Notorious UserBenchmark is already throwing shade at AMD's new Ryzen 7 5800X3D

I'll be happy with my Ryzen™ 9 5900HX..

Not sure why its emphasized as if Alder Lake running DDR5 is kind of an advantage, when it is definitely not for gaming, enough benchmarks prove that.

On 14/04/2022 at 05:52 PM, ice-dogg said:

Not sure why its emphasized as if Alder Lake running DDR5 is kind of an advantage, when it is definitely not for gaming, enough benchmarks prove that.

Because usdefewgewgewcchark are Intel shills.

not a fan or try to defend userbench

"has turned sour towards AMD and anyone who praises" yeah...no, they just don't care about mainstream opinion or vendor marketing; they even go against Intel's own products too

They don't sugar coat anything & it's entertaining to read that summary. What do you find wrong with it ?

there's no IPC changes -true

some -fps vs 5800x -true

some +fps vs 5800x -true

canned , sponsored AMD titles showing huge gains -been true for a while with Ryzen vs Intel numbers, especially with NDA saying how to review said product

On 14/04/2022 at 06:02 PM, tjtremor7865 said:

not a fan or try to defend userbench

"has turned sour towards AMD and anyone who praises" yeah...no, they just don't care about mainstream opinion or vendor marketing; they even go against Intel's own products too

They don't sugar coat anything & it's entertaining to read that summary. What do you find wrong with it ?

there's no IPC changes -true

some -fps vs 5800x -true

some +fps vs 5800x -true

canned , sponsored AMD titles showing huge gains -been true for a while with Ryzen vs Intel numbers, especially with NDA saying how to review said product

Are you the UserBenchmark site owner? /s

Btw, yes the 3D V-cache that AMD has added is also contributing to IPC improvement as it is indeed a part of the architecture. If some one says gaming performance gains isn't part of IPC changes then clearly they are lying or they are uninformed.

On 14/04/2022 at 06:02 PM, tjtremor7865 said:

not a fan or try to defend userbench

"has turned sour towards AMD and anyone who praises" yeah...no, they just don't care about mainstream opinion or vendor marketing; they even go against Intel's own products too

They don't sugar coat anything & it's entertaining to read that summary. What do you find wrong with it ?

there's no IPC changes -true

some -fps vs 5800x -true

some +fps vs 5800x -true

canned , sponsored AMD titles showing huge gains -been true for a while with Ryzen vs Intel numbers, especially with NDA saying how to review said product

https://www.youtube.com/watch?v=AaWZKPUidUY

https://www.trustpilot.com/review/www.userbenchmark.com

https://www.reddit.com/r/build...benchmark_should_be_banned/

But sure, you're not trying to defend them, lol. I honestly don't see what people use this site anyway. The percentages are super misleading and don't make sense.

in my experience, userbenchmark is biased toward Intel

TL:DR - UserBenchmark sucks - The 3D cache isn't a gimmick. Some games won't see any improvement because ###### devs. Some games will see significant improvement.

The 5800X3D does not clocking as high as the 5800X - mentioning this just to emphasize the extra cache is not a gimmick. So, even in games with crappy code, the cache is doing enough to keep pace with the higher clock of the standard 5800X. Again, not a gimmick.

Userbenchmark's take is just silly and technically ignorant.

----

"UserBenchmark"

I wouldn't trust them to time an egg.

Why?
1) Benchmarks tools have issues
2) Benchmark ratings are shifted by the site.
3) Benchmarks and ratings are shifted by users 'votes'. The 'popular' dumb wins the spelling test.

When a CPU that hits 90% of the speed of another CPU at the same or lower price - and it showcased in their 'performance' listing much higher on the list - there is a problem.

Even if they are 'clearly' explaining how their 'user' benchmarks work - most people don't read the fine print - and most people don't understand the fine print when it comes to benchmarking/performance.

When it comes to benchmarks - consistency of testing - very little or no editorial narrative - testing that holds up to real world metrics. There are a few sources I will trust for a general metric -as their testing almost always hits really close to highly calibrated internal testing we have done over the years.

Be careful who you trust and be willing to abandon them when they screw up or change.

If I was going to send people to a site for fairly accurate metrics - go to Passmark's cpubenchmark and videocardbenchmark. Their single core and multi-core numbers hold up to real world experiences for both gaming and heavy workloads. Raw data without inane/insane editorializing. (I have nothing to do with them, and if their numbers start to suck, I will be the first one here to call them out.)

Userbenchmark is bad - avoid. I also say this without regard to the Intel/AMD drama that I see mentioned in surrounding them. They suck, so no matter their spin, it just doesn't matter.

Geekbench also sucks - in their attempt to do cross platform testing - they are too stupid to realize they are OS optimizing. Any good OS engineer will do a spit take when evaluating how Geekbench works. - Followed by uncontrollable laughter at the stupidity.)

PS The 12900KS is expensive for the small gains over the 12900K - this makes the 5800X3D or the 12900K a better value for the upper end pure performance gamer.

I don't trust UserBenchmark for anything whatsoever.