The final specifications for the NVIDIA GeForce RTX 5070 and RTX 5070 Ti appear to be confirmed.
The leaked specifications of NVIDIA’s future GeForce RTX 5070 and RTX 5070 Ti graphics cards have been finalized thanks to kopite7kimi.
Starting with the RTX 5070 Ti, it will have 8,960 CUDA cores and 16 GB of GDDR7 memory on a 256-bit memory bus, with a bandwidth of 896 GB/s. According to reports, the card’s total board power (TBP) is 300 W. The PG147-SKU60 board design with a GB203-300-A1 GPU seems to be used by the Ti model. With specifications indicating 6,144 CUDA cores and 12 GB of GDDR7 memory on a 192-bit bus, along with 627 GB/s memory bandwidth, the standard RTX 5070 is positioned as a more power-efficient choice. It is anticipated that this model will run at a little lower 250 W TBP.
Remarkably, the GB205-300-A1 GPU will be used by both the PG146 and PG147 board types of the non-Ti RTX 5070 card. Although the pricing structure is unknown to us, it is evident that NVIDIA has decided to make significant differences across its SKUs. In addition to an additional four gigabytes of GDDR7 memory, the Ti variant’s CUDA core count is increased by an astounding 45%, from 6,144 to 8,960. The GeForce RTX 5070 and RTX 5070 Ti are anticipated to debut later, potentially after the RTX 5080 and RTX 5090 GPUs, as we wait for the first wave of GeForce RTX 50 series cards to be unveiled at CES.
