LMG Last Video Comparing Mistake? Waiting Official Answer

[Nasosdem]

In the last video of LMG, which compares i3-7350k to a i5-7400 (downgraded i7-7700k) I found a mistake:https://www.youtube.com/watch?v=Qn_E_pI9GeY

In this video, i3-7350k is clocked at 5Ghz and they have a downgraded i7-7700k to i5-7400 (I assume it's running with "turbo boost" at 3.5Ghz for single thread and 3Ghz or higher for multi thread)..
In single thread, the result is 196 Cb for the i3 and for the i5 it is 151 Cb.
In multi thread, the result is 528 Cb for the i3 and for the i5 it is 436 Cb.

And here comes the problem. The assumed frequency I take for i5-7400 at multi thread is 3.0Ghz, which means 85.71% of 3.5Ghz, resulting in 151 x 4 x 85.71% = ~518. That doesn't make sense, it comes with conflict with the video right?
Especially since the Cb score of i5-7400@Anandtech is 554 (probably turbo boost is running the i5-7400 a bit higher than stock frequency, around 3.2 to 3.2Ghz):
http://www.anandtech.com/bench/CPU/1029

And some other source about Multi Thread score rating the Cb multithreading to 563:
http://www.cpu-monkey.com/en/cpu_benchmark-cinebench_r15_multi_core-8

And the score of i5-6400 is about 490-530 at different sites:
http://www.cpu-monkey.com/en/cpu-intel_core_i5_6400-527
http://www.trustedreviews.com/intel-core-i5-6400-review-benchmarks-page-2
http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/71760-intel-skylake-i5-6500-i5-6400-i3-6100-review-4.html
http://www.hardwaresecrets.com/core-i5-6400-cpu-review/6/

This means that either LMG did a mistake, either editing at video, or at the downgrade of i7-7700k, or the i7-7400 is slower than its predecessor (which comes in conflict with anadtech result and other sources)!

I am waiting for Linus or some Admininstrator to answer to both me, but also to any other who watched his video, and clear the issue. This is a big mistake, if done by LMG, and should be addressed with a good fix and fast.

[slowdeath008]

Other benchmarks seem out of spec based on other youtubers too. They must have gotten the simulation of the i5 7400 wrong.

[RGProductions]

@LinusTech @Slick 

 

And now for my 2 cents. I guess they didn't

t have turbo dialed in quite right. Some of this could also be a variance in the cinebench run, as I usually get around a 20cb difference sometimes.

[slowdeath008]

Yes but 436 to 560 is a long way

[MageTank]

They also tested all of the games at 4k, when the GPU was the bottleneck, and wondered why the results were basically the same. This is exactly why I tell people to check multiple sources when it comes to reviews. LTT's methodologies are getting worse every day. 

[Wolther]

Yep. Everything about this video seemed off and very weird.. Not only that, why'd they even play the games in 4k? If they wanted to see CPU bottleneck then they should've stressed the GPU as little as possible 

[Nasosdem]

2 minutes ago, RGProductions said:

@LinusTech @Slick 

 

And now for my 2 cents. I guess they didn't

t have turbo dialed in quite right. Some of this could also be a variance in the cinebench run, as I usually get around a 20cb difference sometimes.

Well, 10% is what turbo boost gives... And it's a long way from 436 to 550.

[RGProductions]

33 minutes ago, Nasosdem said:

Well, 10% is what turbo boost gives... And it's a long way from 436 to 550.

yeah.

[slowdeath008]

bump.

[Taja]

Copy paste of what i commented on the vídeo:

 

what is going on with techtubers? Kyle from bitwit tested the processors with a GTX 1060, thus the results were Limited by the GPU. here this did not happen, but you tested in 4k??? that makes no sense, this also creates GPU Bottleneck. All CPU tests should use the most powerful GPU and a lower resolution, like 1080p (or test all 3 resolutions) because this way you can really see what is happening. Just for example, If you test a game in 16k res, a 1080 Will deliver like 10 FPS with a Pentium or a i7. but this is not a good test, obviously.

 

 

[Prysin]

23 hours ago, MageTank said:

They also tested all of the games at 4k, when the GPU was the bottleneck, and wondered why the results were basically the same. This is exactly why I tell people to check multiple sources when it comes to reviews. LTT's methodologies are getting worse every day. 

@Slick@nicklmg@LinusTech Is become so unserious that they should just stop doing reviews. They are complete garbage when it comes to consistency and quality, not to mention being completely detached from their forums when it comes to feedback.

[KE2012]

To be honest, if you want reliable information LTT is not the place to go for it. While they are good at entertaining and giving a general review, they are abysmal at things that need a scientific / engineering approach. 

[Wolther]

On 1/15/2017 at 4:35 PM, KE2012 said:

To be honest, if you want reliable information LTT is not the place to go for it. While they are good at entertaining and giving a general review, they are abysmal at things that need a scientific / engineering approach. 

This. But people should be looking at multiple sources anyways. One source is never sufficient enough evidence; especially a YouTube channel