Production The First - Monsterlabo High Performance SFF case & Fanless CPU / GPU

dudeperm87

Caliper Novice
Mar 10, 2019
24
13
There is a german reviewer named Igors lab who did a test on the case and came to the conclusion that IT CAN NOT cool a 2060 because the vrms got to a temperature of 121 °C.

Test is here (https://www.igorslab.media/monsterl...albestueckung-im-exklusiven-labortest-teil-1/)

He even limited the power target and clocked the VRM down with no success. He then also tried to apply VRM cooling blocks which also didn't help, just delayed the temps a little.

So my question:
1. Will the case cool my 2070 super properly and will I have to lower the power target?
2. does anyone have any experience with a 2070 super in the case?
3. why can't igors lab cool the 2060 vrm properly. One of his commentators has gpu errors because of this and had to lower the power target of his 5700xt.

thanks
 
Last edited:

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
Hello,

Yes we have seen this article. It's a very good one with a lot of DATA !
We really like this guys that has a strong technical spirit and a lot of knowledge.
I really wish that one time we see same technical analysis on standard board. Most of the time we would be really surprised (badly).

My comment:

- VRM cooling that we propose in option was not used during the test. Unfortunately the GPU was not compatible.
Results are slightly better with taped heatsink and a lot better with the big aluminium version we have on the shop.
-->This would have changed the feedback.

- 120°C is really a high temp close to the specifified limit, 125°C. This component has a nice margin that can reach to 150°C/175°C
This is not recommended but this works. In Data Center business, boards reach this kind of value during years ....

Now regarding the 2070 super, please note that this board would have to be downclock. This is mandatory !!
AAA games with this GPU can create heat load up to 215W - 230W. This is way over the maximum heat load specified !! Warning !!
I have not the feedback if downcloking is easily feasible with this board. We did it on some boards with succes as VEGA 64 or 2080 / 2080 ti.
Regarding power/price/TDP ratio I would suggest to go for a classic 2070.

BR,
 

Eduard

Trash Compacter
May 31, 2019
35
7
There is a german reviewer named Igors lab who did a test on the case and came to the conclusion that IT CAN NOT cool a 2060 because the vrms got to a temperature of 121 °C.

Test is here (https://www.igorslab.media/monsterl...albestueckung-im-exklusiven-labortest-teil-1/)

He even limited the power target and clocked the VRM down with no success. He then also tried to apply VRM cooling blocks which also didn't help, just delayed the temps a little.

So my question:
1. Will the case cool my 2070 super properly and will I have to lower the power target?
2. does anyone have any experience with a 2070 super in the case?
3. why can't igors lab cool the 2060 vrm properly. One of his commentators has gpu errors because of this and had to lower the power target of his 5700xt.

thanks

Hello, I don't speek german... The 2060 VRMs reached 120C in totally passive config ot with 14cm fan?
I don't know about 2070 super (TDP 215-225 Watt), but I'm using 1080Ti (TDP 250Watt) with 20cm fan on top.
I had 1860 Mhz (not very stable) in heaven in my preveous rig (reference 1080Ti with Arctic accelero III on it). For "the First" I undervolt it to 1810Mhz on 0.875V. I lost very little in performance, but the clock is stable and much less heat. It reaches 70C in heaven with closed case. VRM was like 80C, but with opened case (don't have tools to measure temp in the closed one).
All this with TOP 20cm fan at 350-400 RPM - compleately inaudible.
Was playing metro exodus several hours, had no problems. HWMonitor reported 220Watts consumption in avarage with 72C maximum.

I advise to undervolt a GPU (if you dont overclock it) in any case. You will keep the same perf or gain a little bit, but the temp will be better controlled.
 

dudeperm87

Caliper Novice
Mar 10, 2019
24
13
THe 2060 VRMs reached 121 °C with the 12cm fan. The major problem was also that the PCB also reached 120 °C. All this is in part because the 14 cm Fan doesn't do anything for the VRM and PCB cooling because of the orientation of the GPU in the case.

The thermal engineer at Gigabyte says that 100 °C is the maximum safe temperature for a PCB. Any temperatures above that lead to gassing out and lanes breaking while cooling down in the long term. Igors lab makes the point that the VRM chips each need 2 watts and that there is 30 watts in total which the card draws besides the GPU itself.

the problem is that these 30 watts produce heat which needs to go somewhere and needs to be cooled. that why the PCB reaches these high temperatures of >120 °C.
 

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
THe 2060 VRMs reached 121 °C with the 12cm fan. The major problem was also that the PCB also reached 120 °C. All this is in part because the 14 cm Fan doesn't do anything for the VRM and PCB cooling because of the orientation of the GPU in the case.

The thermal engineer at Gigabyte says that 100 °C is the maximum safe temperature for a PCB. Any temperatures above that lead to gassing out and lanes breaking while cooling down in the long term. Igors lab makes the point that the VRM chips each need 2 watts and that there is 30 watts in total which the card draws besides the GPU itself.

the problem is that these 30 watts produce heat which needs to go somewhere and needs to be cooled. that why the PCB reaches these high temperatures of >120 °C.

