NVIDIA GeForce RTX 2080 Ti To GTX 980 Ti TensorFlow Benchmarks With ResNet-50, AlexNet, GoogLeNet, Inception, VGG-16
The AlexNet model with FP16 precision shows again the strong performance potential with the RTX 2080 Ti...
With AlexNet at FP16 we see again how the average power consumption is actually lower than the GTX 1080 Ti though with the peak power consumption still being higher... Likely due to the efficiency of the tensor cores on Turing? The GPU temperature of the RTX 2080 Ti Founder's Edition continues being quite good compared to the other Founder's Edition hardware benchmarked.
On a performance-per-Watt at FP16, the GeForce RTX 2080 Ti continues as a big deal for those into deep learning.