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

microclue

Minimal Tinkerer
Aug 20, 2019
3
1
Sorry y'all. It seems like my i7-9700kf does throttle after a few minutes, I just never ran a long enough test to notice. In the balanced test, it takes around 3 minutes to start throttling, but in the small/smallest it throttles instantaneously. Also, I hover around 120W, before throttling to ~100W. And by throttling, it goes below the turbo clock (all core 4600mhz), but never below the base clock. Sorry for the wrong info I provided...
 

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
1. Hm... Gamers Nexus made several tests grinding IHS, but never got better results than -2C. Do you have any links to tests with the difference of 10-15C?
2.How there can be a difference in cooling if there is no stractural difference (if descreet GPU used)? F version has a desactivated iGPU and normal version has activated, but not ACTIVE iGPU. This should result in exactly same results.

Comparison between different CPU, 9900K - 9900Kf

About the temp, it was at the time of the famous 6700K. I'm not sure to find any media "official" comparison test. But you will find a lot of discussions on the forum.
Exemple:

--> Now this aspect seems to be less an issue for people, but the fact is that it's still the case.
I have seen some really warped 7700K & some 9900K recently.I had some really nice one also.
It's less problematic with CPU like 9600 regarding the TDP.

It's

For Gamer Nexus, could you send us the link ?I'm curious & also It really depends of the problem at the beginning. I have seen guys starting from good ones ...

Also I'm working now in Data Center business for many years. It's the same case. You can put two blades with same chips in same rack. You can notice performance difference, sometimes +/-5°C.

& honestly I don't think it's really needed for the case.
I think the best way to use intelligently is to improve performance vs power consumption, like the "most efficient PC".
--> It may be to soon for PC market ;) (I have seen TDP of 280W for next threadripper).
On my side I still need to work on it but I have some ideas to test & try to do a kind of process.

Sorry y'all. It seems like my i7-9700kf does throttle after a few minutes, I just never ran a long enough test to notice. In the balanced test, it takes around 3 minutes to start throttling, but in the small/smallest it throttles instantaneously. Also, I hover around 120W, before throttling to ~100W. And by throttling, it goes below the turbo clock (all core 4600mhz), but never below the base clock. Sorry for the wrong info I provided...

So what's the stability frequence you have found and corresponding power ?
Maybe you can fix clock to 4.2 or 4.4GHz. It's typically what I use with 9900K.
 

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
Update about the Top Fan, other version made by Clement with buttons.



The Top fan only



In this case we are not compatible with all 200mm Fan. For exemple the Noctua is not going into due to the square shape.
We can only fit this kind of fan.



So we have two solution:



Please let me know your feedback & maybe vote !!
 

Nord1ing

Average Stuffer
Dec 5, 2018
82
38
Update about the Top Fan, other version made by Clement with buttons.



The Top fan only



In this case we are not compatible with all 200mm Fan. For exemple the Noctua is not going into due to the square shape.
We can only fit this kind of fan.



So we have two solution:



Please let me know your feedback & maybe vote !!
I had an experience with TT 200mm fan just once, with core V21 case. In less than a year it start to have a ticking noise at low rpm :|
Maybe, need to drill a hole in Noctua fan to fit?
 

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
It will be compatible in the case you choose to what we can call solution A/ PWR button at the back & smooth top skin = noctua compatible

For solution B/ you will need to grind the Fan a little bi if you wan to stay with it.



I'm not used to brand other than Noctua, but maybe a CoolerMaster MegaFlow 200 would be nice ? if somebody has tested ?
 

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
Hello, still need your help !!



 

Eduard

Trash Compacter
May 31, 2019
35
7
It will be compatible in the case you choose to what we can call solution A/ PWR button at the back & smooth top skin = noctua compatible

For solution B/ you will need to grind the Fan a little bi if you wan to stay with it.



I'm not used to brand other than Noctua, but maybe a CoolerMaster MegaFlow 200 would be nice ? if somebody has tested ?
I already have the noctua fan so will grind it =) Not a problem

PS: I like this look with thinner edges (Top hat with buttons)
 

yeahe - ML

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

Elisa attended to Innovatech show yesterday. We winally finished 3rd over 88 startups from Belgium Wallonia and yes we loose againt this "tech".



Even if I'm sure it's really usefull , I honestly doubt that's really a new tech...
Badly, we promised ourselves to not cheat for the vote (there was no technical issue to do that). Maybe we should had to ;)
A picture of the show when available.


We have made final fine tuning on the top Fan. We added a little plate at the back to mount properly U skin and back skin (come back) as the top structure goes into the back structure. Easier & cost effective solution.
The top structure will move to 2mm steel to improve rigidity.



Still 2 solutions compatible:
- PWR button at the back + Noctua 200mm + top smooth skin
- PWR button at the top + Round 200mm fan* + top with connector holes

*or Noctua modified

Solutions will come with
- Back skin
- Top skin from the option
- New improved grid.
- 3D printed back panel

New grid will be based on this material. Aesthetic is better, available in white / black and natural.


