NVIDIA’s Lovelace architecture provides 9,728 CUDA cores to the NVIDIA 4080 GPU compared to 8,704 CUDA cores of the previous generation NVIDIA 3080 GPU. This GPU is designed for a PCIe 4.0 interface. The interface bi-directional transfer speeds of 64GB/s is twice that of the previous generation using a PCIe 3.0 connection interface. 304x 4th generation Tensor cores delivering up to 48.74 FP32 TFLOPS improve the capabilities of this GPU for data science applications. 76x 3rd generation Ray-Tracing cores support 2x to 3x the speed compared to 2nd generation RT cores, along with a four factor increase in rendering abilities with deep learning super sampling 3 (DLSS 3) versus brute-force rendering.
The NVIDIA RTX 4080 Ada generation supports 16GB GDDR6X Error Correction Code (ECC) memory with a bandwidth of 716.8GB/s.
With a Thermal Design Point (TDP) of 320W, the NVIDIA 4080 has a suggested PSU rating of 700W. A single 16-pin power connector is available on this card. In order to regulate temperatures, there are 3x fans available on the 4080.
Offering impressive performance for gaming, simulation, AI modeling, rendering and graphical applications, the NVIDIA GeForce RTX 4080 GPU is ideal for intensive workloads. It has 16GB of GDDR6X memory with 76x RT cores, 304x Tensor cores, and 9728x CUDA cores.