One of the benchmarks that can increase the capabilities of the GPU is Octanebench. It is used for display purposes and can increase the throughput of graphics cards with almost no need for driver optimization. In other words, you can think of it as the best case scenario for any GPU. The NVIDIA RTX 4070 Ti was recently spotted (via LaptopVideo2Go via Videocardz) in the Octanebench database and outperforms the older RTX 3090 Ti by 5%.
NVIDIA RTX 4070 Ti is 5% faster than RTX 3090 Ti in first Octanebench score
Now that we’re dealing with a benchmark that doubles the theoretical throughput of a GPU, let’s look at the theoretical numbers first. The RTX 3090 Ti has 10,752 CUDA cores clocked at 1860MHz which yields an FP32 performance of 39.99 TFLOPs. On the other hand, the RTX 4070 Ti has 7680 CUDA cores clocked at 2610MHz resulting in an FP32 performance of 40.09 TFLOPs.
Now Octanebench can also take into account the RT cores on the card, of which the NVIDIA RTX 4070 Ti has 60 compared to the 3090 Ti’s 84. All said and done, the RTX 3090 Ti is very close, on paper, to the RTX 4070 Ti and interestingly enough, We see that there is at least one benchmark that can take full advantage of this.
The RTX 4070 Ti gets an overall score of 725.3 points while the RTX 3090 Ti gets 688.17 points (or 5% faster). However, it should be noted that this score is a score with aggregated and heavily weighted scores. I noticed that the RTX 4070 Ti is significantly strong on the Teletext test (probably due to the updated VRAM) scoring 953 compared to the RTX 3090 Ti’s 697. On the Direct Lightning and Path Tracing side, it scores roughly the same with scores 705 and 696 compared to the RTX. 3090 Ti’s 687 and 687.
This looks like a decent performance upgrade from NVIDIA and a solid replacement for the RTX 3090 Ti at face value. However, this is where the comparison falls apart. There are a lot of caveats associated with this. 1) This is an introduction test, not a play test. Gaming is greatly affected by driver optimizations and has only a partial correlation to the FP32’s theoretical peak performance. 2) We don’t know if the card used here is a heavily OC-ed version or a vanilla stock version. It is always best to obtain a larger sample size before drawing any conclusions.
All of this means that in real-world scenarios, the RTX 3090 Ti should continue to stay faster than the RTX 4070 Ti on an overall gaming basis for the foreseeable future. However, the increased clock speeds and architectural improvements can make this card desirable to video enthusiasts and video professionals.