Nvidia recently introduced its latest graphics card, the GeForce RTX 4070, which operates on the Ada Lovelace architecture and incorporates the company’s DLSS 3 AI performance multiplier.
Compared to the RTX 4070 Ti, the RTX 4070 has fewer CUDA Cores (5888 versus the Ti’s 7680) and a lower boost clock (2.48 Ghz compared to the Ti’s 2.61). Despite these differences, both GPUs include 12GB of GDDR6X RAM and can support a maximum display resolution of 4K at 240 Hz or 8K at 60 Hz.
Similar to the other graphics cards in the 40-series lineup, the RTX 4070 comes equipped with 3rd generation RT cores and 4th generation Tensor cores, which enable features such as shader execution reordering and DLSS3 frame generation.
Nvidia asserts that the RTX 4070 delivers performance on par with the RTX 3080 while consuming half the power.
The RTX 4070 boasts 12GB of memory, compared to the standard 10GB found on the RTX 3080. With regard to frame generation, Nvidia reports that the RTX 4070 is 1.4 times faster than the 3080 and 2.6 times faster than the 2070 Super, both of which lack this capability.
Additionally, the RTX 4070 contains an 8th generation NVENC encoder that supports encoding in AV1, although it only has one of these compared to the two included on the other 40-series cards. With a total graphics power of 200W, the RTX 4070 has an average gaming power consumption of 186W according to Nvidia’s claims.
The RTX 4070 is now available for purchase starting at โฑ40,590 in Asus Branches nationwide, including Nvidia.
GeForce RTX 4070, better performance, lower power consumption
