Clevo Inferior GPU's?

Can someone please check if these results are actually the same one every notebook or if they are just caused by 1 faulty model?


The sager has the highest frames per second, which is expected from a desktop 7700K but why on earth is it so terribly unstable compared to the others? The min is half of the other notebooks, which every CS:GO player would hate with passion.


Here we got a GPU intentisve game, and compared to all others the frames are really low. This is a massive 50FPS less than the others????


Direct X 12, which uses the GPU more than the CPU made the game way more GPU intensive and again, the performance is down the drain? This time it isn't as bad as for honor, but again, noticeably less than the others.


Same story here, Direct X 12 less performance than the others. (when fixing some settings the Acer predator had same result as the GX from asus)

When switching from DX12 to DX 11 1080p on tombraider and mankind devided the performance is better on the sager, but why on earth is the performance dropping down when making the GPU load bigger? Can someone confirm or show that the testsystem might be at fault here, I refuse to believe that the GPU's in the Clevo are inferior to the others. Because of this is actually not caused by only 1 unit but all the others, that would mean the sager is inferior to the others in 4k and direct x 12 gaming, which is a huge deal. If you got one of those games, could you share your results?

All games are maxed out 1080p. Please give us feedback what results you get.

 

well, first question: what kinda hardware config are we talking about here? is it an apples to oranges comparison? stock vs stock or stock vs overclocked?

 

Unfortunately I don't have any of these games, but I can say that the benchmarks that I have ran with my clevo p870km1-g SLI 1080s varied from their other test that these screen shots came from

Sent from my SM-T560NU using Tapatalk

 

Also it was a 6700 not a 7700 I believe

Sent from my SM-T560NU using Tapatalk

 

That is a video made by linustechtips

 

It's clear the MSI was throttling in the CPU tests because ICCMax wasn't even set from the default. But I guess this was more of a cancer firmware test anyway.

 

My result from their in-game benchamrk:

1.)Rise of the Tomb Raider - Overall Score - 166FPS at a settings of DX12 Very High w/ stock gpu clock and cpu running at 4.5ghz on all cores.

2.)Deus Ex Mankind Divided - 98.4FPS at a settings of DX12 Ultra w/ stock gpu clock and cpu running at 4.2ghz on all cores. *Note for some reason I'm getting less fps i.e 94FPS when cpu is running at 4.5ghz on all cores in this game!

 

@ jaybee83

it's a 7700K, GTX 1080SLI setup. CPU Oc'ed to 4ghz as far as I can tell.

@cj_miranda23:
Can you post a pic of those benchmarks?

Seems that tomb raider is running as it should, but some problems in to Mankind devided? Do you have CS:GO and maybe for honor?

 

Just watched the video, it's a 6700k

Sent from my XT1585 using Tapatalk

 

Here you go!




Unfortunately I don't have CS GO and For Honor. Tomorrow I'll try testing with different type of SLI rendering. Both games are benchmark using NVIDIA recommended SLI setting.

 

Oh yeah. It's not like there is any difference between the 6700K and 7700K except for the more steamy temps anyways. Also which tests were you refering too?

@cj_miranda23:
Min frames of 61.75 fps on tomb raider 1080p???? Oo

Seems like mankind divided got a bit of frames. Interesting.

 

I am pretty sure that's where the info came from

Sent from my XT1585 using Tapatalk

 

I'm not following, you said you didn't have any of those games but the other tests differed on your build? Pretty sure cinebench should be slighlty higher given the fact that you're running a 7700K on 4.2ghz but other than that there werent other benchmarks?

 

AIDA64 and cinebench, and yes after watching the video again I realized it was a 6700 so yeah my results would vary

And as far as AIDA64, I'm just talking about the tests they did to see throttling

Sent from my XT1585 using Tapatalk

 

It's no secret that clevos cooling isn't the best. Also the only notebook that throttled was the Acer, even tho that got fixed later on in the video. It seems that miranda has some issues on tombraider considering that his FPS min is on 61.75, but mankind devided seems ok, even tho the moment he OC's to 4.5ghz it gets under 94FPS on average.

