Recent developments indicate that the launch of NVIDIA’s GeForce RTX 50 SUPER GPUs has been rescheduled to the third quarter of 2026, primarily due to ongoing shortages in DRAM.
NVIDIA GeForce RTX 50 GPU Launch Delayed to Q3 2026
Initially, the anticipated release of the NVIDIA GeForce RTX 50 SUPER series was set for early 2026, specifically Q1. However, supply chain challenges concerning 3 GB GDDR7 memory modules have necessitated this shift to a later timeline. Contrary to circulating rumors about a potential cancellation, insiders confirm that these GPUs are very much in the pipeline.
No cancellation yet, just delayed. Now planned to launch in 26Q3, previously set to launch in 26Q1.https://t.co/C3RmFv1uIq
— MEGAsizeGPU (@Zed__Wang) November 7, 2025
According to sources like MegaSizeGPU, the GPU series remains on track, with the unveil likely coinciding with Computex in 2026. NVIDIA is expected to introduce three models within this SUPER edition:
- GeForce RTX 5080 SUPER
- GeForce RTX 5070 Ti SUPER
- GeForce RTX 5070 SUPER
The updated lineup is projected to leverage advanced 3 GB GDDR7 memory, which could significantly enhance performance. Nonetheless, inflated production costs and components scarcity may prompt NVIDIA to prioritize models with more robust profit margins, such as the RTX PRO 6000 Blackwell GPUs, which utilize the same GDDR7 dies but offer up to 96 GB VRAM, or the RTX 5090 Laptop GPU, boasting up to 24 GB VRAM through a 256-bit bus architecture.
Previously aimed for the Q1-Q2 of 2026, NVIDIA had plans for a formal unveiling at CES 2026 or a dedicated GeForce event. While these intentions had not been formally announced, leaks have suggested otherwise, and our investigations corroborate that the RTX 50 SUPER series is not slated for release in 2025.

In the current competitive landscape, NVIDIA faces limited competition from AMD’s Radeon RX 9000 series. However, consumer demand for higher VRAM configurations highlights the importance of the upcoming RTX 50 SUPER family. The expected specifications include:
- RTX 5080 SUPER: up to 24 GB VRAM
- RTX 5070 Ti SUPER: up to 24 GB VRAM
- RTX 5070 SUPER: up to 18 GB VRAM
This notable enhancement in VRAM will undoubtedly make the GPUs appealing. However, the escalating costs associated with DRAM could hinder NVIDIA’s ability to maintain competitive pricing relative to their non-SUPER variants. Ultimately, these forthcoming GPUs promise to rejuvenate the current lineup just ahead of the anticipated next-generation Rubin gaming series, projected for release in 2027.
Anticipated NVIDIA RTX 50 SUPER GPU Models:
| Graphics Card Model | NVIDIA GeForce RTX 5080 SUPER | NVIDIA GeForce RTX 5080 | NVIDIA GeForce RTX 5070 Ti SUPER | NVIDIA GeForce RTX 5070 Ti | NVIDIA GeForce RTX 5070 SUPER | NVIDIA GeForce RTX 5070 |
|---|---|---|---|---|---|---|
| GPU Name | Blackwell GB203-450 | Blackwell GB203-400 | Blackwell GB203-350 | Blackwell GB203-300 | Blackwell GB205-400 | Blackwell GB205-300-A1 |
| GPU Streaming Multiprocessors | 84 (Full) | 84 (Full) | 70 (Full) | 70 (Full) | 50 (Full) | 48 (Partial) |
| Total CUDA Cores | 10752 | 10752 | 8960 | 8960 | 6400 | 6144 |
| Base Clock Speeds | TBD | 2.62 GHz | TBD | 2.42 GHz | TBD | 2.51 GHz |
| Memory Capacity | 24 GB GDDR7 | 16 GB GDDR7 | 24 GB GDDR7 | 16 GB GDDR7 | 18 GB GDDR7 | 12 GB GDDR7 |
| Memory Interface | 256-bit | 256-bit | 256-bit | 256-bit | 192-bit | 192-bit |
| Memory Speed | 32 Gbps | 30 Gbps | 28 Gbps | 28 Gbps | 28 Gbps | 28 Gbps |
| Memory Bandwidth | 1024 GB/s | 960GB/s | 896 GB/s | 896 GB/s | 672 GB/s | 672 GB/s |
| Power Connector | 1 12V-2×6 (16-Pin) | 1 12V-2×6 (16-Pin) | 1 12V-2×6 (16-Pin) | 1 12V-2×6 (16-Pin) | 1 12VHPWR (16-Pin) | 1 12VHPWR (16-Pin) |
| Expected Launch Date | TBD | January 30, 2025 | TBD | February 20, 2025 | TBD | March 5, 2025 |
| Thermal Design Power (TDP) | 400W+ | 360W | 350W | 300W | 275W | 250W |
| Estimated Price | TBD | $999 | TBD | $749 | TBD | $549 |
Leave a Reply