Apologies for the double-post but I didn't know of a good place to ask this...
So since this also means that Windows 8 and Windows 8.1 support is ditched, I've thought that it might actually be best to simply use either 5.0-16380 beta or 5.0-16391 dev for the updated CPU benchmark since being compatible with multiple OSes is arguably more beneficial and even more common for benchmark situations compared to real-world use-cases.
That being said, I would like a bit of insight - is there a preference on whether I use the 16380 beta or the 16391 dev version for such a thing? EDIT: I personally would prefer 16380 beta since, on Linux, that means one can just refer to the flatpak.
Secondly is the name - the existing benchmark was simply the "Dolphin 5.0 CPU benchmark" but that doesn't work for this, and it's certainly not 6.0 yet. I've thought about calling it the "Dolphin 5.7 CPU benchmark" to refer to it's compatibility with Win7 as well as the implication that it's a version between 5.0 and 6.0, and the theoretical implication that it's a version more similar to an elusive 6.0 than to the existing 5.0 since it's almost 6 years old now.
But I don't want to call it the "Dolphin 5.7 CPU benchmark" if the dolphin devs would prefer otherwise or if they have a better idea for a name.
Also, if there really doesn't end up being a "Dolphin 6.0" and Dolphin has basically switched to a rolling-release model, then this seems like as good of a time as any to put out an updated benchmark (I mean, it'll be the same "luabench" homebrew - just on a much newer version of Dolphin of course).
Additionally, I actually wouldn't be surprised if such a thing ends up being the final version of the Dolphin CPU benchmark since even modern U-series CPUs are getting fast enough that Dolphin's CPU performance is starting to become a non-issue, and modern CPUs are starting to lean more heavily on cache but Dolphin seems to be too old or light-weight enough or something to really take advantage of CPU cache (the 5800X3D doesn't seem to provide any benefit, and even the halved-L3 of the 5700G barely hurt performance compared to the 5800X) considering that more demanding emulators like RPCS3 actually do see benefit from the 5800X3D despite the CPU's lower clocks.
In other words, for future benchmarks, I may have to move on to more demanding emulators, especially since multi-threading above 2 threads is becoming more relevant in emulation due to the corresponding emulated console actually having multiple CPU cores and the like.
So since this also means that Windows 8 and Windows 8.1 support is ditched, I've thought that it might actually be best to simply use either 5.0-16380 beta or 5.0-16391 dev for the updated CPU benchmark since being compatible with multiple OSes is arguably more beneficial and even more common for benchmark situations compared to real-world use-cases.
That being said, I would like a bit of insight - is there a preference on whether I use the 16380 beta or the 16391 dev version for such a thing? EDIT: I personally would prefer 16380 beta since, on Linux, that means one can just refer to the flatpak.
Secondly is the name - the existing benchmark was simply the "Dolphin 5.0 CPU benchmark" but that doesn't work for this, and it's certainly not 6.0 yet. I've thought about calling it the "Dolphin 5.7 CPU benchmark" to refer to it's compatibility with Win7 as well as the implication that it's a version between 5.0 and 6.0, and the theoretical implication that it's a version more similar to an elusive 6.0 than to the existing 5.0 since it's almost 6 years old now.
But I don't want to call it the "Dolphin 5.7 CPU benchmark" if the dolphin devs would prefer otherwise or if they have a better idea for a name.
Also, if there really doesn't end up being a "Dolphin 6.0" and Dolphin has basically switched to a rolling-release model, then this seems like as good of a time as any to put out an updated benchmark (I mean, it'll be the same "luabench" homebrew - just on a much newer version of Dolphin of course).
Additionally, I actually wouldn't be surprised if such a thing ends up being the final version of the Dolphin CPU benchmark since even modern U-series CPUs are getting fast enough that Dolphin's CPU performance is starting to become a non-issue, and modern CPUs are starting to lean more heavily on cache but Dolphin seems to be too old or light-weight enough or something to really take advantage of CPU cache (the 5800X3D doesn't seem to provide any benefit, and even the halved-L3 of the 5700G barely hurt performance compared to the 5800X) considering that more demanding emulators like RPCS3 actually do see benefit from the 5800X3D despite the CPU's lower clocks.
In other words, for future benchmarks, I may have to move on to more demanding emulators, especially since multi-threading above 2 threads is becoming more relevant in emulation due to the corresponding emulated console actually having multiple CPU cores and the like.
Dolphin 5.0 CPU benchmark
CPU: Pentium G3258 @ 4.5GHz 1.24v
GPU: Intel integrated
RAM: 4x4GB Corsair Vengence @ DDR3-1600
OS: Linux Mint of some variety + [VM] Win7 SP1 x64
CPU: Pentium G3258 @ 4.5GHz 1.24v
GPU: Intel integrated
RAM: 4x4GB Corsair Vengence @ DDR3-1600
OS: Linux Mint of some variety + [VM] Win7 SP1 x64