An AMD Ryzen 9 3900XT desktop processor sample has been put through its paces on UserBenchmark, where it managed some decent scores and a respectable average bench result. However, the score of 91.8% would put the 12-core Matisse Refresh Ryzen 9 3900XT in 28th place in the UserBenchmark table, which is even behind the quad-core Intel Core i3-10320!
https://www.notebookcheck.net/Ryzen-9-3900XT-sample-gets-highest-average-bench-result-for-an-AMD-processor-tested-on-UserBenchmark-but-it-s-still-behind-over-25-Intel-chips-including-the-Intel-Core-i3-10320.477390.0.html
I can't believe NotebookCHECK would write news about UserBenchmark. Their ranking system is ridiculous.
This article is so bad, it feels paid for. Parroting that the i3 is better than the ryzen 9 sounds like the article from a little over a year ago saying that the 8700k beats the 2700x by 50 something percent in gaming
I assumed there was an implied /s at the end of the article.
UserPoopStain is a load of garbage. I hope this article was trying to get that point across, but I feel it wasn't heavily enough implied for the normal user who may not be super into the tech world. The Ryzen 9 3900XT is laughably more powerful than the I3, it's not even a comparison. Userbenchmark not only rates everything on single core, but then also massively weights memory latency for when AMD chips have better single core than an Intel variant. It's just so obviously biased to the point that everyone needs to call them out on their crap. Some type of intervention needs to be done here because they are either delusional fanboys or completely paid off by Shintel.
Even a 3300X can run every gpu at full load at 1080p high settings or better, and only starts to be bottlenecked compared to an Intel cpu at 1080p medium settings or lower. So 1080p high, 1440p, 4k, even a 3300X is equal to a 10900K with a 2080super, and will only be 8-15% behind even with a 2080ti. You can see all this right on Gamersnexus youtube channel. So any Zen 2 cpu is just as good gaming as any Intel cpu for the 2080super or lower graphics card, at 1080p high settings or higher resolutions. And the 3700X or better cpu's will also match with a 10900K with a 2080ti in most modern games at 1440p or higher resolutions. It just makes no sense to recommend Intel when all they are better for is 1080p medium and low settings gaming, which makes no sense with a 2080ti outside of 240hz gaming which is rather niche. But at normal resolutions, AMD is just as good at gaming, while also providing much more multithreaded power for the money. Userbenchmark is a steaming pile of 💩
My Ryzen 3 3300X performs way better 8), Single core performance is the king.
This is boatload of s*** couldn't find the report button, notebookcheck you've led everyone down this time didn't expected that from you , it's all paid in my opinion cause everyone knows about the shady tactics intel is pulling of in order to prove amd bad. #Boycott intel #boycott notebookcheck
There are some sense mofos commenting...
"Lastly, UserBenchmark results, especially AMD/Intel comparisons, require all the salt you can find to pinch at while reading them."
If that wasn't enough to make it clear they were mocking userbenchmark, I don't know you would ever get it.
Find it shocking that many here don't get the sarcasm in the articles.. indeed most geek are like sheldon cooper
This is useless and a shame that notebookcheck publish this userbenchmark crap
Why are you guys complaining? Although I am a big AMD fan, Intel still bests AMD in single core performance (which is still important, no matter what anyone says). Just look at GeekBench 5 results, it will just confirm these findings here. AMD needs to get on Intel's level with single-core performance, and if/when they do, Intel will bite the dust.
1) Relax people, the article is making fun of UserBenchmark, nobody at Notebookcheck thinks it's reliable.
2) Nobody thinks that that the i3 can compete with an R9.
3) Given the majority of posts here clearly showing they haven't read the article (why come here then?), it may be worth to consider getting your message across either in the title or the short summay that follows. Yes, not everyone is super into the tech world, and I couldn't blame those who'd take both a title and an article summary at face value. When Renoir chips are obliterating Intel CPUs with twice the TDP, is it such a long shot to assume the same thing the other way around (especially how it was the norm for the better part of a decade)?
I read the article and yes, sarcasm is dripping from it, and I'm not advocating for boring objectivitity all the time. An article can be entertaining while getting its message straight up-front; that's what the summary is about.
Quote from: Flomotion on June 25, 2020, 09:15:10
Why are you guys complaining? Although I am a big AMD fan, Intel still bests AMD in single core performance (which is still important, no matter what anyone says). Just look at GeekBench 5 results, it will just confirm these findings here. AMD needs to get on Intel's level with single-core performance, and if/when they do, Intel will bite the dust.
Nobody set their PC to use 1 core for any tasks. We use all system resources to finish our workload.
Quote from: Nikhil Sharma on June 25, 2020, 06:14:13
This is boatload of s*** couldn't find the report button, notebookcheck you've led everyone down this time didn't expected that from you , it's all paid in my opinion cause everyone knows about the shady tactics intel is pulling of in order to prove amd bad. #Boycott intel #boycott notebookcheck
Can you not tell the sarcasm from this article? ......
Anyone who can't tell that this article is criticizing the heck out of userbenchmark here is unable to read at a basic level. It's that simple.
Quote from: Orlynow on June 25, 2020, 12:54:59
Anyone who can't tell that this article is criticizing the heck out of userbenchmark here is unable to read at a basic level. It's that simple.
Absolutely agree. I honestly cannot comprehend whatsoever how one might read this text and not see how it comprehensively criticizes the ranking system and the results presented. The same goes for calling it "sarcasm", though. There is no trace of sarcasm here, just a well written, clearly worded and matter-of-fact examination of the bias visible in these rankings.
Not only are many not able to properly understand this article but they and NoteBookCheck seem to be unable to understand UserBenchmark ratings. The Bench % score is a gaming rating. UserBenchmark claim is that most games show very little benefit after 6 cores/threads. Their memory latency score under normal is an important factor in their % Bench (Game) rating. If you are interested in the performance in highly threaded workloads they give that information also. Bur their claim is most users seldom have much use for moar cores in excess of 4-6. This may not meet your requirements but for most it is likely the case. While I find their rating system somewhat opaque it is not as ridiculous as this article would have you believe. Chill a little.
Just making sure we all know that UserBenchmark is Intel biased and likely literally took a payout from Intel to change their numbers to favour Intel CPU's.
We all remember that, correct? Ok, good. Lol, UserBenchmark can kiss my a**.
Are you for real? You're actually taking userbench seriously, please delete this entire 'review' you absolute joke.
A have the 3900xt and i can say, its not the fastest single core but! I managed to set it on 1,15 volts by 4,2GHz all core OC. Single core in passmark is 3000points and thats more then enough for gaming. The CPU takes about 100Watt by 12 cores 24 threads!!! Runs max to 60celsius. Absolute winner. There is no intel CPU with these preferences..