Yeah, once you start using M.2 drives there’s no going back.
-
Save 15% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!
WYSIWYG - BC1 Mini
- Thread starter tinyitx
- Start date
Installed Windows 10 Pro 1803.
Now I need to do much tweaking and optimization (aka bringing back the look of Win 7 as much as possible).
And I need to do performance and thermal testings to ensure all components are working properly.
This pic shows all the RGB effects:- mobo, ram and display card.
I think Asus Aura can display the CPU temp. So, I might keep it if I get it to work.
I do not think GSkill RGB has any functional value and so it is going to be disabled.
MSI Mystic Light does not seem to be able to show the GPU temp and so this will have to go as well.
But, I will keep it like this until Christmas is over as it can pretend to be a little Christmas tree.
Now I need to do much tweaking and optimization (aka bringing back the look of Win 7 as much as possible).
And I need to do performance and thermal testings to ensure all components are working properly.
This pic shows all the RGB effects:- mobo, ram and display card.
I think Asus Aura can display the CPU temp. So, I might keep it if I get it to work.
I do not think GSkill RGB has any functional value and so it is going to be disabled.
MSI Mystic Light does not seem to be able to show the GPU temp and so this will have to go as well.
But, I will keep it like this until Christmas is over as it can pretend to be a little Christmas tree.
Get my system set up for some thermal testing.
stock i9 9900K without overclocking
Prime 95 v26.6 for 30 minutes (Small FFTs test)
Asus MCE: disabled
Ambient temp: 24C
NH-C14S fan speed @ 850rpm
CPU max temp: 59-64C
It is amazingly silent, especially given that this is an open benchtable (ie without a case to muffle any noise) and I am about 0.5m from the BC1.
Given the max rpm of the NF-A14 fan is 1500rpm, I might further adjust the fan curve to spin faster to further drive down the temp.
I might also try to enable MCE to see how much the temp will rise.
Also, will run some OCCT to double check the thermals.
Here are some final pics:-
stock i9 9900K without overclocking
Prime 95 v26.6 for 30 minutes (Small FFTs test)
Asus MCE: disabled
Ambient temp: 24C
NH-C14S fan speed @ 850rpm
CPU max temp: 59-64C
It is amazingly silent, especially given that this is an open benchtable (ie without a case to muffle any noise) and I am about 0.5m from the BC1.
Given the max rpm of the NF-A14 fan is 1500rpm, I might further adjust the fan curve to spin faster to further drive down the temp.
I might also try to enable MCE to see how much the temp will rise.
Also, will run some OCCT to double check the thermals.
Here are some final pics:-
Last edited:
Nice! I've always been of the opion most of the fan noise was from being (too?) close to vents which causes turbulence noise, so your observation makes sense to me.
Hey man! awesome build!
I have the same board plus a i5 9600K and I'm interested in buying this same cooler. I was afraid that it would interfere with the IO panel or the GPU, but now that I see your pics its going to fit perfectly in my Core V1.
I still have one little concern: As I'm looking for a decent air cooler that fits in my case and I'm planning to overclock the CPU... But I got a little concerned with the 64C without OC.. Did you overclocked the CPU and re-tested?
I'm currently at 4.9Ghz but the temps go way up with my old TT Water 2.0 performer.. So I've dialed a little bit down until I get the new one..
I have the same board plus a i5 9600K and I'm interested in buying this same cooler. I was afraid that it would interfere with the IO panel or the GPU, but now that I see your pics its going to fit perfectly in my Core V1.
I still have one little concern: As I'm looking for a decent air cooler that fits in my case and I'm planning to overclock the CPU... But I got a little concerned with the 64C without OC.. Did you overclocked the CPU and re-tested?
I'm currently at 4.9Ghz but the temps go way up with my old TT Water 2.0 performer.. So I've dialed a little bit down until I get the new one..
I did not have time to play with my new build in the past month as I have some life matters to attend to. So, I have not done much testing. But I will gradually.
I thought 64C max for a stock i9 9900K is quite OK, is it not?
And that was obtained with the CPU fan at about 800rpm.
I have done a little more testing today with Intel XTU for 15 minutes:- (HWiNFO64 monitoring temperature)
1) MCE disabled
CPU (package) temp = 65C
CPU (DTS) temp = 60-64C
fan speed about 60%
2) MCE enabled
CPU (package) temp = 73C
CPU (DTS) temp = 69-74C
fan speed about 70%
Frankly, I do not fully understand MCE much. It seems its function differs from one board to another and from one manufacturer to another. For my Asus Z390I (BIOS 0401), all cores are at 5GHz at 100% usage (well, it is more like 4998.8 MHz to 5001.2MHz) during this test.
2 questions for anyone to comment one:-
a) with MCE enabled, all the cores do not downclock and are always at 5GHz even when idling or at low usage. Normal?
b) with MCE enabled, Prime 95 v26.6, somehow, 'crashes' by a BSOD and Windows 10 reports IRQL_NOT_LESS_OR_EQUAL error after a minute or two into testing. Temp seems OK (certainly under 80C) just prior to 'crashing'. Anyone knows why?
