KitGuru KitGuru.net – Tech News | Hardware News | Hardware Reviews | IOS | Mobile | Gaming | Graphics Cards
Here we present a range of AI and productivity benchmarks, designed to offer insight into the sort of workloads that might be undertaken by someone purchasing this calibre of GPU.
For our AI image generation test, the first thing to know is the 5070 actually failed to run our usual Stable Diffusion XL test due to a lack of VRAM. We tried again using the Stable Diffusion 1.5 tests and this worked fine, though performance isn't able to match the RTX 4070 Super.
AI text generation fared slightly better, with up to an 11% gain over the 4070 Super using the Llama 3.1 model, though Llama 2 performance is slightly worse.
Geekbench AI continues to be a mixed bag, with single precision performance lacking, but the half precision metric shows vastly superior results, this time with a 9% lead over the 4070 Super.
There's not much to split the two GPUs in Blender 4.3, however, though Nvidia remains far superior to AMD's offerings.
V-Ray also shows more appreciable gains when using the RTX path, this time with an 18% lead over the 4070 Super, one of the biggest differences we have seen today.
