• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

UserBenchmark Bias

Soldato
Joined
24 Feb 2003
Posts
4,187
Location
Stourport-On-Severn
All this shows is how desperate Intel have become, which doesn't bode well for the future as they clearly think they have to play the game like this.

The thing is though @humbug, Intel as you well know, have always played "the game like this". No amounts of legal action and multi billion $ fines are ever likely to stop them.
What will stop them though is the buying public just moving to AMD as the cpu of choice. We and Intel know this has already happened in the retail cpu space. They obviously still hold something over OEM's because there is no proper business case for them to still be offering 90% Intel inside when AMD's entire range is cheaper and faster than Intel's.
 
Soldato
Joined
6 Feb 2019
Posts
17,403
It's an absolute joke they have become for this, changing the weight they assign makes sense but certainly not in the way they have been doing it. For example I can understand reducing the effective multicore score IF they increased the the quad core to octocore. But leaving the quadcore in there benefits no one but Intel, as modern programs as well as games increasingly uses cores past 4 (and more than 8 threads too). Leaving single core in there makes sense for applications that continue to use single core and allows you to compare the effective single core between different CPU's

The operators of that website are so incompetent they screwed Intel as well with the changes - not sure if it's still the case but at one stage UB was saying an i3 9100 is the best gaming cpu Intel has and that it beats everything else intel and amd makes
 
Caporegime
Joined
18 Oct 2002
Posts
33,188
It's also more recently come to light that they are now doing the same things with Nvidia vs AMD in GPU. A 5700XT is apparently slower than a 1080, despite every single review agreeing very much otherwise.

Userbenchmark are taking money to give better results to anyone who wants to pay for it. This is nothing new or surprising, neither is the fact that Intel and Nvidia continue to be happy to pay such sites for positive coverage. Ignore then and move on.
 
Caporegime
Joined
17 Mar 2012
Posts
47,309
Location
ARC-L1, Stanton System
It's also more recently come to light that they are now doing the same things with Nvidia vs AMD in GPU. A 5700XT is apparently slower than a 1080, despite every single review agreeing very much otherwise.

Userbenchmark are taking money to give better results to anyone who wants to pay for it. This is nothing new or surprising, neither is the fact that Intel and Nvidia continue to be happy to pay such sites for positive coverage. Ignore then and move on.

lol... just gone from a 1070 to a 5700XT, its at least half as fast again, damn near as fast as a 1080TI......

Is it that Intel and Nvidia are sending brown envelops? or is it that they just have some sort of hate for AMD, like some petulant people on the Internet sometimes do just have a brand religion and pathologically hate anything else.

With the latest drivers i could find...

https://www.techpowerup.com/review/sapphire-radeon-rx-5600-xt-pulse/27.html

u5qldmH.png
 
Associate
Joined
31 May 2020
Posts
2
good thing workstation users aren't idiots and wouldn't touch a 10900k with a 10ft pole

Nah, i would certainly benefit from a 10900k compared to a 3900x in my workloads(fea and cad +occasional gaming) it would be faster. Except, if you factor in that where i live it costs $200 more to buy a 10900k, it doesn't worth it anymore. It comes out to around 130 weeks of work to pay for itself... Bad deal.
 
Last edited:
Soldato
Joined
18 Oct 2002
Posts
19,354
Location
South Manchester
Soldato
Joined
29 May 2005
Posts
4,896
Both Passmark (Also had Intel favouring changes recently) and UserBench are right at the top of Googles sponsored links for "CPU Benchmark" search results, to get to the top of those sponsored search results cost mega money, far more money than these people will be earning from ad revenues.
For sure! Intel is paying for their Google fees. AMD could do the same but instead relies on internet reviews and word of mouth. I think not a great deal of self build people use those benchmark sites. I personally have never ever consulted these sites. It is pointless.

I buy the platform and CPU based on my budget and what’s best at the time. Not some stupid website tells me this is what or not. Benchmarks are fun thing to compare like for like overclocks and see where you tally up with your peers that’s pretty much it. It is only really useful for the same gen of CPU maybe even the same line. Definitely cannot be used for comparing different platform.
 
