p.2 #1 · LrC Benchmark tests of RTX 5080 and RTX 5090??
What are you doing that is causing problems after only 2 years? Bad mainboards sometimes happen, but just about everything else is an easy replacement.
p.2 #2 · LrC Benchmark tests of RTX 5080 and RTX 5090??
dclark wrote:
Thanks for posting the link to the Puget Systems benchmarks. I am familiar with their benchmarks and use them when configuring the systems I build. I think they are excellent.
The GPU differences are puzzling. They actually show the system with the 5090 performing slightly worse than the 5080. That raises the question, what can a 5080 do faster than a 5090? I can understand that their benchmarks are a weighted average of a bunch of LrC operations, so the impact of the GPU is diluted, but I don't see why the 5090 should be worse than a 5080. If anyone knows, I would like to hear about it.
I am pretty focused on 1:1 preview generation, which is an important part of my workflow and has only recently become dependent on the GPU, which has greatly speeded-up preview generation. The same is true for denoise. I don't have any idea how much those operations impact Puget Systems benchmarks, and I don't know if they are using the GPU for previews. ...Show more →
Preview generation now uses both the CPU and GPU, and a faster CPU can really accelerate it, as can one with more cores. For example, my M4 Max MacBook Pro has faster GPU and CPU clock speeds and a faster/newer GPU architecture overall, and it nearly matches my older M2 Ultra for preview generation.
In your case, if previews are that important than just upgrading your GPU will probably have a very minimal effect.
p.2 #3 · LrC Benchmark tests of RTX 5080 and RTX 5090??
jhapeman wrote:
Preview generation now uses both the CPU and GPU, and a faster CPU can really accelerate it, as can one with more cores. For example, my M4 Max MacBook Pro has faster GPU and CPU clock speeds and a faster/newer GPU architecture overall, and it nearly matches my older M2 Ultra for preview generation.
In your case, if previews are that important than just upgrading your GPU will probably have a very minimal effect.
Speculation may be interesting, but some real data would be more convincing. The problem is that no one is likely to have both a 5080 and a 5090 to do a comparison. If I were to upgrade my 3090 to a 5090 (or a 5080) I could produce some data, but I'm having trouble generating sufficient motivation to spend the money to upgrade.
p.2 #4 · LrC Benchmark tests of RTX 5080 and RTX 5090??
dclark wrote:
Speculation may be interesting, but some real data would be more convincing. The problem is that no one is likely to have both a 5080 and a 5090 to do a comparison. If I were to upgrade my 3090 to a 5090 (or a 5080) I could produce some data, but I'm having trouble generating sufficient motivation to spend the money to upgrade.
I think a lot of people have both, but maybe not at FM. I don't use LR or I could probably find a 5090 for testing. I would not be trying to interpolate or extrapolate any Apple data to nVidia RTX with x64 CPUs.
p.2 #5 · LrC Benchmark tests of RTX 5080 and RTX 5090??
EB-1 wrote:
I think a lot of people have both, but maybe not at FM. I don't use LR or I could probably find a 5090 for testing. I would not be trying to interpolate or extrapolate any Apple data to nVidia RTX with x64 CPUs.
EBH
I wasn't suggesting that, although if you go look at all of the historical mixed data I have shared--both Apple and PC with Nvidia cards--there's very little real difference between the two. This is pretty logical as the application is 100% functionally identical on both platforms and until recently both versions were using some of same CPUs and GPUs.
What doesn't change--Mac or PC--is that CPU core count and core speed matter a great deal for things like JPEG exports and preview generation. On both the Mac and PC side, the addition of GPU assistance to preview rendering is new, and I would venture to guess not yet fully optimized, as if you track GPU use, it never manages to fully saturate the GPU, unlike DeNoise. Either that, or some of the cores in the GPU simply aren't beneficial to preview rendering and therefore remain idle.
This is why I said OP wouldn't likely see a big gain from a 3090 to a 5090 for this task. To start with the gains between those two on the most intensive tasks is good, but not massive. But the fact that preview generation relies so heavily on the CPU and still 100% saturates every CPU core--no matter what your CPU is--means that if you want to improve that function, the CPU is the foremost place to seek performance gains--and in this day and age, a lot cheaper than an Nvidia 5090.
