I appreciate satire, and I appreciate self-deprecation, and I appreciate self-awareness.

But I still see Steve as being quite belligerent in this.

There was a time when it was a standard for reviewers to do 3 resolution testing (720, 1080, 1440 first, and 1080, 1440, 4k later on).

Steve was very calculated in this video - by choosing 4k only, on max settings, with RT, in known GPU-centric titles. It was a tongue-in-cheek contrived test to prove himself 'right' and say 'see I told you so', when he and everyone knew what the results were before any test was ever conducted.

None of the CPU-testing critics are asking to create GPU-limited-only testing scenarios for CPU-testing. None.

1080p testing on $4000 worth of hardware is still stupid and not representative of any real-world use case. Swinging in the complete opposite direction out of spite is not a good alternative either (though of course the video was done out of jest, not presented as a true solution).

I also understand that thorough testing is an impossible job, because every variable that is testable would literally multiply the amount of testing (labor hours) required.

We know 2 things;

• 1080p low testing with insane hardware just gives fantasy, exaggerated, unreal results. Might as well do 720p testing to really exaggerate those bar graphs.

• 4k max testing will commonly result in useless data. (not useless if the data is contextualized properly. Yes, if you are GPU limited and if a consumer is presented with those results it could inform them that a CPU upgrade is not necessary.)

The valuable information for consumers and enthusiasts is going to lie in the middle.

A happy medium would be two 1440p test runs, with the assumption that the end user would be aiming for satisfying a 120hz monitor.

Test Run 1 would be something like a 'low-medium' settings, or perhaps rasterization oriented, results. The idea being that the user is going to accept compromises on video settings so that their system can output a higher framerate. Perhaps we leave premium features like FG, DLSS, FSR, RT out of these tests.

Test Run 2 would be something like a 'medium-high' settings. The idea here being maybe the end user wants to dabble in, but not max out, some featuresets. Maybe the lowest level of RT, maybe DLSS quality, maybe the highest texture quality setting, high shadow quality, etc. (DLSS / FSR being critical here; the idea being can the CPU support the higher framerates, at higher quality, enabled by Nvidias software tricks which are now industry standard.)

So instead of testing the farthest left of the spectrum which is 1080p low settings, and instead of testing the farthest right of the spectrum which is 4k max gpu-limited, instead we aim to test the two counter balanced points in the middle (imagine a horizontal line divided into 3 equal parts by equidistant vertical lines).

The results would show tighter margins than 1080p low testing, and are unlikely to present GPU-limited results (but may indicate just how close we are to being GPU-limited). However, it just miiiiight expose some failures or successes of different CPU SKUs. Like, maybe a 9800x3d dominates 1440p with DLSS in a way that maybe a 12600k doesn't, or maybe the opposite is true and the 3D vcache is exposed as irrelevant past 1080p.

And of course, infinitely closer to real-world system use where everyone is making compromises, and no one is playing at absolute maxes and mins.

I feel that everyone would benefit from and be more informed by this testing than 1080p low testing. Even if this isn't used as de-facto product release review testing methodology, it's at least worth a 1-off video to gage the viewer appetite for this type of testing, or at least evaluate the state of mid-range gaming in 2025.

1080p low bar graph-maxxing is a masturbatory exercise.

If 'real world' testing tells us that the 7800X3D and the 9800X3D and the 14900k and the 285k are all equivalent, well then, so be it and thanks for saving us $500 and thanks for being the pioneer that was willing to do more than the bare minimum to combat mindless consumerism.