The NVIDIA RTX 4070 SUPER is the objective winner of this comparison, securing victories in nine out of thirteen measured specifications. Its most significant advantage over the NVIDIA GTX 1070 Ti is its FP16 performance, which is 27192% higher. While the older card maintains a lead in memory clock speed, outperforming the newer model by 52% in that specific category, the overall technical superiority of the newer architecture is clear across the majority of the evaluated metrics.

Users prioritizing raw computational throughput and modern architectural capabilities should choose the NVIDIA RTX 4070 SUPER for its massive lead in FP16 performance. Conversely, the NVIDIA GTX 1070 Ti remains a viable option for specific use cases that rely heavily on a higher memory clock speed. Selecting between these two components depends on whether your workload demands the superior processing power of the newer card or the specific memory frequency advantage offered by the older hardware.