Intel Core i5-10600 sample manages higher bench result on UserBenchmark than AMD Ryzen 5 3600 despite overall lower test scores

[Redaktion]

A Comet Lake Intel Core i5-10600 sample has been spotted on UserBenchmark, where it managed an "outstanding" bench result and scored particularly highly in the single-core test. However, even though its overall score is lower than the average total score of the AMD Ryzen 5 3600, the Intel chip was awarded a higher benchmark percentage.

https://www.notebookcheck.net/Intel-Core-i5-10600-sample-manages-higher-bench-result-on-UserBenchmark-than-AMD-Ryzen-5-3600-despite-overall-lower-test-scores.461737.0.html

[william blake]

userbenchmark guys are kinda intel fans
the 2 useless things on this site-
-USER RATING
-BENCH SCORE
too bad they are always in sight, always first in any table or comparison.
other than that its all fine, just check points for single or multicore to compare things.

[Mark Renton]

boy! does that benchmark verdict make little sense!

[RicoVIking9000]

Y'all commenting on this, perhaps you should remember that gaming favors single core performance and high clock speeds over core count. Yes, these are desktops, but there is like a 2% gaming FPS difference when you compare the laptop i5 9300H and the i7 9750H, and the latter is certainly "more powerful" than the former. There is a reason why the steam surveys are in favor of Intel; not all games, in fact not many games at all are optimized to perform better with more cores at lower clock speeds.

Also, if anyone actually bothered to do their research before commenting, literally 10 months ago, in June of 2019, the description for the 3600 says that the 9400F even performs better in games, and by no means will the 9400F be better than the 10600. Guys, it isn't hard to do your research, because there's actual gaming performance benchmarks that favor the Intel for a reason. Just because you don't like the Intel, or just because numbers in some tests show the AMD to be higher doesn't mean it performs higher in what the article is about, which is overall performance. And overall performance is not video editing, because this is an i5 and 3600 we're comparing, which is not what people will prefer for extreme tasks anyway. Perspective and context does matter. Yes, the AMD offers better multi core performance, but at that point, the real comparison is between the Intel core i7 chips and other AMD chips in that ballpark

I see some of the same people commenting on numerous articles with some comments with no evidence backing their statements, just disagreement, which is not constructive at all

[william blake]

Y'all commenting on this, perhaps you should remember that gaming favors single core performance and high clock speeds over core count

you missed the point. 3600 is leading in ALL THREE categories, normal, heavy and extreme. but they turned it into 10600 lead somehow.
a>x
b>y
c>z
but
a+b+c<x+y+z
isnt this amazing?
btw nobody is talking about exactly games.

[resn]

Most workloads including games do not utilize more than 4 cores. So in real works the Core i5 is expected to outperform ryzen 5 in most scenarios. The userbenchmark is thus more realistic than cinebench. Although it is always best to see the benchmark for individual application that you care about instead of a general benchmark like userbenchmark or cinebench.

[william blake]

Most workloads including games do not utilize more than 4 cores.

fyi 4 core cpus are not good in games anymore.

So in real works the Core i5 is expected to outperform ryzen 5 in most scenarios.

"real"? like what?
btw check this test, only real world applications, the test package is made with some help of a bunch advanced users after a long discussion.
https://www.ixbt.com/platform/amd-ryzen-5-3600-3600x-intel-core-i7-8086k-test.html (sorry, in russian, use g.translate or something)
keep in mind(and i see you missed it), i5 frequency is 4.3, not 5+, and both has 6 cores/12 threads.
so, i5 is expected to lose in basically ALL, except browsing.

The userbenchmark is thus more realistic than cinebench

yep, and it says exactly what i said,
i5 is expected to lose in basically ALL, except browsing.
183:193 in 1-2 core light workloads
642:645 in 3-4 cores medium
955:1042 in all cores extreme
on top of that, i5 consumes more and requires new, outdated at the day of birth, socket.
zen 2 cores is superior, and userbenchmark hates this fact so much they even denied their own test results.

[albert_1]

To the fanboys tooting the "gaming doesn't use more than 4 cores" horn: welcome to 2020, where 4 cores are the bare minimum, not cutting edge desktop chips. Not even close. No, not even in games. Please find actual arguments that didn't lose validity years ago.

On the 9400F edging out the 3600 in gaming: just no. Even the 9600K is trading blows with it, then again uses double the power once you overclock it and needs a beefy cooler, so it's not really a fair comparison, either. The total cost of the Intel part will be at least $100 higher considering the higher power requirements and the cooler extra cost. We didn't even talk about buying 3 different socket motherboards for the same CPU gen lol.

Finally, just look at the benchmark numbers. The way they account to the final percentage has been twisted quite a lot now. For those unaware, it all started with Zen 2, last June, once AMD started to rise to the top of the rankings in this joke of a benchmark site. The funniest thing about this is, before the fact, Intel 18 core extremes have been sitting on top of the charts. That wasn't a problem though, I guess - as long as it blue, it can stay. Just use the wayback machine to find out for yourself and for some laughs. For even better lulz, check out their CPU bench method description in the wayback machine. The projected shilling accusations are getting more and more hilarious over time.

[Lolman]

reddit.com/r/hardware/comments/g2uf7a/userbenchmark_has_been_banned_from_rhardware/ Press F

[Thibault]

Do you guys realize that the CPU score also depends on the memory latency score, which is not displayed here, right ?

Or you just wanna say UserBenchmark bad ?

[happydaydreamer]

Shame to say that the testresults is clearly biased toward the Intel CPU. The AMD scores better overall in every test and cost 35% less.
Interesting to see that the overall score for Intel is better, even if the results clearly is worse. A recommendation score should also be based on TDP and memory latency and considering this the AMD is probably two years ahead of Intel.
Have been buying Intel the last years myself, but so happy to see AMD is back so strong!