NVIDIA officially unveils the complete specs of GeForce RTX 5070 Ti & RTX 5070 cards in the Blackwell lineup, detailing everything.
NVIDIA GeForce RTX 5070 Ti & RTX 5070 Specs Unveiled: Powered by GB203 and GB205 GPUs Respectively
Even though NVIDIA revealed the RTX 50 GPUs, including the 70-class cards at CES, we still didn't have the complete information on their specifications, which has now been revealed in the PDF officially. We do know the core specs of both the GeForce RTX 5070 Ti and RTX 5070, but there are some minor changes in the specifications, plus a lot of additional stuff has been released by NVIDIA.
As you may know, the NVIDIA GeForce RTX 5070 Ti graphics card is currently the third-fastest GPU in the Blackwell lineup and succeeds the RTX 4070 Ti. The RTX 5070 Ti is based on the GB203 GPU die, which is also found on the RTX 5080. Some specs are, therefore, technically equal or closer on both cards, including the VRAM, which is 16 GB GDDR7. The RTX 5070 Ti, however, brings the usual 28 Gbps speed vs 30 Gbps on the RTX 5080.
The NVIDIA GeForce RTX 5070, on the other hand, will be based on the GB205 die, making it the slowest among the revealed GPUs, and will succeed the RTX 4070. Unlike the latter, there seems to be a downgrade in the GPU die as 70-class cards in the Ampere and Ada lineups used the 104 die. Nonetheless, we can see that the GPU does bring an uplift in Cuda Cores, Tensor Cores, RT Cores, and other areas.
However, the VRAM capacity remains the same at 12 GB on a 192-bit bus except for a faster GDDR7 memory. The GPU boost clock for the RTX 5070 is 2512 MHz, which is slightly higher than the 2475 MHz on the RTX 4070. On the other hand, the RTX 5070 Ti got a noticeable downgrade in boost clock, bringing 2452 MHz vs 2610 Mhz to the RTX 4070 Ti. Still, other areas have seen an uplift, including the memory configuration, which now brings 16 GB GDDR7 memory vs 12 GB GDDR6X and a wider memory bus width of 256-bit.
It is rated at 300W while the non-Ti will work at 250W for the FE model. AIB editions can consume more depending on their cooling solution and the boost clock out of the box. It's expected that the RTX 5070 Ti will be launched on 20th February and the RTX 5070 will follow a few weeks later in March. Both cards feature a $50 price cut compared to their predecessors, but their custom editions can cost significantly more.
NVIDIA GeForce RTX 5070 Ti 16 GB Graphics Card
The NVIDIA GeForce RTX 5070 Ti will feature a cut-down GB203 GPU with 8960 CUDA Cores. The memory side will feature the same 16 GB VRAM as the RTX 4070 Ti SUPER but upgraded with brand-new GDDR7 memory chips, offering a bandwidth boost of 78% over the 4070 Ti to 896 GB/s. As per the specs page, the GeForce RTX 5070 Ti will operate at a boost clock of 2.45 GHz.
In terms of performance, the NVIDIA GeForce RTX 5070 Ti will offer 2x the uplift over the RTX 4070 Ti while being priced at $749 US, which is better than the 4070 Ti & 4070 Ti SUPER given its improvements and support for DLSS 4 capabilities.
NVIDIA GeForce RTX 5070 12 GB Graphics Card
Moving over to the NVIDIA GeForce RTX 5070, this graphics card will come with the GB205 GPU on the PG147 reference PCB. The GPU side will feature up to 6144 CUDA cores and the memory side will feature a 192-bit bus interface with 12 GB of memory, similar to the existing 12 GB RTX 4070 (192-bit).
The card will feature 28 Gbps memory speeds for up to 672 GB/s of total bandwidth. This will be a 33% increase in bandwidth versus the RTX 4070 & RTX 4070 SUPER GPUs. This specific variant will feature a 250W TBP, marking a +14% increase over the RTX 4070 SUPER (220W).
The NVIDIA GeForce RTX 5070 is said to be faster than the RTX 4090 at a price of just $549 US. That should make the card a potent offering in the high-end segment that is closer to the $500 US mark and poses a big threat to AMD's recently announced RX 9070 series graphics cards. Both the NVIDIA GeForce RTX 5070 Ti and RTX 5070 are listed as powerful 1440p cards, while the 5080 goes the 4K gaming route.
