 was $999, but today it has virtually no bearing on the card’s actual price.

According to TechSpot, as of early November 2026, the situation at US retailers is as follows:

All three models currently on sale have become more expensive compared with their recommended prices. The difference is especially large for the RTX 5080.

Today, a buyer pays approximately $950 more for the RTX 5080 than for the RTX 5070 Ti. That is a premium of around 76%, even though the performance advantage is far smaller.

The situation could be even more difficult for the RTX 5080 SUPER. More expensive memory and the higher power limit make inexpensive production unlikely. At the same time, neither the exact price nor even the price range of the upcoming card can currently be confirmed.

Is It Worth Waiting for the RTX 5080 SUPER?

For gaming, buying an RTX 5080 for $2,200 right now offers poor value. The RTX 5070 Ti at $1,250 provides similar performance in many games, while the Radeon RX 9070 XT at $870 looks even more attractive if high ray-tracing performance and CUDA capabilities are not required.

The RTX 5080 SUPER is primarily worth waiting for if you need 24 GB of video memory for local neural networks, rendering, or other demanding workloads. In these scenarios, the increased memory capacity could provide more benefit than a small FPS increase.

But neither the release date, the price, nor the final specifications are known yet. Therefore, it is worth delaying a purchase for the RTX 5080 SUPER only if there is no urgent need to upgrade your graphics card.

If NVIDIA maintains the current pricing level for high-end GeForce models, even 24 GB of memory may not be enough to justify the purchase.