Price will depend on the quantities, as we will do pre-order. Nevertheless for small batch we will be around 80€ + shipping.

Have a good day,
 
  • Like
Reactions: Eduard

Goatee

King of Cable Management
Jun 22, 2018
739
1,513
My build is fully assembled with a 2080 (210W) and 9600k (110w) using default board (MSI) / card (EVGA) settings.

I have a noctua 140mm fan @ 1000rpm which is pretty quiet, just a gentle whoosh.

I did a synthetic Burn test last night, fully assembled with case shell and meshed top fitted. CPU hit 91c and GPU got to 75c both temps stable based on a 30 min combined fully synthetic burn using Furmark and Prime95 combined run.

Total power at the wall was 350w using a Corsair gold 450w SFF PSU. So much worse than any real life situation, no throttling on either component (Tjunction is 100c for the CPU).

Pretty impressed with the cooling ability.
 
  • Like
Reactions: Eduard

Eduard

Trash Compacter
May 31, 2019
35
7
Comparison between different CPU, 9900K - 9900Kf

It is really difficult to make any conclusion of this article. He had 200 i9-9900K, but only 2 i9-9900KF. Possible that he just gets very lucky with this one. I didn't see anywhere else confirmation that F versions are better for Overclocking. And it is not logical. Intel just decided not to scrap chips with defective graphics, so it should behave exactly the same as K version with not used graphics.

About the temp, it was at the time of the famous 6700K. I'm not sure to find any media "official" comparison test. But you will find a lot of discussions on the forum.
Exemple:

--> Now this aspect seems to be less an issue for people, but the fact is that it's still the case.
I have seen some really warped 7700K & some 9900K recently.I had some really nice one also.
It's less problematic with CPU like 9600 regarding the TDP.
For Gamer Nexus, could you send us the link ?I'm curious & also It really depends of the problem at the beginning. I have seen guys starting from good ones ...

I was talking about lapping of IHS. Delidding is not an option for me as i7-9700K has already soldered IHS.
Here are some results from Gamers Nexus:

Also I'm working now in Data Center business for many years. It's the same case. You can put two blades with same chips in same rack. You can notice performance difference, sometimes +/-5°C.

Good example on the article you posted on overclocking 200 i9-9900K. Same SKU, but the temp goes from 75C to 96C for same load...
 

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
In the comment of the post, there are a moderator of Toms that don't understand the interest of a middle power, 100% air & no noise product that do the job that other don't.... I wondering if this people are still taking time to read and understand before making a comment.

Update: News about pre order on VRM and Top Fan are coming. Sorry for the delay, we were busy working for our other jobs.
 

yeahe - ML

Cable-Tie Ninja
Original poster
Monster Labo
Dec 19, 2017
230
203
Update today: still 13 Top Fan available for first batch.

 

yeahe - ML

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

Our feedback regarding last configuration we had tried to mount. I would like to share my experiment for people that want to use this kind of configuration.

Ryzen 3700X + Gigabyte X570 I Aorus Pro WiFi

Motherboard side:
The motherboard we had, had a lot of difficulties to boot. The first boot has been obtained after the board itself did a factory reset.
After that, standard windows installation, reboot and ... nothing !
I achieved to install windows by flashing previously the bios prior to windows. Last bios was not working, so "I pick one" .
After that, standard installation of software into windows, driver installation. I made some thermal test, playing some hours with the CPU setting & unfortunately I tried to re-update the bios.
--> Bad idea as the board is impossible to boot now !!
I can't even flash the bios anymore. I sent back the board.

I had an other one (two PC to do for a company), and this one has never boot at all withtout reset of the bios.
Same fate for this one !!

CPU side:
I haven't have a lot of time to really test the CPU and try to optimize the behaviour but my feedback below:
  • CPU iddle power & temperature are high ... I don't know how to optimize it but you can have more than 20/25W and 45/50°C in iddle
  • Temperatures are really high regarding power consumption.
    The TDP seems to be optimized (80-90W for me depending of software) but temperatures are really high.
    It's a lot worst than any Intel CPU I tested so far.
    It's a lot worst than any Ryzen 2000 also ...
    --> More than 85°C on the Die with fan at full speed. It's clearly not normal !!
  • I checked over the web, a lot of people have trouble, even with watercooling.
  • The frequencies goes down directly close to 4.0Ghz when temperatures rise.

My conclusion:
Regarding X570 motherboard, I recommend to not use this Gigabyte reference.
I haven't had any problem with older gen. So for now I will still recommend to use B450 or X470 board.

Also, the older gen board has no active fan on the chipset. I modified the heatsink for one of the board for my purpose but it's not easy (you need some tooling), and also you loose the warranty. Other point, even if you find a motherboard compatible in standard, the fan can be heard at any speed. At Full speed, It's quiet annoying for a fanlessPC.