Just for reference, we do know what each max configuration of the Blackwell "RTX 50" GPUs will have to offer and the GB205 maxes out with 5 GPCs or 50 SMs and a total of 6400 CUDA cores. It can be seen how the RTX 50 SKUs are arranged in the list below:
- GB202 (FULL) -> 192 SMs / 24576 Cores (100%) / RTX 5090 -> 170 SMs / 21760 Cores (88.5%)
- GB203 (FULL) -> 84 SMs / 10752 Cores (100%) / RTX 5080 -> 84 SMs / 10752 Cores (100%)
- GB205 (FULL) -> 50 SMs / 6400 Cores (100%) / RTX 5070 -> TBD SMs / TBD Cores (TBD %)
The NVIDIA GeForce RTX 5080 will be available on the 30th of January while the RTX 5070 Ti & the RTX 5070 will be launched in February. All cards will be available in custom variants and the very liked Founders Edition models.
NVIDIA RTX 50 Founders Edition Models:
NVIDIA GeForce RTX 50 GPU Specs (Official):
| Graphics Card Name | NVIDIA GeForce RTX 5090 | NVIDIA GeForce RTX 5080 | NVIDIA GeForce RTX 5070 Ti | NVIDIA GeForce RTX 5070 | NVIDIA GeForce RTX 5060 Ti | NVIDIA GeForce RTX 5060 | NVIDIA GeForce RTX 5050 |
|---|---|---|---|---|---|---|---|
| GPU Name | Blackwell GB202-300 | Blackwell GB203-400 | Blackwell GB203-300 | Blackwell GB205-300 | Blackwell GB206-300 | Blackwell GB206-250 | Blackwell GB207-300 |
| GPU SMs | 170 | 84 | 70 | 48 | 36 | 30 | 20 |
| GPU Cores | 21760 | 10752 | 8960 | 6144 | 4608 | 3840 | 2560 |
| Clock Speeds | 2017 MHz (Base) 2407 MHz (Boost) | 2295 MHz (Base) 2617 MHz (Boost) | 2295 MHz (Base) 2452 MHz (Boost) | 2325 MHz (Base) 2512 MHz (Boost) | 2407 MHz (Base) 2572 MHz (Boost) | 2280 MHz (Base) 2497 MHz (Boost) | 2317 MHz (Base) 2572 MHz (Boost) |
| Memory Capacity | 32 GB GDDR7 | 16 GB GDDR7 | 16 GB GDDR7 | 12 GB GDDR7 | 8-16 GB GDDR7 | 8 GB GDDR7 | 8 GB GDDR6 |
| Memory Bus | 512-bit | 256-bit | 256-bit | 192-bit | 128-bit | 128-bit | 128-bit |
| Memory Speed | 28 Gbps | 30 Gbps | 28 Gbps | 28 Gbps | 28 Gbps | 28 Gbps | 20 Gbps |
| Bandwidth | 1792 GB/s | 960 GB/s | 896 GB/s | 672 GB/s | 448 GB/s | 448 GB/s | 320 GB/s |
| Power Interface | 1 12V-2x6 (16-Pin) | 1 12V-2x6 (16-Pin) | 1 12V-2x6 (16-Pin) | 1 12V-2x6 (16-Pin) | 16-Pin or 8-Pin | 8-Pin | 8-Pin |
| Launch | 30th Jan, 2025 | 30th Jan, 2025 | 20th Feb, 2025 | 4th March, 2025 | 16th April, 2025 | 19th April, 2025 | 1st July, 2025 |
| TBP | 575W | 360W | 300W | 250W | 180W | 145W | 130W |
| Price (MSRP) | $1999 | $999 | $749 | $549 | $379-$429 | $299 | $249 |
News Source: NVIDIA
About the author: Sarfraz Khan is a hardware reporter with a focus on PC components and the builder community. With years of experience writing about PC hardware and laptops, his work has been featured on several reputable technology publications. Sarfraz's hands-on experience is demonstrated through his first-person accounts of using and comparing different hardware configurations, providing practical and relatable insights for everyday users. His technical analysis is respected by peers in the enthusiast community and has been cited by specialized hardware sites such as Germany's Igor's Lab.
Follow Wccftech on Google to get more of our news coverage in your feeds.
