News UserBenchmark bashes AMD GPUs and claims they lack real-world performance

[DavidLejdar]

There definitely is a massive benefit to having a more powerful CPU with middling graphics cards and that is shown in the 1% fps lows, and there are plenty of games that massively benefit from the extra cache such that if you don't have it the game is nearly unplayable in certain circumstances (Beyond all reason). Now, if you had 700 dollars for a CPU and a GPU, on average, you would be better off with a 200$ CPU and 500$ GPU which would exclude all X3D CPUs from consideration. AMD drivers have been par with Nvidia for nearly 10 years now. I still have many super annoying driver issues with Nvidia as we speak. CPU pro and GPU pro are extremely biased and consistently lie about AMD performance on both fronts, I would not call that a "bend."

Yeah, I can't complain, as the 9800X3D increased the low 1% (compared to a 7600X before). E.g. just yesterday, I happened to run a benchmark in Far Cry 5, and did have a minimum FPS of 97, at 3440x1440 with a RX 6700 XT. And likely to upgrade the GPU soon enough anyhow (where I may even see better performance with a 9070 XT, than someone who spends around double for a RTX 5080 and has a by UB previously recommended i5-13600K).

Sure though, on a budget, or simply just not into fast-paced gaming, that's money there, which can help to go for a higher tier GPU instead.

 

[Roland Of Gilead]

AMD should use a small percentage of their profits and sue that site for slander and put them out of business

Yeah, agree with you there. Why would they not though. It doesn't seem to be a thing for them.

purely based on disagreement of opinion.

Well, it's not just a disagreement of opinion. It's much more akin to either stupidity or as others have suggested some unknown issue where UB rag on AMD. Without really knowing though, it's conjecture.

but I do believe their is some brand vs brand bias inherent at userbenchmark.

some brand to brand bias? It's much more than bias.

 

[NickT300]

UserBenchmark is heavily pro Nvidia and Intel. They hate AMD and their bogus benchmark results are proof of this.
That site had a chance for Legitimacy, but choose to be very unreliable with its benchmarks.

 

[helper800]

It's like TDS, except for tech.

AMD GPUs were garbage some years back, I tried out a R9290X and what a piece of junk that was. Went back to Nvidia real quick.

Fast-forward to 2021, I grabbed a 6800XT during the crypto craze because it was available, and I knew it would pay for itself even at scalper prices. Never looked back since, it was a great gaming GPU and since then I've owned another 6800XT, 2x 7900XTXs and 2x 7900XTs, and been very happy with every one of them. And you don't want to get me started on how many AMD CPUs, normal and gaming laptops, and Minis I've owned. Every single one has been excellent.

My first dGPU was a HD 7950 from AMD and it ran flawlessly years before the 200 series cards. My best friend got a r9 290 at my recommendation and had no issues except that it ran hot.

 

[Roland Of Gilead]

My best friend got a r9 290 at my recommendation and had no issues except that it ran hot.

I remember those. Apparently, most of the overheating R9's were from really cheap thermal paste AMD used. Once re-pasted the card ran well below it's max temp.

 

[ohio_buckeye]

I will say on the cpu comments the guy makes in the article, he may have a point. AMD has done really well at marketing the x3d lineup.

Not to say this is every case, but many times people don't need to spend almost 500 on a 9800x3d for example when a cheaper 7700x or 9700x will perform similarly in some cases. That said, if you're running 4090 or 5090 or 5080 etc, yes you want that extra cpu power. But if you're running say a 7800xt or below, maybe the gains aren't as pronounced. As someone said, maybe you see it in 1% lows etc. For a casual gamer who plays a couple of times a week or on the weekends, or a PC that is doing a mix of tasks, maybe the mainstream part makes sense.

 

[bigdragon]

I really hate how prominently displayed Userbenchmark is in search results. Their information comes across as misleading when you look into what they say about different products. There's been a few times in the past where Userbenchmark (and Wccftech) fooled me with some incorrect information -- never again!

However, I don't fully disagree with them criticizing AMD GPUs. We need a stronger competitor to Nvidia, and AMD could be that competitor with some improvement. However, they do lack the stability and compatibility of Nvidia. AMD seems to have stability issues with Unreal Engine and performance limitations with Blender in particular. Their hardware seems good, but their software needs more attention.

 

[hotaru251]

"High average fps are worthless when they are accompanied with stutters, random crashes, black screens, excessive noise and a limited feature set."

i mean those are all better than literal melting connectors & fire risks on team green...

