Userbenchmark does more shady stuff

[GDRRiley]

Userbenchmarks as done more ridicules stuff by calming a 3700X gets beaten by a 9350KF in 5 modern popular games. The 3700x also bottlenecks a 2070 super.

Ryzen 4800HS should only be used with an RTX 2060 as anything faster will be crippled by irs higher latency. 2 further lines strait from that page

Quote

With low power consumption and high core counts, the 4000 range, on paper at least, is a perfect fit for the datacenter. AMD should focus on delivering a platform that offers performance where end users actually need it rather than targeting inexperienced gamers with the same old "moar cores" mantra

 

My thoughts:

My god userbenchmark that some shady stuff. Ether your benchmarks are so incompetent or you really like that intel money. I want to see them removed from the top of google and be killed off. this is just a bunch of misleading and shady work by them.

 

Note: not a joke if you check their pages they do say it. little late for april fools guys

 

[mr moose]

I haven't watched the video, but is this just another one of those cases where if you select the right measurement metrics you can get anything to sit at the top?

 

The last couple of times people tried to claimed benchmarks were biased it turned out to be user error.

[GDRRiley]

1 minute ago, mr moose said:

I haven't watched the video, but is this just another one of those cases where if you select the right measurement metrics you can get anything to sit at the top?

 

The last couple of times people tried to claimed benchmarks were biased it turned out to be user error.

this is mostly about their writings below processors making 0 sense. a 3700X doesn't bottleneck a 2070 super in any normal configuration but they claim it does.

a 4800H/HS shouldn't be used with anything faster than a 2060 as it has to much latency

[Jurrunio]

13 minutes ago, GDRRiley said:

Ryzen 4800HS should only be used with an RTX 2060 as anything faster will be crippled by irs higher latency. 2 further lines strait from that page

Quote

With low power consumption and high core counts, the 4000 range, on paper at least, is a perfect fit for the datacenter. AMD should focus on delivering a platform that offers performance where end users actually need it rather than targeting inexperienced gamers with the same old "moar cores" mantra

 

Since we've already seen how Intel's 8 core are slower than AMD's 8 core when limited to the same power target (cuz laptop, can't just slap a huge cooler onto it), that means Intel CPUs cant even be paired with anything that says RTX on laptops without CPU bottleneck then. Thx userbenchmark

[mr moose]

That does appear to be a pretty stupid thing to say,  I'd like to know what other reviews say about that (I haven't looked at the 3700X because it's not in my sights as a purchase). 

[Fasauceome]

more performance in "5 popular titles"

 

those titles being CSGO, fortnite, pubg, overwatch, and GTA V. All games cherry picked to benefit intel. On top of that:

The only tangible difference is in CSGO. otherwise, performance is within testing error basically (not that testing error means anything on this site, since there's no controlling for variables.)

[Techstorm970]

Cinebench or GTFO, honestly...

I mean, the 1,2,4, and 8-core performance benchmarks on CPU Userbenchmark are actually decent, but when it comes to GPUs or any other CPU metric, they're absolute CLOWNS and always have been.

[mr moose]

2 minutes ago, Fasauceome said:

more performance in "5 popular titles"

 

those titles being CSGO, fortnite, pubg, overwatch, and GTA V. All games cherry picked to benefit intel. On top of that:

The only tangible difference is in CSGO. otherwise, performance is within testing error basically (not that testing error means anything on this site, since there's no controlling for variables.)

It actually says comparable in performance not more performance than.

 

 

Quote

real world gaming performance comparable to the $80 USD entry level 4-core, 4-thread Intel Core i3-9100F.

 

And these are user generated stats not controlled tests.  So it maybe that more of the 9100F users have 2080ti's (because they are gamers) and those who have the 3700X have an earlier vega (because they are content creators). 

 

 

[Fasauceome]

5 minutes ago, mr moose said:

It actually says comparable in performance not more performance than.

Actually the scores I snipped were of the 9350KF. The 9100F scores swing notably in favor of the 3700X on userbenchmark's site. Either way the cherry picking is real.

Quote

And these are user generated stats not controlled tests.  So it maybe that more of the 9100F users have 2080ti's (because they are gamers) and those who have the 3700X have an earlier vega (because they are content creators).

I sincerely hope people aren't tossing a 9100F with two 2080 tis

Edited April 11, 2020 by Fasauceome

[PyroTheWise]

@GDRRiley Thanks for the heads up on the article.   I have watched a few of their videos and often wondered about their results.  They seem to not always be in line with lots of other testing/benchmarking I have seen on other sites/videos. 

 

This does open my eyes to the site.

[mr moose]

5 minutes ago, Fasauceome said:

Actually the scores I snipped were of the 9350KF. The 9100F scores swing notably in favor of the 3700X on userbenchmark's site.

I sincerely hope people aren't tossing a 9100F with two 2080 tis

I didn't say 2 2080ti's,  in fact I wasn't actually trying to be specific but just trying to highlight that many gamers skimp on the CPU so they can put more money into the GPU, which (for better or worse) would be reflected in userbenchmark results on the games listed.