The heat will be released by natural convection due to the GPU position. Indeed, this position helps to the cooling as you are not blocking the heat to go up.
Even if the Fan is not blowing directly on the GPU, the position creates a flow that sucks all the air from bottom to top.
Imagine a classic board mounted upside down & horizontally with no thermal contact to the VRM. That still happends...

Maybe adding a front Fan could be a solution for such high power... Let see !

Off course I trully think that the VRM cooling is necessary with this board so as mounting some Memories heatsink (VGA-02). For really high power TDP, a full covering would be really nice but this cost a lot and is really engineering expensive. Maybe for a bigger & powerfull version ;)
 

bbalex

Chassis Packer
Nov 30, 2018
18
1
I've seen the same test on tomshardware.fr at this link, with the RTX 2060 (and they did try to use small heatsinks on the memory chips in a second attempt): GPU 60C, VRM 121C, VRAM chips with no heatsinks on them 121C, VRAM chips with heatsinks on them 105C

The system was one with ASRock Z390, i5-9600K, GeForce RTX 2060. As for the CPU temps: CPU 67C, VRM 84C after 30 minutes, when running Prime95 with AVX instructions enabled for tests.

It's exactly the same review, as the author is the same, but in French and posted on another website.

I've been saying this since the beginning: there really is a need for a better way to cool the graphic card VRM and memory chips, as the good temp of the GPU chip seen in monitoring tools give the feeling that everything's fine, while the rest of the graphic card is undergoing an accelerated degradation due to extremely high temps. Otherwise cards with such high power consumption figures should not be recommended as safe to be used.
 

cptarrr

Efficiency Noob
Feb 16, 2019
5
1
  • also can you send us the reference of the small copper heatsinks for memories ?

I'm not him, but those heatsinks look like those sold by Enzotech: BMR-C1 or BCC9 on the memory chips, MOS-C1 on the VRM caps or chips. Manufacturer page here.

@yeahe - ML -> @bbalex <- was absolutely right. And as i can see you put them on your shop - you owe me one of your huge beers

After a few months i am still happy with my case. I only have sometimes problems with the temperature (in summer) after a couple hours of gaming in one session (PC will shut down) but thats my fault, cause my hardware may be a little bit too OP for this case. Anyway i will print the 'lower skin with dust filters' for a better airflow and then the problems should be gone
 
  • Like
Reactions: yeahe - ML

Eduard

Trash Compacter
May 31, 2019
35
7
Hello Nord1ing. Thank you for publishing your results.

If I got it right, you have here a 20cm fan on the top and a 14cm fan inside? I know it is not an easy operation, but did you test without 14cm fan inside? To know if it is really beneficial for temps...
In same test, my i7-9700k generates 170W of heat, that is too much for the case and it throttles. Not much more (Watt) than generate your R5 1600x, but the difference in the temps may due to the size difference of IHS (it is much bigger on Ryzen) or something else... Not sure that we can compare CPU from different manufacturers.
So to control the temp I set up the package power limit in bios to 120W => The score on Cynebench 15 got from 1468 to 1430 (not much of a difference), but the temps don't exceed low 70th now. So I'm pretty happy with this taking in account the totally silent operation (500RPM for 20cm fan).

Concerning VRM overheat on the graphics card. I have 1080Ti with small radiators from arctic accelero III and undervolted a little bit.
It reaches in heaven 190W and don't exceed 65C (the hottest part of the PCB is 85-90C after 30 min) with opened top part of the case. Wasn't able to measure the VRM temps with closed case, but the GPU temps gets lower (62C) with closed case. Due to more efficient air flow I think (still 20c fan at 500RPM).

So no need to worry about accelerated degradation... Top 20cm fan creates good airflow through graphics card (without noise).
 
  • Like
Reactions: yeahe - ML

Nord1ing

Average Stuffer
Dec 5, 2018
82
38
Hello Nord1ing. Thank you for publishing your results.

If I got it right, you have here a 20cm fan on the top and a 14cm fan inside? I know it is not an easy operation, but did you test without 14cm fan inside? To know if it is really beneficial for temps...
In same test, my i7-9700k generates 170W of heat, that is too much for the case and it throttles. Not much more (Watt) than generate your R5 1600x, but the difference in the temps may due to the size difference of IHS (it is much bigger on Ryzen) or something else... Not sure that we can compare CPU from different manufacturers.
So to control the temp I set up the package power limit in bios to 120W => The score on Cynebench 15 got from 1468 to 1430 (not much of a difference), but the temps don't exceed low 70th now. So I'm pretty happy with this taking in account the totally silent operation (500RPM for 20cm fan).

So no need to worry about accelerated degradation... Top 20cm fan creates good airflow through graphics card (without noise).

Yes, I have both fans installed. Mainly because the 20 cm fans do not have a lot of static pressure, and I installed dust filters not only on the side panels (as seen above), but also on the bottom of the case. I will try to test without a 14 cm fan in the future when I switch 1600x to 3800x.

At the moment, I have reduced the CPU voltage (Vcore offset) -150 mV in the BIOS (the default voltage is terribly high on the Asrock AB350 ITX), and my thermal package does not exceed 125 W .
 