I have not had an AMD gpu since they still ati (ati radeon x1300 512mb back in early 2000's) but even I am thinking about going team red next as I don't do framegen (let alone MFG which is gross) and how amd actually gives you vram.

 

[salgado18]

This has nothing to do with AMD's quality. If their cards were really bad, none of them would receive positive reviews. That includes overheating, crashing, etc., even though some of them did receive remarks in these aspects (RX 480 was too power hungry, for example).

This is all about that site's malicious intent towards AMD. To take a card that is indeed a very good card and try to make it look bad is a deliberate act of aggression towards a brand. UserBenchmark is an anti-AMD site, period.

But the worst part is that, if it weren't for Tom's Hardware, I would never suspect a thing. I did use UserBenchmark in the past to compare some cards and CPUs, and it all looks too legit. How many people just Google stuff and end up on their site? How many people decide not to buy a Ryzen 9800X3D because they said so?

AMD should really take action. I don't know the legal grounds for this, but it is almost a direct attack at the company. Either that, or take action against Google for favoring such false content, misleading millions of people away from good products.

 

[joshglen]

Not that this makes a huge difference, but the current sample size is one on Userbenchmark for the 9070XT. I don't think that will fix too much when more results are uploaded but I think the percent difference has a reasonable chance of going down as more results are added

 

[salgado18]

However, they do lack the stability and compatibility of Nvidia. AMD seems to have stability issues with Unreal Engine and performance limitations with Blender in particular. Their hardware seems good, but their software needs more attention.

Nonsense. AMD's failure to compete with Nvidia is based on macro features, not micro features. RT performance was awful, FSR was (considered) worse than DLSS, AI performance worse, and even thermals were a bit worse. AMD cards, until 7000-gen, were only viable on rasterization, and even then because of a better price/performance, and even then are not selling well. When AMD had competitive cards in performance, they almost matched Nvidia's market share. And now AMD seems to be back in the big game (and Nvidia made a bad launch), it will sell a lot more. Issues in one or other software is often overlooked, except for users of said software.

 

[helper800]

Here is a perfect example of the complete BS of Userbenchmark... When these CPU's are actually benchmarked against each other you get this, 285k being 12.1% slower instead of just 1%:

Source.

 

[HardwiredWireless]

AMD has never been known as the top of the high performance GPUs merely the cheapest alternatives. Of course with Intel entering the picture they may lose that crown as well.

 

[cyrusfox]

some brand to brand bias? It's much more than bias.

Why the hyperbole? Bias perfectly describes the issue at hand

Bias (noun): prejudice in favor of or against one thing, person, or group compared with another, usually in a way considered to be unfair.

How would you describe it?

 

[jimi411]

His statement that there is no advantage in buying a 3D CPU if it is married to an average GPU is correct. The GPU will be the bottleneck. You need a top of the line GPU for the 3D CPU to shine. The majority of users do not own a fast GPU although that will increase over time as people upgrade. It is also true that AMD drivers are usually not as good as Nvidia's.

He may have an anti AMD bent but his points are not without some merit. Simply denouncing him as crazy just lowers the conversation.

Considering one of the primary perks of PC gaming is ease of upgrading it's kind of a null point. No, infact it's a counter point. You WANT your GPU to be the bottleneck. A top of the line CPU bought today will eventually bottleneck a midrange GPU though it may be several years in the future at which point you have to upgrade the entire platform instead of just the GPU.

A 5800x3d has kept the nearly decade old am4 platform relevant even with these newly released cards. Likewise a 9800x3d is unlikely to bottleneck a midrange GPU released in 2030.

 

[Confused_Counsel]

Unfortunately, there is a lot of misinformation out there. User Benchmark is just one example, but there are more. That is why independent review provided by Tom's Hardware, JayzTwoCents, Gamers Nexus, TechPowerUP, Daniel Owen, Hardware Unboxed, etc. are important.

I honestly wish these sites had a way to compare CPU and GPU directly - especially laptop configurations. Sure, the hierarchies Tom's Hardware provides for CPUs and CPU are nice, but they could be more interactive. Like letting a user search for a CPU / GPU and see the review scores or comparisons without having to scroll through a chart. Just a suggestion. Mr. Alcorn and Mr. Walton please make the hierarchies more usable and searchable.