I start to think, that this might be a power draw issue.

 

Power Limit Throttling maybe is the main culprit for the result. Not knowledgeable about this things so maybe some experts i.e. ( @Prema @Mr. Fox @Papusan @hmscott @Meaker@Sager) here in this forum can enlighten us noobs hehehe!

 

Still, I am wondering about the CS:GO result. The min FPS kinda throws everything off, it's not GPU intensive, but CPU intensive only, why would the game run so unstable? It really bothers me.

 

CS:Go wouldn't power limit throttle.
MSI was 100% power limit throttling in the CPU tests if the clockspeed decreased there. Easily fixed by changing ICCMax.

 

Those results being as varied as they are for what is very similar hardware tells me either

Linus' testing methodology is slack
And/or
Stock config (firmware, paste job, etc) on these laptops is terribly unpolished

I think it's a little column A, little column B.

They all have 1080 sli. It's not rocket surgery to do a gpu only bench to see if the raw graphics power is there, and what clocks can be maintained over the long term.

But to answer op the 180W tdp limit on the clevo 1080s I'm pretty sure is the highest of all 4.

 

Despite the issues that some have it is actually the best, we are just cooling a higher class of hardware.

 

The MSI has 200W TDP on their 1080's but I do NOT know if the 1080 SLI has 200W TDPx2 or not.

*EDIT*
No video bioses anywhere to even edit and look at for the TDP.

 

It does. In fact, much more than that with overclocking. Even if you had the vBIOS files from all of them it wouldn't be relevant. You can put whatever TDP you want in the vBIOS and it means nothing if the system firmware power limits everything.

Almost anything from Linus in terms of notebook testing is going to be flawed. One search is all it takes to show which notebook rules. https://www.3dmark.com/search

Pick any benchmark you want and Clevo is kicking ass and taking names. Totally dominating everything else with stock vBIOS. Add a Prema vBIOS and the carnage is even greater.

Some clown is probably going to quote or reference this post with silly comments about "real world" nonsense and a make lame excuses about synthetic benchmarks not being relevant. Take it however you want to. The numbers don't lie.

Clevo gimps their stuff like all the other idiots that manufacture notebooks, but not gimped as bad as all of their competitors. It's also nice they use real CPUs with a full dose of TDP instead of soldered toy crap CPUs that are poorly binned TDP-deprived trash.

 

From what I could tell of the video, it mostly came down to CPU. In particular most of the machines using the 7820HK are overclocked by default (with a select-able profile in most cases) that puts them at ~4.2ghz which is a small bump over the 6700K stock 4.0ghz across all cores.
I imagine a stock 7700K version of the P870DM-G would close that gap (assuming it doesn't have a dud IHS that will cause it to throttle at stock speeds). I think they mention the MSI Titan is using a stock CPU profile which is why it tails behind the rest in most of the tests.

Rather simple explanation really....

Another thing to keep in mind, is when they're all stacked up, the P870 is easily the smallest of the lot. It's the only one of those machines you'd have a remote chance of sticking into a regular backpack and travelling with.

 

@Danishblunt you have got your answer right there. if anyone to purchase the laptop as is at stock, all seems okay and thats really the problem isnt it. OEM/intel purposely put in throttling, high voltage in order to control the speed of your laptop which you bought with your own money.

HOWEVER, anybody, putting in a little bit of effort to learn hardware, overclocking etc a clear winner comes out which is clevo. every other laptop doesnt have a modded bios and has all these restriction, only clevo has prema bios, at least for now.

i know what your favourite argument "clevo doesn't have the best cooling". well cooling dont matter if all the other laptop overheat and restricted, without good modded bios you dont even get the options to adjust thing, it is simply an irrelevant comment.

what mr. fox said is the takeaway and hope people like you and many others blinded by BGA can open their eyes to the truth.

 

I definitely agree and particularly in the case of Tomb Raider and Deus Ex in-game benchmark, at least in my case clevo cooling solution doesn't affect result since I never reach 75C but still getting a lower FPS compare to other laptop in the Linus video.

I wish I can get the @Prema firmware/bios even if I have to pay for it.

