The official launch of NVIDIA’s new RTX 50 series graphics cards is fast approaching, and they are expected to set new industry standards. While the release was initially rumored to be in early 2025, recent leaks have revealed possible specifications for these GPUs.
These leaked details cover key aspects such as the number of cores, memory capacity, power consumption, and other significant features. Here’s a closer look at what to expect from the upcoming release.
Leaked technical specifications of NVIDIA RTX 50 graphics cards
Recently, it was confirmed that Jensen Huang, CEO of NVIDIA, will be attending CES 2025 and is scheduled to deliver the opening keynote on January 6, 2025. While no specific details have been provided about what will be announced, many are hopeful that this will be the ideal moment to unveil the next generation of GeForce RTX 50 series graphics cards.
What has truly surprised everyone, though, is the recent leak of potential specifications for these GPUs, which has sparked a lot of excitement among users. According to Winfuture, WccfTech shared a list of technical details for the RTX 5090, 5080, and 5070 models. While these details are intriguing, it’s important to remain cautious, as the sources haven’t provided any further confirmation.
| Graphics Card Name | NVIDIA GeForce RTX 5090 | NVIDIA GeForce RTX 5080 | NVIDIA GeForce RTX 5070 |
|---|---|---|---|
| GPU Name | Blackwell GB202-300 | Blackwell GB203-400 | Blackwell GB205 |
| GPU SMs | 170 (192 Full) | 84 (84 Full) | TBD |
| GPU Cores | 21760 (+33%) | 10752 (+11%) | TBD |
| Clock Speeds | TBD | TBD | TBD |
| L2 Cache | TBD | TBD | TBD |
| Memory Capacity | 32 GB GDDR7 (+33%) | 16 GB GDDR7 (0%) | 12 GB GDDR7 |
| Memory Bus | 512-bit (+33%) | 256-bit (0%) | 192-bit (0%) |
| Memory Speed | 28 Gbps | 32 Gbps | 28 Gbps |
| Bandwidth | 1792 GB/s | 1024 GB/s | 672 GB/s |
| TBP | 600W (+33%) | 400W (+25%) | 250W (+14%) |
| Power Interface | 1 12V-2×6 (16-Pin) | 1 12V-2×6 (16-Pin) | 1 12VHPWR (16-Pin) |
The RTX 5090, expected to be the flagship model, is rumored to feature 32 GB of DDR7 memory with a 512-bit interface and a Blackwell GB202-300 GPU with 21,750 CUDA cores. It’s also said to have a power consumption of 600 watts.
Next, the RTX 5080 is reported to be a slightly lower-tier model, featuring a GB203-400 chip with 10,752 CUDA cores, 16 GB of memory, and a 256-bit interface. Its power consumption is expected to be around 400 watts.
As for the RTX 5070, details are more limited. It is said to come with 12 GB of memory and a power consumption of 250 watts, but no information about the number of CUDA cores has been leaked yet.
As always, this information should be taken with caution until NVIDIA officially confirms the specs. If these details turn out to be accurate, the new generation of GPUs will be incredibly powerful. We’ll have to wait for more updates, possibly during CES 2025, although no official announcement date has been set yet.
