News UserBenchmark suggests you buy the i5-13600K over the Ryzen 7 9800X3D — says AMD drives sales with 'aggressive marketing' rather than 'real-world p...

[Scraph]

Actually, tons of people that are into competitive shooters or sim racing play as fast as the game can run on their machine for reduced latency.

And they are consumers, not developers, and don't understand that the same could be achieved without all that overkill.

A game can render one frame right before each refresh and also get low latency. It's a simple problem with an equally simple solution.

 

[stuff and nonesense]

The entire point of using those as a benchmark is that after that games can become GPU limited, obviously. But it's pretty simple to assume that a cpu isn't "optimized" just for the 1080p low benchmark, but that the CPU performance will scale up at higher resolutions with better GPUs.

Put simply, the best "1080p low" CPU is probably the best CPU for 2k/4k so that it's not dragging performance down on the system overall. Feel free to test this theory by putting a 4090 in a system with a core i3 2120, and compare 1440p benchmarks in any modern game of your choosing at high settings.

Spot on. A lower resolution with really fast gpu is used to make the GPU transparent to the test. This makes the limiting factor the CPU.

Likewise if you are testing GPUs take the fastest processor(s), your chosen samples of GPUs and test at the current range of resolutions. Results will be similar to: 1080 fast, 1440 as fast or a little slower, 4k dropping, 8k through the floor. (It’s not long since 4k was through the floor). The thing is that the CPU should be as transparent as possible when testing GPUs

 

[stuff and nonesense]

And they are consumers, not developers, and don't understand that the same could be achieved without all that overkill.

A game can render one frame right before each refresh and also get low latency. It's a simple problem with an equally simple solution.

And each year the AAA titles become more demanding. So as your “lower spec” purchase tanks and the “overkill” purchase remains relevant you find yourself buying a new cpu, cpu/motherboard combo.. Buying overkill if it is well priced isn’t such a bad idea.

 

[Kamen Rider Blade]

I wonder how much of Intel's "Marketing Fund" is going to pay these buffons to run their website?

 

[tony.dimarzio]

The argument here is basically that it doesn't matter if the 9800x3D is 500% faster than Intel's best because the 13600k and 14600k are already more CPU than anyone actually needs and there's zero reason to buy anything else. So, we've reached peak and AMD and Intel should stop making newer & faster stuff.

 

[SkyBill40]

How that heavily bias and laughably out of touch clown show continues to get pub and have people trust it is simply unbelievable.

 

[MergleBergle]

Spot on. A lower resolution with really fast gpu is used to make the GPU transparent to the test. This makes the limiting factor the CPU.

Likewise if you are testing GPUs take the fastest processor(s), your chosen samples of GPUs and test at the current range of resolutions. Results will be similar to: 1080 fast, 1440 as fast or a little slower, 4k dropping, 8k through the floor. (It’s not long since 4k was through the floor). The thing is that the CPU should be as transparent as possible when testing GPUs

Thanks! I mean, if CPU advancements weren't important as opposed to GPU ones, I'd be brawling through The Veilguard on my old 486 DX 100.

 

[Scraph]

But it's pretty simple to assume that a cpu isn't "optimized" just for the 1080p low benchmark, but that the CPU performance will scale up at higher resolutions with better GPUs.

Put simply, the best "1080p low" CPU is probably the best CPU for 2k/4k so that it's not dragging performance down on the system overall.

That would be pretty simple to assume. Too simple.

CPU performance doesn't scale linearly once you start implicating cache and memory.

Remember the X3D part in the 9800X3D name? The cache is why X3D is such a big deal. So, it's also a big deal when a benchmark interacts differently with the cache than your usage.

Measuring "1080p low" to predict 2K/4K performance is just silly. You can't really think that's a sensible thing to do?

If you want to predict 2K performance... crazy idea... benchmark at 2K.

If you want to predict 4K performance... another wild one... benchmark at 4K.

 

[Amdlova]

Think only way to justified a 9800 x3d is when you have a 1000 usd graphics or more...
Rest of the world is using a rx6600 or a 4060,
The best and affordable gaming gpus...

