Octanebench is one of the benchmarks that can maximise a GPU’s capabilities. It is utilised for rendering and can increase a graphic card’s throughput with little to no driver optimization. In other words, consider it the best-case situation for any particular GPU. The NVIDIA RTX 4070 Ti was recently discovered in the Octanebench database (through LaptopVideo2Go via Videocardz) and outperforms the previous RTX 3090 Ti by 5%.
The RTX 3090 Ti features 10,752 CUDA cores clocked at 1860 MHz, producing 39.99 TFLOPs in FP32. The RTX 4070 Ti, on the other side, has 7680 CUDA cores clocked at 2610 MHz and delivers 40.09 TFLOPs in FP32.
Octanebench can now allegedly take into account the RT cores on the card, of which the NVIDIA RTX 4070 Ti has 60 vs the 3090 Ti’s 84. Overall, the RTX 3090 Ti is extremely similar to the RTX 4070 Ti on paper, and it is interesting to see that at least one benchmark can take full advantage of this.
The NVIDIA RTX 4070 Ti receives 725.3 points, whereas the RTX 3090 Ti receives 688.17 points (or 5% faster).
It is important to note, however, that this score is a combination score, with the findings substantially aggregated and weighted. The RTX 4070 Ti performs substantially better in the info channels test (presumably due to the additional VRAM), scoring 953 versus the RTX 3090 Ti’s 697. However, it scores approximately the same in Direct Lightning and Path Tracing, with scores of 705 and 696 compared to the RTX 3090 Ti’s 687 and 687.
At first glance, this appears to be a decent performance increase from NVIDIA and a reasonable replacement for the RTX 3090 Ti. This is not a gameplay test, but rather a render test. Driver optimizations have a significant impact on games, with only a partial link to peak theoretical FP32 performance. And there’s no way of knowing whether the card used here was heavily OC’d or the vanilla stock version. Before drawing any conclusions, it is always preferable to have a bigger sample size.
However, the RTX 3090 Ti should continue to outperform the RTX 4070 Ti in aggregate gaming performance for the foreseeable future. However, the faster clock speeds and architectural changes may make this a desirable card for rendering enthusiasts and video professionals.
Also Read:
