NVIDIA is gearing up to unveil a new wave of GeForce RTX 50 "Blackwell" gaming GPUs at the start of 2025, with at least four models on the horizon, followed by more mainstream options later in the year.
In the first quarter of 2025, we can look forward to seeing the NVIDIA GeForce RTX 5070 Ti and 5070 join the 5090 and 5080. However, other RTX 50 "Blackwell" versions will roll out subsequently. Over the past few days, a flurry of leaks and speculations has shed light on these much-anticipated GPUs. Initially, many believed NVIDIA would kick things off with just the RTX 5090 and 5080. Now, it seems two additional models could debut around the same period.
According to insights from Benchlife, NVIDIA seems poised to release both the GeForce RTX 5070 Ti and 5070 shortly following the high-end RTX 5090 and RTX 5080. The anticipated launch in early 2025 suggests that the Consumer Electronics Show (CES) might serve as the grand stage for unveiling these new flagship models.
Interestingly, while the spectacular RTX 5090 and 5080 are likely to be showcased at CES, the 5070 Ti and 5070 are expected to emerge a bit later, perhaps in February 2025. Initially, speculation was rife that we’d only see the RTX 5070 as the sole representative in the ’70 series early on, but it now looks like the 5070 Ti is part of the action too. This means NVIDIA is likely packing all these models into the same launch timeframe.
Come about May 2025, we could see NVIDIA expanding their lineup further with the potential introduction of the GeForce RTX 5060 Ti and 5060. While several sources suggest this May timeline, it’s crucial to highlight that not all have confirmed these details. As the initial launches progress, we might uncover more about these slower models and their specs.
Diving into specifics, the GeForce RTX 5070 Ti is said to deliver a powerful punch as the third-fastest GPU in its series. It’s rumored to feature a hefty 8,960 CUDA cores, 16GB of GDDR7 memory, and a substantial TGP of 300W. Benchlife has noted that while most reports point to this 300W figure, one NVIDIA Board Partner (AIB) has hinted it could be generous and reach 350W. Meanwhile, the RTX 5070 would clock in with a 250W TGP, 6,400 CUDA cores, and 12GB of GDDR7 VRAM.
As for the RTX 5060 Ti and 5060, details remain more elusive. We’ll be keeping an ear to the ground and updating you as soon as more concrete information about these GPUs becomes available.
Here’s a quick look at some rumored specifications for the impending NVIDIA GeForce RTX 5090 and 5080 GPUs:
| Graphics Card Name | NVIDIA GeForce RTX 5090 | NVIDIA GeForce RTX 5080 |
|---|---|---|
| GPU Name | Blackwell GB202-300 | Blackwell GB203-400 |
| GPU SMs | 170 (192 Full) | 84 (84 Full) |
| GPU Cores | 21,760 (+33%) | 10,752 (+11%) |
| Memory Capacity | 32 GB GDDR7 (+33%) | 16 GB GDDR7 (0%) |
| Memory Bus | 512-bit (+33%) | 256-bit (0%) |
| Memory Speed | 28-32 Gbps | 32 Gbps |
| Bandwidth | 1792-2048 GB/s | 1024 GB/s |
| TBP | 600W (+33%) | 400W (+25%) |
So, which NVIDIA GeForce RTX 50 GPU are you most excited about? Keep following for more insights and news from the world of gaming.
News Source: Benchlife
Share this story on Facebook and Twitter.