For The CPU, I think that it's not fitting nicely for the First. Temperatures are really high and so it's not the best in a Fanless PC.

For my needs & my customer, we moved to a 9700KF & Asrock Z390m-itx board.
The CPU will even perform better for most application due to higher stable frequencies.

Maybe @Nord1ing could complete my post ? You had tested deeper than me.
 

Nord1ing

Average Stuffer
Dec 5, 2018
82
38
Update:

Our feedback regarding last configuration we had tried to mount. I would like to share my experiment for people that want to use this kind of configuration.

Ryzen 3700X + Gigabyte X570 I Aorus Pro WiFi

Motherboard side:
The motherboard we had, had a lot of difficulties to boot. The first boot has been obtained after the board itself did a factory reset.
After that, standard windows installation, reboot and ... nothing !
I achieved to install windows by flashing previously the bios prior to windows. Last bios was not working, so "I pick one" .
After that, standard installation of software into windows, driver installation. I made some thermal test, playing some hours with the CPU setting & unfortunately I tried to re-update the bios.
--> Bad idea as the board is impossible to boot now !!
I can't even flash the bios anymore. I sent back the board.

I had an other one (two PC to do for a company), and this one has never boot at all withtout reset of the bios.
Same fate for this one !!

CPU side:
I haven't have a lot of time to really test the CPU and try to optimize the behaviour but my feedback below:
  • CPU iddle power & temperature are high ... I don't know how to optimize it but you can have more than 20/25W and 45/50°C in iddle
  • Temperatures are really high regarding power consumption.
    The TDP seems to be optimized (80-90W for me depending of software) but temperatures are really high.
    It's a lot worst than any Intel CPU I tested so far.
    It's a lot worst than any Ryzen 2000 also ...
    --> More than 85°C on the Die with fan at full speed. It's clearly not normal !!
  • I checked over the web, a lot of people have trouble, even with watercooling.
  • The frequencies goes down directly close to 4.0Ghz when temperatures rise.

My conclusion:
Regarding X570 motherboard, I recommend to not use this Gigabyte reference.
I haven't had any problem with older gen. So for now I will still recommend to use B450 or X470 board.

Also, the older gen board has no active fan on the chipset. I modified the heatsink for one of the board for my purpose but it's not easy (you need some tooling), and also you loose the warranty. Other point, even if you find a motherboard compatible in standard, the fan can be heard at any speed. At Full speed, It's quiet annoying for a fanlessPC.

For The CPU, I think that it's not fitting nicely for the First. Temperatures are really high and so it's not the best in a Fanless PC.

For my needs & my customer, we moved to a 9700KF & Asrock Z390m-itx board.
The CPU will even perform better for most application due to higher stable frequencies.

Maybe @Nord1ing could complete my post ? You had tested deeper than me.
I tested 1800x vs 3700x on the same hardware(both at stock speed/voltage), with fan constant speed (same bios stock settings):

StatusPower draw (at wall), WTemp
Ambient
Hwinfo temp
CPU Tctl/Tdie
Hwinfo temp
CPU Tdie
Hwinfo CPU package power
Idle 1800x
51​
~23N/A32-37
26​
Idle 3700x
50​
~2336-4332-36
20​
temps are much more variable
Prime95 1800x
208​
~23N/A
74,8​
120​
temps = T max
Prime95 3700x
173​
~23
71,6​
73,3​
102​
temps = T max, throttle below 3.6
Vega 4k 1800x
228​
~23N/A
70​
115​
Vega 4k 3700x
193​
~23
83,9​
82,3​
99​
Interestingly, fan speed had almost no influence on 3700x, in fact heat-pipes stayed much colder than with 1800x.



After that, I tested same CPUs with NH-D15 to validate the issue, and i had the same behavior: 3700x (stock) was a bit hotter compared to 1800x@3.95Ghz@1.37V, but heatsink with 3700x was colder.

I tested with both, thermal pate NT-H1 and thermal grizzly carbonaut (thermal pad) with no difference.

In my opinion, thermal transfer between cpu core and heatsink itself (cf NH-D15 results) is bad.

P.S.: still, stock 3700x is faster in video rendering, compared to 1800x@3.95Ghz :)
 
Last edited:

Papa Rob

Minimal Tinkerer
New User
Jul 18, 2019
3
0
Thanks for the awesome data @yeahe - ML and @Nord1ing! I wonder how a voltage offset 3950x would fare. Maybe having two, more efficient, ccd's would spread the heat better, considering more heatpipes are alligned directly above the chiplets and the 3950x package draws less than 150w stock. Pairing that with a low power gpu, I remain hopeful it could work in The First.

I'm curious if lapping the IHS would help in this instance.

Unlucky streak of motherboard problems.. About the chipset fan, Der8auer used a cheapo passive heatsink during his chipset power consumption tests and never got above 74C. He talks about it here (15:40). With your heatsink mod, were you able to install an m.2 drive in that slot?

Edit: a lot.
 
Last edited: