But, the 9070XT is a lower tier card (its a mid level card) with worse bandwidth and compute. 960GB/s vs 640GB/s and 64TFLOPS vs 40something FP32 on the 9070XT.
Depending on what you do in said project, those could matter.
Also variable framerate is not Resolve friendly, you should convert that, before editing, to constant.
DefinitelyNotAnAIBot wrote:I'm using project server to store all my projects so they can be edited from both of my workstations. ... Why is System 2 so much slower rendering the project?
To answer that question you first need to answer is it systems or is it "networking" that is impacting you. Move your project locally to each system, render again, and then compare times.
DefinitelyNotAnAIBot wrote:I'm using project server to store all my projects so they can be edited from both of my workstations. ... Why is System 2 so much slower rendering the project?
To answer that question you first need to answer is it systems or is it "networking" that is impacting you. Move your project locally to each system, render again, and then compare times.
Exactly could be a faulty cable going from 1000 > 10/100. Also if on Windows open task man and monitor resources and compare
Desktop: AMD 9950x, 64GB, AMD 5700XT, Couple NVMes. CachyOS (Arch) and Nobara (Fedora) Laptop: AMD 5700U, 32GB, Nobara (Fedora), Macbook Pro M3 Max
I stand corrected, it must be something about the network speed. Even though it never came close (according to Windows Task Manager) to saturating the gigabit network speed, bring the project local make the project render in about 9 minutes on the 9070XT.
My other system has a 25Gb direct connection to the NAS, which according to task manager, never exceeded 200Mb of usage.
I also noticed that power usage on the 7900xtx is typically higher during the export. Usually between 150 and 200 watts, but as high as 350 watts. The 9070XT typically runs around 100 watts.