 

[Confused_Counsel]

Considering one of the primary perks of PC gaming is ease of upgrading it's kind of a null point. No, infact it's a counter point. You WANT your GPU to be the bottleneck. A top of the line CPU bought today will eventually bottleneck a midrange GPU though it may be several years in the future at which point you have to upgrade the entire platform instead of just the GPU.

A 5800x3d has kept the nearly decade old am4 platform relevant even with these newly released cards. Likewise a 9800x3d is unlikely to bottleneck a midrange GPU released in 2030.

Yes the 5800X3d was one heck of strong chip. Hardware Canucks just did a video about how it still holds up today. Though the bottleneck calculators say differently, but then they always tell me I should have bought Intel... So, not really reliable either.

 

[Roland Of Gilead]

Why the hyperbole? Bias perfectly describes the issue at hand

How would you describe it?

No hyperbole at all. Just a reflection of UB and how they denigrate anything AMD. It's no surprise. It's been like that for years. There's a pattern there if you care to look.

I would describe it as unsavoury! Does that suit your narrative?

UB are not objective in anyway with regards to how they posture themselves in relation to AMD. They mislead users of their apps/benches or whatever, some of which pay for the privilege of running UB in Pro mode. It's deceptive at least, and a horrible practice in general.

But, look, the notoriety serves them well. While they continue like that, UB will be the ever running joke of the community.

 

[mrmessma]

... It is also true that AMD drivers are usually not as good as Nvidia's...

I don't know what to say except that having used nVidia and AMD (last 16 years) I've only had amd driver issues once through 6 cards and that was a single game's issue with crossfire.

 

[SpenReyn2]

UserBenchmark claims that AMD's Radeon GPUs fail to deliver real-world performance and are backed by an army of influencers it blames for duping customers.

UserBenchmark bashes AMD GPUs and claims they lack real-world performance : Read more

Not sure why any legitimate computer journalist would pay any attention to what UserBenchmark says. They should be ignored by everyone until they disappear.

 

[Typhon666]

You can tell by the responses that people don’t like the truth. Truth is the i7 and i9 are the most powerful. I have 3 gaming systems and I build them for a living. X3D still has micro stuttering issues. So at that price I feel you kinda get ripped off cause on my weak 12700k with just 8 cores enabled its beats my 7800x3d…

 

[ravewulf]

If you click the three dots next to the Google search result, there's an option to leave feedback flagging the site for inaccurate/misleading content

 

[Allan Walsh]

UserBenchmark claims that AMD's Radeon GPUs fail to deliver real-world performance and are backed by an army of influencers it blames for duping customers.

UserBenchmark bashes AMD GPUs and claims they lack real-world performance : Read more

I stopped taking that site seriously years ago. I love my AMD CPUs (currently the 9800X3D) but I also run with Nvidia GPUs (currently a 5080, but waiting on a 5090 coming). And no, I'm not rich! I won both of the GPUs in raffles after upgrading the rest of the PC, but I'm using the 5080 until the 5090 is back in stock. Lol
I've had no issues with the current card, and the ROPS are fine. I also like to undervolt to keep my hardware running longer.

 

[TerryLaze]

Unfortunately, there is a lot of misinformation out there. User Benchmark is just one example, but there are more. That is why independent review provided by Tom's Hardware, JayzTwoCents, Gamers Nexus, TechPowerUP, Daniel Owen, Hardware Unboxed, etc. are important.

They are not independent at all, they heavily depend on add revenue and clicks and will tell the story in any way that will get them more clicks and more comments.
The more out there or controversial they can make a story sound the more we will be sitting here arguing for days the more adds they will get by showing the add firms higher numbers.

 

[mac_angel]

But it does suck when it's compared to it's own previous gen cards. If it can't beat it's own 7900XTX in rasterization performance, then, yea, it lacks real world performance.
And your own quote, quoting them, "For context, it once recommended readers purchase a Core i5-13600K over the Ryzen 7 9800X3D, asserting, and I quote, "Spending more on a gaming CPU is often pointless."" I've often said your benchmarking and reviews of CPUs in gaming is greatly flawed. You take a top of the line graphics card, then run it at 1080p to benchmark the CPU. No one is buying an RTX 4090 or RTX 5090 and gaming at 1080p. If you have those cards, you're gaming at 4K. What kind of difference in fps do you have then between the 13600K (which is highly overclockable) and the Ryzen 7 9800X3D?
You never do real world benchmarks like that. How much of a difference would it be from the core i7 to the core i9 in those situations when the core i7 can overclock much more, and have a lot less thermal throttling?