The NVIDIA GeForce RTX 4070 Ti graphics card will be released in early January with the same specs as the unreleased GeForce RTX 4080 12 GB graphics cards. According to rumors, the graphics card may have the same $899 US price tag as the previous-generation card, although this remains to be seen.
The Geekbench 5 OpenCL benchmark test reveals that the NVIDIA GeForce RTX 4070 Ti can boost to 2,730 MHz, and according to the performance metrics, the card is comparable to the AMD Radeon RX 7900 XTX, which costs $999 US. The card is around 5% quicker than the RX 7900 XT in the same benchmark, and if NVIDIA is able to achieve the same performance boost in AAA titles, it will interpret why NVIDIA wants to maintain the pricing at $899 US. The RTX 4070 Ti is 45-50% faster than its predecessor, which is a significant improvement, but we must wait to see if games show a comparable rise in performance.
The “Ada Lovelace” AD104-400 GPU configuration used by the NVIDIA GeForce RTX 4070 Ti will have 7,680 cores, or 60 SMs enabled; this is the maximum SKU. The GPU will have 48 MB of L2 cache and a 192-bit bus interface; thus, we may expect up to 12 GB of GDDR6X capacity. The memory is supposed to run at 21 Gbps, with a memory bandwidth of 504 GB/s. The card is stated to have a boost clock of 2610 MHz, but the peak frequency will be greater.
NVIDIA GeForce RTX 4070 Ti 12 GB “Official” TBP – 285W
NVIDIA GeForce RTX 3070 Ti 8 GB “Official” TBP – 290W
The TBP now has a power rating of 285W, which is a 35W drop from the RTX 3080 10 GB variant and a 5W decrease from the RTX 3070 Ti. The new graphics card is projected to outperform the RTX 3080, but considering the specifications reduction over the 4080 16 GB, there will be a significant difference between the two variants of roughly 30-40%.