Product photos

The NVIDIA GeForce RTX 4070 SUPER graphics card is engineered to enhance gaming and creative applications. It incorporates ray tracing, Deep Learning Super Sampling (DLSS), and AI-assisted technologies. Utilizing the AD104 silicon, similar to the RTX 4070 Ti, this card activates 56 out of 60 streaming multiprocessors. This configuration results in 7168 CUDA cores, compared to the 7,680 in the RTX 4070 Ti

It includes 56 RT cores, 224 Tensor cores, 224 Texture Mapping Units (TMUs), and 80 Raster Operations Pipelines (ROPs). The RTX 4070 Super has a 192-bit wide memory interface with 12GB of 21Gbps GDDR6X memory, offering a memory bandwidth of 504GB/s.

The GeForce RTX 4070 Super Founder Edition is designed with a sophisticated aesthetic and a robust build. The card occupies 2.5 slots, marking a significant presence in any system configuration. With the ADA series replacing the Ampere lineup, NVIDIA introduces the GeForce RTX 4070 Founder Edition.

The design mirrors that of the RTX 4080/4090 Founder Edition, retaining the Dual-Axial Flow-Through cooling architecture, now improved with larger radiators and an enhanced fan design. You'll also notice that the card is colored a bit darker compared to the previous models.

A notable feature of the RTX 40-series is NVIDIA's exclusive provision of power adapters for both the Founders Edition and custom-designed boards by partners. The GeForce RTX 4070, while adhering to PCIe Gen 5 specifications, operates with a PCIe Gen 4.0 x16 host interface. The card is equipped with 12GB of GDDR6X memory and a 192-bit wide memory bus. It has a Total Graphics Power (TGP) of 220W, with a recommended minimum power supply of 500~550W. However, in practice, the Founder Edition card typically consumes between 175-200W on average. Board partners may utilize higher TGP limits, leading to increased wattage requirements for the card though.