 

EDIT: just to give an example, my son has a i3 4XXX, that combined with a 1050ti gives reasonable results in csgo.  Does this make the haswell i3's better than ryzen7 (fuck no).

 

Also I think this topic is getting confusing as there are several different issues being discussed and the page for the 3700X doesn't actually compare to the 9350KF.  Maybe it's just because there is not a single point of reference and we are all looking up the userbenchmark site for different things relating to the issue.

[Fasauceome]

1 minute ago, mr moose said:

the page for the 3700X doesn't actually compare to the 9350KF. 

I was referring to the video in that regard, the statement from userbenchmark suggests that the 9350K is a superior gaming chip in every way.

[williamcll]

Why go on userbenchmark when you get can all your CPU comparisons from Linus Tech Tips.

[GDRRiley]

9 minutes ago, williamcll said:

Why go on userbenchmark when you get can all your CPU comparisons from Linus Tech Tips.

cus GN is better than LTT at detailed testing

[Godlygamer23]

28 minutes ago, GDRRiley said:

cus GN is better than LTT at detailed testing

Yeah, GN is definitely better, especially now with their cooler tests. Their in-depth testing is actually pretty great on all sides. Can't wait for them to get into power supply testing.

[xAcid9]

I hope some big tech tuber do a video explaining why website like these is not to be fully trusted so i can just point to their video instead of explaining to these poor lazy souls every single fking time. 

[GDRRiley]

1 minute ago, xAcid9 said:

I hope some big tech tuber do a video explaining why website like these is not to be fully trusted so i can just point to their video instead of explaining to these poor lazy souls every single fking time. 

hardware unboxed has a few

 

[leadeater]

2 hours ago, mr moose said:

And these are user generated stats not controlled tests.  So it maybe that more of the 9100F users have 2080ti's (because they are gamers) and those who have the 3700X have an earlier vega (because they are content creators). 

Anyone with a 2080 Ti that also has a 9100F is probably right up there with the worlds biggest idiots. If you want an Intel CPU fine, just not that paring lol

[mr moose]

3 minutes ago, leadeater said:

Anyone with a 2080 Ti that also has a 9100F is probably right up there with the worlds biggest idiots. If you want an Intel CPU fine, just not that paring lol

Again, I was just making an example of how pairing a cheaper CPU's  with bigger GPU's would effect userbenchmarks results.   Remember when the Pentium was heavily reccomended in budget builds so people could get better GPU's? 

[xAcid9]

19 minutes ago, GDRRiley said:

hardware unboxed has a few

 

How tf did i missed this. 42mins though lol. Thanks. 

[Guest]

35 minutes ago, xAcid9 said:

I hope some big tech tuber do a video explaining why website like these is not to be fully trusted so i can just point to their video instead of explaining to these poor lazy souls every single fking time. 

You don't have to worry about that when you have a handful of overclockers on this forum that benchmark competitively. 

One thing we all need to remember is how difficult it is to be non-bias with giving and receiving information without favoritism. 

Also the term "bottleneck" is used way way too loosely to give proper meaning to Electrical hardware. 

[i_build_nanosuits]

the video doesnt seem to say what you say it does in your post what so ever...it says AT LAUNCH the best GPU available in those ryzen 4000 LAPTOPS is the RTX 2060'' which is by all means a MID RANGE GPU...which makes those lap top...mid range at gaming...i dont see his point.

 

is there something i'm not getting right?!

[GDRRiley]

2 minutes ago, i_build_nanosuits said:

the video doesnt seem to say what you say it does in your post what so ever...it says AT LAUNCH the best GPU available in those ryzen 4000 LAPTOPS is the RTX 2060'' which is by all means a MID RANGE GPU...which makes does lap top...mid range at gaming...i dont see his point.

 

is there something i'm not getting right?!

userbenchmark says the best GPU you should run with 4800H/HS is a RTX 2060 as any other would be hurt by ryzen bad latency which is BS

[i_build_nanosuits]

9 minutes ago, GDRRiley said:

userbenchmark says the best GPU you should run with 4800H/HS is a RTX 2060 as any other would be hurt by ryzen bad latency which is BS

well that would greatly depend on what game you test and the settings including rendering resolution and framerates target...

i can find situations easily where even my 5 ghz 8700K is preventing a gtx 1080ti to deliver more frames...it happens all the time...so a 4800H? in a laptop...i dont see that driving high demand games that well.

[GDRRiley]

18 minutes ago, i_build_nanosuits said:

well that would greatly depend on what game you test and the settings including rendering resolution and framerates target...

i can find situations easily where even my 5 ghz 8700K is preventing a gtx 1080ti to deliver more frames...it happens all the time...so a 4800H? in a laptop...i dont see that driving high demand games that well.

its faster than intels top end mobile chip at a low wattage. therefore if ryzen shouldn't run above a 2060 nether should intel.

Well maybe it is time to up your res and lower your targets.