Email from Puget Systems:
"We’re excited to announce the official 1.0 release of Puget Bench for Lightroom Classic. This is a major step in our mission to bring real-world performance testing to the creative industry. Puget Bench runs directly on your installed version of Lightroom Classic, using real photos and real workflows to deliver results that reflect how your system performs in practice. "
However I find this benchmark only partially meaningful. Most of it is Import Export Smart Previews and Library operations. That says very little about how fast a system really is when you actually edit photos. GPU heavy tasks like masking AI noise reduction and detailed local adjustments hardly show up in the scoring. At the end about 70 percent CPU performance and only about 30 percent GPU performance are weighted.
If you work with 61 megapixel RAW files and create HDR raws from twelve exposures you also need a lot of RAM. None of that is reflected in this benchmark and it can strongly influence real world performance with complex workflows. And if you work with fifty or more masks GPU memory becomes critical as well. Even with 24 GB VRAM it can get tight at times but this is completely missing from the benchmark.
For photographers who do heavy editing it is therefore difficult to tell from this test whether the GPU VRAM or the system memory might become the bottleneck. So I have mixed feelings about this benchmark. Apple users will probably like it because with this type of test their weaker GPU performance is not as noticeable.
One more thing that bothers me. The program does not uninstall properly. On my system it left about 30 GB of data behind. I suspected this might happen which is why I monitored the installation. Anyone who wants to try it should manually delete this folder as well:
And one more note:
“GPU for Preview Generation: Disabled” – why?
In Lightroom this feature is fully enabled for me, but the benchmark doesn’t use it. As a result, the GPU appears to be underutilized again.
Honestly, it’s a rather odd test. I’m really disappointed.
This benchmark is only useful for office-style Lightroom use, not for serious photography workflows.