 

[stuff and nonesense]

That would be pretty simple to assume. Too simple.

CPU performance doesn't scale linearly once you start implicating cache and memory.

Remember the X3D part in the 9800X3D name? The cache is why X3D is such a big deal. So, it's also a big deal when a benchmark interacts differently with the cache than your usage.

Measuring "1080p low" to predict 2K/4K performance is just silly. You can't really think that's a sensible thing to do?

If you want to predict 2K performance... crazy idea... benchmark at 2K.

If you want to predict 4K performance... another wild one... benchmark at 4K.

Guess what, some sites do and they show diminishing returns as the resolution rises and the GPU bottlenecks.

The stronger CPUs have the potential to supply the upcoming GPUs with data and not immediately be a bottleneck for the GPUs.

Every system has its bottlenecks be it GPU, CPU or memory. When your PC runs out of steam, what you want to play won’t hit your desired frame rates then choose your upgrade path.

I know one guy who bought a new top end rig each product cycle just to be on the bleeding edge.. his choice - no criticism is implied. In the same vein who is anyone to say what someone should or shouldn’t buy?

 

[tony.dimarzio]

Thanks! I mean, if CPU advancements weren't important as opposed to GPU ones, I'd be brawling through The Veilguard on my old 486 DX 100.

DX4/100 crew represent

 

[stuff and nonesense]

DX4/100 crew represent

Far too modern. 6502!

 

[bgunner]

This is totally opinion based on my end, This article and charts, UserBenchmarks, were put together by Intel fans boys that are having issues with the fact that AMD has now found, for the time being, a solution that gives it unpearalleled performance in gaming scenario's .

Even TomsHardware "had" one of these type of writers on staff a few years back, maybe quite a few as the year slips my mind, but he soon went to the wayside after the article I read where it was not an objective view but forced his own views into the article for Intel fanboy-ism.

Other than the Intel's 13th-14th gen frying themselves, Intel had a good offering last gen, till the frying issues. There performance is not in dispute but there longevity is and this fear will steer those looking to upgrade away from those CPU's. This is a fact. These issues are also not taken into consideration with UserBenchmarks review or charts. I personally will not buy a 14th gen for another 3+ years till it is proven to be safe but by then there will be a different CPU with better performance and newer tech as it moves so quickly these days. 3+ years is going to put it well past the end of life for this CPU so why bother at this point.

Fanboy-ism has no place in benchmarks or reviews. The facts, just the facts matter. The conclusion is subjective but don't force your views on your readers without saying "Opinion Based". User Benchmarks is Opinion based and they have a flawed system to evaluate hardware. Toms Hardware does a good job and so does Gamers Nexus of being transparent and try to be objective and only give opinions in there conclusions, which is Opinion Based to start with.

Just my 2 cents on this article in discussion.

 

[Dntknwitall]

Geeze, they must be taking a stance like the mainstream media. Just make stuff up and say it like you mean it. Misinformation should be outlawed as it just creates ridiculous ideologies in the manipulative minds which leads to arguments and in some cases violence. Everyone with a brain should boycott these sites and push for strict Misinformation laws/rules on the internet.

 

[Dntknwitall]

Looking at the user benchmark ratings is hilarious the top 11 cpus are all Intel. AMD has better cpus than all 11 of them well 10 the 14900K is a beast or it was anyway. It's crazy what Intel fan boys will do to discredit the competition. User Benchmark, just give the scores that are deserved and stop lying to people, Seriously!

 

[MergleBergle]

a

DX4/100 crew represent

I forgot the 4 after the DX lol. MY BAD! Lots of Tie Fighter and Elder Scrolls: Daggerfall on that beast of a machine

2 button joystick attached to my SoundBlaster 16 port (never understood why the gaming devices were attached to the soundcard, but it is what it is), flew through space with the greatest of eaaaaase.. This Tie Pilot and his ATI 64 graphics carrrrd...