I thought 64C max for a stock i9 9900K is quite OK, is it not?
And that was obtained with the CPU fan at about 800rpm.
I have done a little more testing today with Intel XTU for 15 minutes:- (HWiNFO64 monitoring temperature)
1) MCE disabled
CPU (package) temp = 65C
CPU (DTS) temp = 60-64C
fan speed about 60%
2) MCE enabled
CPU (package) temp = 73C
CPU (DTS) temp = 69-74C
fan speed about 70%
Frankly, I do not fully understand MCE much. It seems its function differs from one board to another and from one manufacturer to another. For my Asus Z390I (BIOS 0401), all cores are at 5GHz at 100% usage (well, it is more like 4998.8 MHz to 5001.2MHz) during this test.
2 questions for anyone to comment one:-
a) with MCE enabled, all the cores do not downclock and are always at 5GHz even when idling or at low usage. Normal?
b) with MCE enabled, Prime 95 v26.6, somehow, 'crashes' by a BSOD and Windows 10 reports IRQL_NOT_LESS_OR_EQUAL error after a minute or two into testing. Temp seems OK (certainly under 80C) just prior to 'crashing'. Anyone knows why?
I can't comment on your temperatures as I don't have a 9900K, there are people better placed than me to do so.
MCE (and XMP) can change all sorts of settings beyond the obvious changes to voltages and multipliiers. If you have the time to spare and you want to overclock, making manual adjustments in the BIOS should give you better results. I used the guide in the link below.
https://www.overclock.net/forum/6-i...ies-official-support-thread.html#post26396441
Depends on your BIOS settings. You can enable Intel speedstep to adjust clocks according to load. This has no effect on MCE, so your cores will still turbo to 5GHz under load.
Increase vCore. For me vDroop increases when running P95, so i needed to increase vCore to compensate.
MCE does not take your CPU into account, it just applies a predetermined voltage - some CPUs will work with it, some won't.
Last edited:
You're right that MCE implementations vary between OEMs. From what I've seen of 9900Ks, your temps are perfectly fine. Those chips run hot (to the degree that pro overclocker Der8auer delidded one and ground down the protective layer on top of the die to improve temps). No wonder, given that they're pulling 150+W when turboing.
From my understanding of MCE the CPU should still downclock when idle (MCE is only supposed to affect multi-core turbo), but due to the general weirdness and variability of MCE it's difficult to know if this is stock Asus behaviour or a bug. Might be that enabling MCE disables SpeedStep by default on your board? A support email might clear this up, but checking if SpeedStep is enabled or not should also help. As for the second crash, it sounds like MCE isn't entirely stable - it is technically an overclock, after all (as in: running above on-paper all-core boost speeds), and it's entirely possible that the voltages it applies are insufficient to maintain certain workloads. The error message is a bit odd, but might perhaps be down to the crash being related to the PCIe bus (perhaps MCE on your board increases the base clock?). Still, as brt02 says above, a core voltage increase is the first step - but that will of course also lead to increased temperatures.
One benefit of an open test bench is that no case fan is needed, thus reducing cost, noise and fan cable management. But this could be a weakness as well.
I found out the M.2 SSD (Samsung 970 EVO 1GB) on the underside of my Asus Z390I Gaming is a little hot.
With 25C ambient, it idles at around mid 40's C and loads at high 60's C.
Granted that I have already added a heatsink (the EK one), there is absolutely zero airflow. The only fan around it is the Corsair SF600 fan but it rarely spins.
So, I add a 40mm fan to blow air directly and perpendicular to the EK heatsink. Fan has a max rpm of 7000 and noise becomes an issue at around 4000rpm. So, I have it spinning at a quiet 3000rpm. Now, it has mid 30's C idle and mid 50's load temperature. The ave temp drop is 10C.
Now, I have to observe the top M.2 SSD more to see if it needs a similar treatment. It benefits some from the NH-C14S's downward draft but not as much as I like though. This might be because the NH-C14S is installed in a direction (the only orientation without removing the motherboard's heatsinks) that does not blow much air in the direction of the M.2 drive.
I found out the M.2 SSD (Samsung 970 EVO 1GB) on the underside of my Asus Z390I Gaming is a little hot.
With 25C ambient, it idles at around mid 40's C and loads at high 60's C.
Granted that I have already added a heatsink (the EK one), there is absolutely zero airflow. The only fan around it is the Corsair SF600 fan but it rarely spins.
So, I add a 40mm fan to blow air directly and perpendicular to the EK heatsink. Fan has a max rpm of 7000 and noise becomes an issue at around 4000rpm. So, I have it spinning at a quiet 3000rpm. Now, it has mid 30's C idle and mid 50's load temperature. The ave temp drop is 10C.
Now, I have to observe the top M.2 SSD more to see if it needs a similar treatment. It benefits some from the NH-C14S's downward draft but not as much as I like though. This might be because the NH-C14S is installed in a direction (the only orientation without removing the motherboard's heatsinks) that does not blow much air in the direction of the M.2 drive.
