Why you can trust Tom's Hardware Our expert reviewers spend hours testing and comparing products and services so you can choose the best for you. Find out more about how we test.
Nvidia RTX 4070 Ti Super AI Performance
GPUs are also used with professional applications, AI training and inferencing, and more. Along with our usual professional tests, we've added Stable Diffusion benchmarks on the various GPUs. AI is a fast-moving sector, with many of the publicly available projects designed specifically to run on Nvidia GPUs. That may be changing, slowly, but Nvidia has a reputation as the leader in AI for a reason right now.
For our AI testing, we're using Automatic1111's Stable Diffusion project for the Nvidia cards, with the TensorRT extensions enabled. Intel has an OpenVINO fork of Automatic1111's repository, which shows about a 50% improvement over our previous testing. Finally, AMD has a DirectML fork of Automatic1111 that appears to be mostly tuned for the RDNA 3 GPUs we're testing.
Given what we've seen so far, it should be little surprise that the 4070 Ti Super lands right between the vanilla 4070 Ti and the 4080 in our Stable Diffusion tests once again. We're not using models that will benefit from the extra VRAM, but VRAM bandwidth does appear to be a bigger factor here, and the new GPU ends up closer to the 4080 than the vanilla 4070 Ti this time.
AMD's fastest consumer GPUs are still far behind what we're able to get from Nvidia's offerings. The results here mimic the heavy DXR results at higher resolutions, with the 4070 Ti Super beating the 7900 XTX by 81% for 512x512 image generation, and 66% for 768x768 images.
Nvidia RTX 4070 Ti Super Professional Workloads
We use SPECviewperf 2020 v3 to assess professional workstation graphics performance on the various GPUs. Nvidia only offers full performance tuning in its drivers for professional GPUs like the RTX 5000 Ada Generation, while AMD tends to be more forthcoming with driver tuning for its consumer cards.
The overall view of performance using the geometric mean of all eight tests puts the 4070 Ti Super just 12% ahead of the RX 7600, and well behind all the other AMD GPUs. In the individual tests, the best Nvidia's new GPU can do is to basically tie the RX 7900 XT in 3dsmax and trail by only 7% in Maya. But the real discrepancy is in Siemens NX, where even the 7600 is about seven times faster than the 4070 Ti Super, and the 7900 XT is over fifteen times faster.
For GPU-accelerated 3D rendering, the only app that currently supports all three GPU vendors is Blender. It now leverages the ray tracing hardware to boost performance, and that generally puts Nvidia in the pole position. We're using Blender Benchmark, version 3.6.0, though version 4.0.0 is also available. We've found some GPUs, including all of the Nvidia models, perform better with version 3.6.0, and it's not clear whether 4.0.0 executes the same workload or if it does extra calculations to deliver a better result.
As we've seen in our ray tracing gaming tests, Nvidia's GPUs come out far ahead of AMD's competing offerings. The 4070 Ti Super in this case outperforms AMD's 7900 XTX by 118% in our Geomean. Also, it's slightly closer to the 4080 than the 4070 Ti again, 12% slower than the former and 15% faster than the latter, so Blender appears to like the extra bandwidth as well.
Nvidia also has RT hardware acceleration for Otoy OctaneRender and V-Ray, but the other GPU vendors currently lack hardware acceleration for these benchmarks. As such, we've only tested Nvidia GPUs in these applications. Once more, the 4070 Ti Super falls between the 4070 Ti and 4080, and memory bandwidth looks to be a big factor. The new GPU is 22% faster in Octane and 20% faster in V-Ray (RTX mode, not CUDA), and only 7% slower than the 4080 in Octane and 11% slower in V-Ray.
Nvidia RTX 4070 Ti Super Content Creation Summary
Professional workloads aren't normally the primary use case for consumer GPUs. yes, you can run them, and some of them do just fine, but content creation professionals will usually opt for the additional support and performance offered by professional GPUs — or at least their employers will.
AI workloads on the other hand are mostly limited by one of two things: raw compute, or the amount of VRAM available. For workloads that fit within the memory of a consumer GPU, such cards are often faster than 'professional' solutions, since they have higher clocks and power limits. If you need more VRAM to hold the LLMs, however, that will generally require a high-end professional GPU to get either 32GB or 48GB of memory — with data center GPUs offering up to 160GB of VRAM.
- MORE: Best Graphics Cards
- MORE: GPU Benchmarks and Hierarchy
- MORE: All Graphics Content