 

>taking LTT seriously

kek

 

I simply can't agree with this. Sure there are some units that actually have a decent cooling solution like the P650SE but way to many clevo models are really bad, the entire P1xx series was terrible with the cooling and even the P750 series has bad cooling as well.
https://www.notebookcheck.net/fileadmin/Notebooks/Schenker/XMG_U507_2017/stresstest2.jpg
100c CPU, 90c GPU, that is NOT acceptable especially for the amount of noise it's producing.

Meanwhile the tornado F5 which is MSI based runs at 97c CPU and 68c GPU, despite still having a high CPU temp, at least the GPU temp is really good.
https://www.notebookcheck.net/fileadmin/Notebooks/Eurocom/Tornado_F5/maxstressturbo.png

Same GPU, same CPU. No excuses here.

There are WAY more clevo notebooks with major cooling issues than other brands out there. There is absolutely no excuse for a notebook that far exceeds 3k EUR to have 90c on the GPU.

 

We agree on this.

I will add that there is no excuse for any notebook, any size, any brand, regardless of CPU or GPU, to ever run above 80C at full stock TDP, at max turbo/boost speed stock performance with no throttling, artificial firmware gimping or cancerware algorithms to aid in cooling or power reduction. If they lack the technical and engineering competence to do that the product should not be allowed to be sold.

The P7 model family is a good example of a thermal management abortion. It should be eliminated from the product line or be fixed before they sell any more of them.

 

that you outta blame both intel/nvidia and the OEM. OEM taking TDP seriously and working together to further control high performing hardware probably also one of the reason intel uses toothpaste in their CPU.

you also have to know most of these hardware can be undervolted and they are purposely restricting it with crippled bios/vbios, undervolting helps with temperature A LOT.

 

Furmark and prime 95 blender? That's like buying your family car based on its lap time round the nurburgring.

I think you see people on here with issues and assume that is the whole picture.

 

I see people here with issues, personal experience and reviewsites showcasing the issues. You're basicially saying at this point, you're completely ignoring the facts I showed you. despite same hardware the Tornado F5 had considerably better temp on GPU and better temp on CPU. Even prema talked about how he dislikes the cooling solution of some of the high end clevos due to them combining the GPU and CPU heatsinks, which causes the GPU to getting heated up by the CPU, which in that example is the case.

You cannot possibly imply with a straight face that people just happen to get bad models. If I were jumping to conclusions like you want me to, then I'd already be complaining about how terrible the power management of high end clevos with desktop CPU's are due to 1 benchmark test of linustechtips, which I don't, hence this thread. Apparently the same issues are happening for other users as well with underperformance with the desktop version, what exacly is causing this has yet to be figured out.

Note, I'm not saying all Clevo notebooks have terrible cooling solution. I'm just saying they aren't exacly known to have the best. There are some, like the P650SE, which got perfecly good cooling solutions, then there are P1xx series and the P7xx series that got a disastrous cooling solution. Most are sadly rather in the "meh" range.

I'm also not saying that you can't fix it, I'm saying that if I pay over 3k EUR for a notebook, then I don't understand why I have to rework the cooling solution to actually get acceptable results with insanely noisy fans, there is absolutely no excuse for this. And if it turns out, that the EC or whatever Is causing the GPU's to powerthrottle as well, then I honestly see some rework which has to be done, it cannot possibly be acceptable to pay a massive premium price for unfinished hardware that has to be tweaked by yourself and people like Prema to fix things issues.

 

i see your double standard here. pointing out people's issue but when others letting you know the pros side of thing, simply ignored. just being ignorant? i think not.

 

The results from the Cinbench was also screwed up.
Clevo P870DM2 reviewed by Anandtech.com with Stock 6700K and normal Cinebench R15 score for([email protected] all 4 cores). The score is about correct for 4.0GHz.


