We got reports in from user Ess.one that some of UL benchmark's 3DMark could be ran at other than the standard default benchmark when running a Custom run.

Nothing special, however these non default benchmark settings still get labelled as VALID when using the newer UL Benchmarks Systeminfo version 5.83 and 5.84 as you can see in the below snippit.

These 'tweaked" scores are totally inflated, messing up the current rankings. A normal 3DMark run at default settings would score around 17K.

After some emailing with The UL benchmarks support team we got informed of the fact that any Custom Run is not taken into account for their Hall of Fame.

This is also an easy rule to impose at HWBOT, yet we also feature specific CPU and GPU benchmarks in our benchmark suite.

So we have to adapt to the ongoing situation to avoid any abuse of the current ruling even before UL Benchmarks will provide a fix.

So depending on the benchmark we are forced to impose specific additional rules:

- For standard UL 3Dmark 3D benchmarks NO CUSTOM RUNS are allowed

- For specific CPU or GPU only Tests of the 3DMark 3D benchmarks we impose that default benchmark settings have to

remain applied, but a custom run is possible by disabling for eg the CPU test when e.g. only the GPU score is required.

- For TimeSpy Extreme (CPU), the default benchmark settings can not be changed, but one is allowed to CHANGE ONLY the

CPU Instruction set to for e.g. AVX 512.

As you can see in the snippit below:

Thanks to Ess.one for finding this unfortunate bug, fingers crossed a software fix from UL Benchmarks will prevent future misabuse.

So plz read-up on the NEW Rules before starting to bench, thank you