Question Reduced RTX 3080 graphics score after platform upgrade

[George9t8]

Hi all,
I recently upgraded my platform from:

Asus Z-390 Prime with i7-9700KF and 32GB DDR4 3200MHz CL16
to
Asus Z-790 Prime with i5-13600KF and 32GB DDR5 6000MHz CL32

My current platform supports resizable bar so that's enabled.

I've kept the same Asus TUF RTX3080 10GB.

In January I achieved a 3DMark Timespy graphics score of 18348, but now I'm consistently averaging around 18,000.

I undervolt my card to get a steady clock rate which was a constant 1950Mhz in both runs.

The only variable under my control is the graphics driver, so I installed one of the earliest versions that was compatible with Win 11 (472.12). This didn't make any difference.

This is a new version of 3DMark but I'd be surprised if this would result in different scores.

Even User Benchmark is reporting the card is performing below expectations: https://www.userbenchmark.com/UserRun/56922080, compared to an earlier run on the old platform: https://www.userbenchmark.com/UserRun/49457921.

I'm curious, is this difference not unexpected? What could be causing this regression in performance?

I would have expected the more modern architecture to perform better if anything.

Thanks.

 

[Phaaze88]

I'm curious, is this difference not unexpected? What could be causing this regression in performance?

It's because folks are constantly adding to the databases; the mean changes, for good or bad.

 

[George9t8]

It's because folks are constantly adding to the databases; the mean changes, for good or bad.

I'm not comparing to other peoples' score though, I'm just comparing to my own result.

Unless you're saying the calculated score is based on benchmark scores from all users at the time?

 

[sizzling]

Did you reinstall Windows after the platform change?

 

[George9t8]

Did you reinstall Windows after the platform change?

Yes, fresh install of Win11

 

[George9t8]

I manged to improve my score by setting Hardware accelerated GPU scheduling on.