From my own testing with [email protected] all 4 cores(CleP870)


Edit. [email protected] all 4 cores

 

I have to agree. I've only witnessed a lack of ownership of Clevo's crappy heatsink problems. There are particular vendors that care enough to mitigate the issue which is good, but it needs come off the production line right and with consistency the first time. I hear Clevo doesn't care to take on these problems for a variety of reasons that boil down to - excuses. Until Clevo step up there's no way in hell I'm dropping 3-4k on another machine. Should they improve I'm definitely behind them but they've got work to do.

 

http://www.tieba.com/p/5352798808
clevo possibly re-using heatsink/chassis and redsign the new mobo same size to incorporate the new CFL 6 cores, this will prob run hot so 5ghz is possibly out of the question unless a miracle chip of some kind.

well some good news for current owners they can simply replace mobo + cpu to upgrade to 6 cores its very likely possible, BGA laptop dont got access to this. @Papusan @Mr. Fox @TBoneSan @manyothers

 

That is super good news, but I am not thinking I want to spend any more money on laptops in the future. I think my money would be better spent on a monster desktop instead. I am just getting tired of the extra effort to fix incompetent engineering and manufacturing practices and my 5.2GHz overclock limit is starting to be too boring. Too many limitations with laptops, and I expect that to get worse rather than better. The OEM/ODM global clown posse is too infatuated with their sadistic fetish for gimping their mobile crap.

That being said, it is kind of nice owning the one and only model of notebook capable of this kind of performance... check out the #1 notebook scores with "inferior" Clevo GPUs, LOL.

Spoiler: Fresh results...

Yup, this is with my "LGA turdbook" as described by @Danishblunt - really horrible ranking in the top 30 competing against liquid cooled desktops.

 

heres some more from hardocp which is from baidu..



rofl 9980x? 9700k lololol. 10nm++ looking damn strong man, 8 cores in a laptop see that 9750k at turbo 5ghz?? possibly the devil canyon version of 9700k.

 

- bored of 5.2ghz i wish i could get that on a 6 core, or even 5ghz on 6c
- i feel you, it is definitely getting worse. looking at clevo simply assumed 6c 95w TDP is close enough as 4c 91w TDP thus possibly reuse same CPU heatsink, we are once again limited on OC. as long as OEM/ODM take those TDP value seriously rather than testing them out themselves, we will always be crippled right off the bat.
- as for danisblunt.. when we post something disadvantage for BGA it seems he turns a blind eye to it.. until he acknowledge the facts I got tired of replying to him.

edit

well the leak from baidu could likely be fake take it with lots of salt at 100kg lol.

spelling error, 10nm+ and 10nm++ too far out, memory speed validation only comes when cpu is out/made, missing cascade lake which is SKL-X refresh in the chart.

 

GTX 1080 SLI turdbook rekt by 980SLI desktop.

That has to be one of the saddest things I've seen in a long time. Keep proving my points, it's hilarious :'D

https://www.3dmark.com/3dm11/9816353 GTX 980SLI > GTX 1080 SLI turdbook
https://www.3dmark.com/fs/3424081 GTX 980SLI > GTX 1080 SLI turdbook

You tried showcasing that your highly overclockedand overpriced turdbook could run with the big boys but now you showcased how people with outdated hardware were beating you :'DDDDDD

Even tho it was a good laugh, please stop trying to showcase your turdbook, all I'm interested in are the benchmarks above, all the other things are not interesting at all. Hey you might do something useful for once and actually run those benchmarks urself and maybe record them and show how good/bad your notebook runs the games.

Provided you can run the benchmarks ofc, whereas something tells me you can't.

 

This being better and better. Compare 6 core 5960x running 5.624GHz vs. oc'd 4 core 7700K. Oh'well I see much here in life. I don't know what to say You mean the desktop setup you show is better in gaming as well? Most likely.

 

I like how you showcased how the combined test result shows some kind of bottleneck in Mr.Fox turdbook (probably caused by a powerdraw issue), thanks for pointing that out, you're the man! Keep up the good work. <3

btw. still waiting for game benchmarks like the thread asked for.

 

Thanks, yes... it is amazing work... world class and industry leading notebook performance that the lesser notebooks cannot even attempt to aspire to. It's nice owning the most powerful notebook available. As far as the game benchmarks, I don't waste my time on irrelevant activities that I gain nothing from doing, so you'll be waiting forever. Don't hold your breath.

