User Benchmark totally manipulating already misleading results!!!

[SpyderTracks]

Skip to 3.30 on this video, just another reason to avoid sites like userbenchmark.com

They totally manipulate the findings, but this time they've really gone too far, it's just rediculous.

 

[Stephen M]

That benchmark stuff has got to be some of the most biased junk I have seen and as a former editor I have seen a lot of biased junk.

 

[SpyderTracks]

That benchmark stuff has got to be some of the most biased junk I have seen and as a former editor I have seen a lot of biased junk.

It’s crazy, isn’t it. Yet an awful lot of people use the site as a valid comparison tool.

 

[SpyderTracks]

Man, I love Steve, just the way he's so brutally honest. You gotta check his surmisation of the UserBenchmark debacle:

Skip to 08:04.... brilliant! He just loses it saying the same thing, blames it on getting over sickness. Love it

 

[debiruman665]

single core is a literal myth, it just simply never happens in real life. Unless you are running the chip in as a dedicated chip for software you've written at the assembly level you'll never see a modern operating system ever run in single-threaded mode

 

[Oussebon]

Single core really isn't a myth.

Ofc there'll be at least a little load on other cores, from the software or other programs running.

But there's plenty that's reliant on the performance of a single core/thread.

Steam appears to encode its game backups on a single thread. Some music production I gather. Certain games. All ultimately dependent on single core performance.

 

[debiruman665]

This is highly specific to the i9 as its the only one I've tested with.

Windows with nothing running just idling will have the processor set to the 7 active core multiplier and then drop down to the 8 core multiplier under heavier load.

Since active cores aren't able to run at different clock speeds at the same time they all drop down to the lowest common denominator.

If all cores are at 1% usage then your chip will run at that highest active core multiplier speed.

I've tried countless experiments and asked around on other boards. the chip will never run in single active core speed (50x) unless you set the multiplier on all cores to 50x and let throttling handle the heavy load situations.

 

[Oussebon]

Ah, so what we're talking about here is that Intel's i9 9900k single core boost clock rating is a myth under load in the Octane chassis? Just to check I'm with you there.

 

[debiruman665]

Ah, so what we're talking about here is that Intel's i9 9900k single core boost clock rating is a myth under load in the Octane chassis? Just to check I'm with you there.

Can you screenshot your xtu settings and then show me your computer running at the single active core multiplier to test my hypothesis?

 

[Oussebon]

My PC's overclocked and the CPU just runs as 4.5GHz because I've told it to. I don't use XTU.