So I'm just now finding out UserBenchMark is widely considered biased against AMD

[FlorbFnarb 537]

News to me.  Sounds like it's been bad for a couple years now.

[Folterknecht 2,257]

Would be nice, if you made it clear what you 're talking about.

[FlorbFnarb 537]

7 hours ago, Folterknecht said:

Would be nice, if you made it clear what you 're talking about.

I ran across people saying it on Reddit, somebody linked this video.  

 

I'm hearing a lot of people say they've even changed their scoring system to bias it towards single core performance, and this started once AMD started beating Intel in multi-core performance.

[Folterknecht 2,257]

Don't know the site, don't use the site - don't care, especially considering the name. "Average" user and usefull benchmarks are two things that exclude themself.

[FlorbFnarb 537]

2 hours ago, Folterknecht said:

Don't know the site, don't use the site - don't care, especially considering the name. "Average" user and usefull benchmarks are two things that exclude themself.

It shows up basically at the top of the google rankings a lot of times; I've googled "compare GPUs" and had it pop up at the top of the list, so they get a lot of traffic.

It sounds like they're just biased as hell.

[Folterknecht 2,257]

https://www.techpowerup.com/review/asus-radeon-rx-6500-xt-tuf-gaming/31.html

 

with interactive graphs

 

https://www-computerbase-de.translate.goog/thema/grafikkarte/rangliste/?_x_tr_sl=de&_x_tr_tl=en&_x_tr_hl=de&_x_tr_pto=wapp

...

[FlorbFnarb 537]

44 minutes ago, Folterknecht said:

https://www.techpowerup.com/review/asus-radeon-rx-6500-xt-tuf-gaming/31.html

 

with interactive graphs

 

https://www-computerbase-de.translate.goog/thema/grafikkarte/rangliste/?_x_tr_sl=de&_x_tr_tl=en&_x_tr_hl=de&_x_tr_pto=wapp

...

Yeah, I think somebody on Reddit said Techpowerup was a good alternative.

[MagicalFlyingFox 1,022]

Userbenchmark has been a meme forever.

[Ezz 1,770]

No you're a meme.

Next gen Intel worth a look?

[MagicalFlyingFox 1,022]

19 hours ago, Ezz said:

No you're a meme.

Next gen Intel worth a look?

I think its been 1 year since I played a battle of WoT!

Yeah, next gen Intel is good.

[Fulcrous 2,073]

userbenchmark was never relevant even if this was not the case.

[FlorbFnarb 537]

On 2/28/2022 at 10:47 PM, Fulcrous said:

userbenchmark was never relevant even if this was not the case.

Really?  It's always near the top in google searches on the subject.  When I google "compare GPU" the first result I get is userbenchmark.

[Folterknecht 2,257]

On 3/6/2022 at 1:45 AM, FlorbFnarb said:

Really?  It's always near the top in google searches on the subject.  When I google "compare GPU" the first result I get is userbenchmark.

You can "buy" your placement in the search results.

[FlorbFnarb 537]

On 3/8/2022 at 5:09 AM, Folterknecht said:

You can "buy" your placement in the search results.

I thought those were clearly marked as ads though?

[Folterknecht 2,257]

17 minutes ago, FlorbFnarb said:

I thought those were clearly marked as ads though?

No. 

I don't know how it works, but I know there are thousands of  advertising/marketing companies that focus mainly on optimizing positioning in online search results and user reviews. Google itself also somehow has fingers in that pie making money directly and indirectly with companies competing for the top spots.

How exactly it works I have no idea, as 'mm not interested in that topic. Maybe someone else here can explain it or point you in the right direction.

[FlorbFnarb 537]

3 minutes ago, Folterknecht said:

No. 

I don't know how it works, but I know there are thousands of  advertising/marketing companies that focus mainly on optimizing positioning in online search results and user reviews. Google itself also somehow has fingers in that pie making money directly and indirectly with companies competing for the top spots.

How exactly it works I have no idea, as im not interested in that topic. Maybe someone else here can explain it or point you in the right direction.

Hmmm.  If I google "bed sheets" some of the first results are marked as ads, but if I google "compare GPUs" then none of them are labeled as ads.

I know search optimization is a thing, but if Google accepts money for higher rankings without marking them as ads, that's just deceptive.  Dishonest.

[Folterknecht 2,257]

Lets stick with userbenchmarks. I ve been active in the PC world (online) for more than 15 years. Before you brought that site up, it never was on my radar. Might be that I stumbeled over it in the past, but I put it aside as useless - the whole idea and presentation is nonsense for everyone who knows a thing or two about benchmarking and PC building.

This site has no buisiness being on top of any generic search listing when it comes to usefull benchmarks. I'm running on a ~7 year old installation of W7 that 's totally cluttered. Do you think that is good datapoint in any performance benchmark? Or how about the PC noob that bought his box at Walmart, half of the SSD filled with useless bloatware, while Norton and McAffee fight over dominance in the background? Sure you can say that averages out over a big number of compareable systems, but that still doesn't make a valid datapoint when it comes to spending hundrets or thousands of dollars, let alone the amounts companies spend on PC hardware.

 

Just think about the XMP-thing you discovered - I wouldn't be suprised if 50% of PCs outside of comercial use in the US that use non-JEDEC RAM don't have XMP enabled. On some systems and with some applications it's not relevant, on others ... .

[FlorbFnarb 537]

3 hours ago, Folterknecht said:

Lets stick with userbenchmarks. I ve been active in the PC world (online) for more than 15 years. Before you brought that site up, it never was on my radar. Might be that I stumbeled over it in the past, but I put it aside as useless - the whole idea and presentation is nonsense for everyone who knows a thing or two about benchmarking and PC building.

This site has no buisiness being on top of any generic search listing when it comes to usefull benchmarks. I'm running on a ~7 year old installation of W7 that 's totally cluttered. Do you think that is good datapoint in any performance benchmark? Or how about the PC noob that bought his box at Walmart, half of the SSD filled with useless bloatware, while Norton and McAffee fight over dominance in the background? Sure you can say that averages out over a big number of compareable systems, but that still doesn't make a valid datapoint when it comes to spending hundrets or thousands of dollars, let alone the amounts companies spend on PC hardware.

 

Just think about the XMP-thing you discovered - I wouldn't be suprised if 50% of PCs outside of comercial use in the US that use non-JEDEC RAM don't have XMP enabled. On some systems and with some applications it's not relevant, on others ... .

Oh, so their method is just to average out user-reported performance or something?  Yeah, that sounds like a pretty bad method.

[MagicalFlyingFox 1,022]

Their scoring also caps out at like 6 cores for CPUs too