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
@yeahe - ML -> @bbalex <- was absolutely right. And as i can see you put them on your shop - you owe me one of your huge beers

After a few months i am still happy with my case. I only have sometimes problems with the temperature (in summer) after a couple hours of gaming in one session (PC will shut down) but thats my fault, cause my hardware may be a little bit too OP for this case. Anyway i will print the 'lower skin with dust filters' for a better airflow and then the problems should be gone

Good to know that you are happy with the case !!
 

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
Hello Nord1ing. Thank you for publishing your results.

If I got it right, you have here a 20cm fan on the top and a 14cm fan inside? I know it is not an easy operation, but did you test without 14cm fan inside? To know if it is really beneficial for temps...
In same test, my i7-9700k generates 170W of heat, that is too much for the case and it throttles. Not much more (Watt) than generate your R5 1600x, but the difference in the temps may due to the size difference of IHS (it is much bigger on Ryzen) or something else... Not sure that we can compare CPU from different manufacturers.
So to control the temp I set up the package power limit in bios to 120W => The score on Cynebench 15 got from 1468 to 1430 (not much of a difference), but the temps don't exceed low 70th now. So I'm pretty happy with this taking in account the totally silent operation (500RPM for 20cm fan).

Concerning VRM overheat on the graphics card. I have 1080Ti with small radiators from arctic accelero III and undervolted a little bit.
It reaches in heaven 190W and don't exceed 65C (the hottest part of the PCB is 85-90C after 30 min) with opened top part of the case. Wasn't able to measure the VRM temps with closed case, but the GPU temps gets lower (62C) with closed case. Due to more efficient air flow I think (still 20c fan at 500RPM).

So no need to worry about accelerated degradation... Top 20cm fan creates good airflow through graphics card (without noise).

Many thanks with you feedback about the VRM temps. It's good to have customer feedback on it. !
 
  • Like
Reactions: Nord1ing

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
Hello,

Some news from Monsterlabo:
- A newcomer in the team, Clément, will help to developp the actual product. Offical news soon by newsletter
- A new design of the VRM heatsink for GPU will be launched next week. Easier to use and compatible with most of the boards.
Soon available on the shop. It will replace the previous version.
- We are now working actively on the top fan solution, sorry for the delay but we had a lot of work (not only monsterlabo) during summer.
- The PC will be presented during IBC show (Amsterdam) by one of our French partner. Pictures to come.
- Tomorrow we will have a big delivery to our Belgium partner, Celem, which are focused on Belgium institution (B2B).

The shop will be soon updated as our PC are now in stock starting tomorrow !!
We will deliver customer that ordered during August.

Also we still have "2nd hand" PCs that was used for videos, testing and media.
There are PC kit (2nd hand & some pre prod) and complete configuration with all options. Same waranties but special price, now 35% off.

Have a good week end,
 

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
The First at IBC show

When I find why I can't load pictures ....
 
Last edited:

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203


Finally ...
 
  • Like
Reactions: Eduard and Nord1ing

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
Some news from the dev.

New VRM heatsink ready for manufacuring





Two width available, 6mm and 15mm so as the 23mm original one (23mm was too big for most of the board).
Both heatsinks are now separated. Easier to mount.
Height is adjusted by using spacer of different dimensions.
This heatsink has nearly same performance than original one and definitely better than taped version.

What do you think ?
 
  • Like
Reactions: Eduard and Nord1ing

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
New Top fan solution





Exploded view



Thx Clement (new team member) for this work. Now the assembly is simplified:
  • New U shape skin which will be same aesthetic than the middle U
  • New top structure plate that allow to assemble directly different kind of fan.
    --> This will replace the strandard one already assembled.
  • New top skin without power button, brushed aluminium.
    --> The power button will be moved to the back panel. 3D back panel delivered at same time.
  • New grid with high ratio opening (already in stock). This grid will be available in black and white.
    --> This new grid will also be available for standard version
We can also propose all the aesthetic in black or white with according price.
A big U skin that replace the top one could also be done.

Now, it's time to have your feedback ?
 
  • Like
Reactions: Eduard and Nord1ing

Nord1ing

Average Stuffer
Dec 5, 2018
82
38
New Top fan solution





Exploded view



Thx Clement (new team member) for this work. Now the assembly is simplified:
  • New U shape skin which will be same aesthetic than the middle U
  • New top structure plate that allow to assemble directly different kind of fan.
    --> This will replace the strandard one already assembled.
  • New top skin without power button, brushed aluminium.
    --> The power button will be moved to the back panel. 3D back panel delivered at same time.
  • New grid with high ratio opening (already in stock). This grid will be available in black and white.
    --> This new grid will also be available for standard version
We can also propose all the aesthetic in black or white with according price.
A big U skin that replace the top one could also be done.

Now, it's time to have your feedback ?
Looks nice!
You do not need necessarily attach the fan to the top plate. At least with noctua fan, thanks to rubber pads, if the upper plate is tithely pushed over the fan, he will never move (and there is no vibrations over screws) :)

Looking forward!
 
  • Like
Reactions: Eduard