 

[jlake3]

Amusing that even UserBenchmark apparently can't defend the new Core Ultra series, seeing as they're comparing the 9800X3D to Intel 13th/14th gen instead and make no mention of the new chips in that screenshot.

 

[d0x360]

I have a 7950x3d lol.. it's no slouch and yes I've had gains at 4k in gaming with it vs the 5800x3d

As for that site... It's shady, their CPU ranking show as much to insinuate some kind of editorial influence in their ranks

They harp on the 7950x3d because 8 of the cores are slower than the normal variant then they trash gaming performance above 1080p using any 3dvc lol..

Yet ranked somehow lower is the "faster' 7950x lol... Of course with little elbow grease you can overclock that 7950x3d to 5.4ghz CCD0, 5.9 CCD1. All you gotta do is toughen up, set the curve optimizer up well then start attacking ECLK/BCLK values... Change from 1 to 1.1 lol

Then load up Alan wake 2 and enjoy it maxed out with DLAA & Path tracing at 4k80, dips into lower 70s sometimes but never below like 67 fps.

My 5800x3d and 5800x (which had a +16% all core clock OC) along with DDR4 overclocked to 6720mhz CL22 if memory serves (no pun intended)... couldn't get anywhere near that on the 4090 so... Eat it userbm

 

[ManDaddio]

So you're writing an article in order to slander some other company. Not cool.
What matters is the benchmarks. I've compared all my benchmarks to user bench and they really haven't been that much different.
Get off the slander train. It doesn't make you look good.

 

[Colif]

The only reason I look at Userbenchmark every new generation is to see how they make the Intel CPU still better than AMD

https://cpu.userbenchmark.com/Compare/AMD-Ryzen-7-9800X3D-vs-Intel-Core-Ultra-9-285K/m2356328vs4173

Sure, they equal... no difference.
It does depend how you test them though, the 9800x3d isn't a business orientated CPU so there will be tests its bad in.

We are hardly the only website to notice their BIAS, and its hardly new

They are constantly moving goal posts to make AMD look worse than Intel.

 

[stuff and nonesense]

So you're writing an article in order to slander some other company. Not cool.
What matters is the benchmarks. I've compared all my benchmarks to user bench and they really haven't been that much different.
Get off the slander train. It doesn't make you look good.

Even if the benchmarks are comparable (their numeric evaluations are not) the overview they give, the write up is completely biased. By their writings there is no such thing as a good AMD processor. They even skewed their benchmark weightings to negate the multicore advantage AMD hold in order to keep Intel looking good by the user benchmark numbers…. It isn’t a slander.

 

[MergleBergle]

So you're writing an article in order to slander some other company. Not cool.
What matters is the benchmarks. I've compared all my benchmarks to user bench and they really haven't been that much different.
Get off the slander train. It doesn't make you look good.

"Get off the slander train." Really? Have you read any of this other "review" websites reviews? They're garbage. They're not reviews, they specifically slander AMD at EVERY turn, to the point of not even reviewing cards or cpus put out by AMD, but just slamming them with copy/paste previous "reviews" they have done. It's a joke. But I guess that's kosher? PLEASE.

 

[wwenze1]

I wish reviewers would start reviewing power-limited settings (*wink wink*)

14700k beats 9800X3D in many multi-core synthetic benchmarks but it also uses 75% more power... and my tower cooler / casing / my room / my safety / my sanity forbids me from going over 150W. Heck, my 5600G with stock cooler is set to 30W.

 

[Colif]

They think every other website gets paid to promote AMD over Intel.
Fact they call AMD Advanced Marketing Devices in some of their reviews shows they aren't neutral in any sense of the word.
The intel subreddit banned all mention of the site as even they could see it was lies.
The hatred is of anything AMD make. GPU included

its just a shame they at top of search results. Can always hope the Google AI doesn't use them as a source.

 

[Copasetical]

Other reviews don't seem to support this theory. Then again, UserBenchmark has long been criticized for being an Intel fanboy. Despite his (weak) arguments refuting this, his reviews do indeed tend to show a pattern. Take it with a grain of salt. UB sure does know SEO.