Soldato
Joined
29 May 2005
Posts
4,896
Before the Ryzen 3000 series came out, Userbenchmarks ranking system was based on a 30% single core, 60% quad core, and 10% multi core performance ratio.
Shortly after the Ryzen 3000 series came out, they “suddenly” changed their 10% multi core performance impact to just 2%, thus lowering AMD scores vs Intel overnight.

Pic21.jpg


This change got a lot of attention but instead of taking in feedback from the tech community to dial back on this heavy focus on single core performance, they instead basically called everyone who disagreed with them chills.
They even called out Hardware Unboxed for being "objectively incompetent smearers” likely after they did this video: https://youtu.be/AaWZKPUidUY?t=214
Now AMD has a CPU targeting every scenario of the score weighing. So where do these clowns go from there to help their intel masters? Lol
 
Associate
Joined
27 Mar 2010
Posts
1,461
Location
Denmark
Now AMD has a CPU targeting every scenario of the score weighing. So where do these clowns go from there to help their intel masters? Lol
In typical userbenchmark style they bash the 3300X, recommends the i3-10100 instead and calls 3700X owners unlucky :rolleyes:
This despite the 3300X is generally better in games, beats the 10100 clearly in productivity tasks and offers overclocking, higher frequency memory support and PCIe4 motherboards.

3uj26d1z.jpg
 
Last edited:
Man of Honour
Joined
19 Oct 2002
Posts
29,503
Location
Surrey
I thought most tech savvy people knew that site was a load of rubbish and wouldn't rely on it.

Unfortunately non tech savvy people could be drawn in by it.
 
Caporegime
Joined
17 Mar 2012
Posts
47,309
Location
ARC-L1, Stanton System
In typical userbenchmark style they bash the 3300X, recommends the i3-10100 instead and calls 3700X owners unlucky :rolleyes:
This despite the 3300X is generally better in games, beats the 10100 clearly in productivity tasks and offers overclocking, higher frequency memory support and PCIe4 motherboards.

3uj26d1z.jpg

Do these people just deliberately say the opposite of what's true?

9600K 6 core 6 thread @ 5.1Ghz: 244 FPS
7700K 4 core 8 thread @ 5.1Ghz: 244 FPS

3300X 4 core 8 thread @ 4.4 Ghz: 238 FPS
3100 4 core 8 thread @ 4.4Ghz: 209 FPS

The 9600K / 7700K are 116% the clock speed of the 3300X with 102% the performance.
The 3100 is the same clock speed of the 3300X but with 15% less performance, they are the same CPU, 4 cores 8 threads even the 16MB L3 is the same, the only difference is the 3300X DOES NOT suffer from architectural latency because it only has a single CCX.

imatw8r.png
 
Soldato
Joined
29 May 2005
Posts
4,896
Do these people just deliberately say the opposite of what's true?
I think they are referring to the architecture latency between AMD and Intel. the 3300X does have higher latency than the intel processors. this is inherent in the architecture and how the IO circuits work. if AMD architecture is similar to Intels we wont have these chiplets concept that are scalable etc.

They are not wrong but they really overemphethised this and completely ignore the benefit of using this type of architecture is that you have a highly scalable design that yields very little loss and cheap to manufacture.

in short, the hole at the bottom.
 
Caporegime
Joined
17 Mar 2012
Posts
47,309
Location
ARC-L1, Stanton System
I think they are referring to the architecture latency between AMD and Intel. the 3300X does have higher latency than the intel processors. this is inherent in the architecture and how the IO circuits work. if AMD architecture is similar to Intels we wont have these chiplets concept that are scalable etc.

They are not wrong but they really overemphethised this and completely ignore the benefit of using this type of architecture is that you have a highly scalable design that yields very little loss and cheap to manufacture.

in short, the hole at the bottom.

That's not what they said, they said the latency in the 3300X is a problem for gaming, it isn't...
 
Back
Top Bottom