Just so you understand what you are looking at, the combined test difference reflects the more powerful HEDT CPU. The 10% / 6 FPS difference is very small considering the 7700K is not really an enthusiast CPU by desktop standards, has 50% fewer cores/threads and was clocked 400MHz less than the hexacore.

One thing you seem to not be aware of is some benchmarks and even some games run better on older hardware. 980, even 780, is able to outperform 1080 and Titan Xp in certain tasks. That's just the nature of software. Likewise, Windows 7 performance beats Windows 10 performance on many things.

 

Learn how Futuremark's 3DMark benchmark software works.

Combined test
"3DMark Fire Strike Combined test stresses both the GPU and CPU simultaneously. The GPU load combines elements from Graphics test 1 and 2 using tessellation, volumetric illumination, fluid simulation, particle simulation, FFT based bloom and depth of field. The CPU load comes from the rigid body physics of the breaking statues in the background. There are 32 simulation worlds running in separate threads each containing one statue decomposing into 113 parts. Additionally there are 16 invisible rigid bodies in each world except the one closest to camera to push the decomposed elements apart. The simulations run on one thread per available CPU core."

And how many cores have 5960X? 4, 6, 8, 10 or 50?

 

That's exactly right. For the longest time a 1080 would give better fps than a 1080ti in PUBG. In fact I'm not sure if it's even been fixed yet. And that's just 1 example without looking too hard.

@Danishblunt

I'm not sure you know what you're arguing here. There's no power draw limitations on Mr Fox's machine. All the power in the world isn't going to turn a 4 core into an 5960x physics score no matter what.

 

Has yet to be proven by benchmarks he won't provide. Thus far it seems the Clevo actually has powerdraw issues when using DX12 or games that are both GPU and CPU taxing, hence this thread to gather information. If this is fixed by prema has yet to be shown.

https://www.3dmark.com/fs/11379496
Here u can see the combined score is around 50FPS. (better CPU, higher clocked GPU)

https://www.3dmark.com/fs/10372425
Here u can see the combined score is 53.74FPS. (worse CPU, lower clocked GPU)

Powerdraw issues are very real when looking at very high end systems. 6Core+, despite the worse CPU, the combined score and even graphics score was better on the system with the lower clockspeed on the GPU and worse CPU. Synthethic benchmarks sadly don't mean that much in actual gaming, since i've seen countless systems that would go through synthethic benchmarks like 3dmark, cinebench but wouldn't last 10 minutes in witcher 3 for instance.

But then again, All i'm asking for are gaming benchmarks from Tomb raider DX12, for honor, CS:GO and DX 12 deus ex, the fact that people like mr.fox keep bragging about scores that gets beaten by GTX 980 and doesn't actually show what I've been asked for lets me believe that his system isn't even stable enough to run those benchmarks without crashing or performing horribly. Which wouldn't matter anyways since stock is actually most interesting here, since it seems that he has a premamod which might have fixed a lot. It's important for me to know wheter or not there are issues, since I want to know if those clevo notebooks are recommendable or not.

 

You don't know enough information to come to that conclusion. His memory bus clock is higher for one. Perhaps he was also using the higher bandwidth SLI cable too. Plus it's also entirely possible to run a bench with higher clocks and a lower score.

Premamod fixed any throttling issues I had when I had my machine. Not sure if Mr. Fox even has thouse games anyway. Why not do a steam gift and ask nicely and I'm sure he'll oblige

 

If you want to buy those games for me I will run the benchmarks. Not going to spend my own money on them.

 

Can't spend a couple of euros on 1or 2 games, but spends over 10k euros on 3 overpriced notebooks. Seems good. At least you can spam other peoples thread with nonsense

 

He does not play games, he enjoys playing with the laptops, benching them and tinkering with them.

Therefore he sees a benefit from the machines but not wasting money on games he never intends to play.

 

Not everyone is a gamer... There are other uses for computers... I don't know any but I hear there are some.

Sent from my XT1585 using Tapatalk