Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Sensor values for MSI Z370 Gaming Plus
#1
Hello,

First, thanks for the great application!

Second, there might be some issues with this board and the readings from the Nuvoton NCT6795D chip. One by one:

- there are 2 CPU readings; I don't think this was always the case.
- the 2nd CPU reading seems to be identical with CPU (PECI)
- System seems pretty much borked and barely moves from 32-33C, regardless of load
- Auxiliary, AUXTIN1 seem borked as well as they are lower than ambient and not budging under load
- one time under load, and additional AUXTIN3 appeared, borked as well, since it oscillated between 0-1C; can see it on the attached picture
- there's a PCH temperature that pretty much stays at 48-49C; might be the Southbridge? Since that would kinda explain the lack of response under load. Could also be borked, I guess. There's an old post from the AIDA64 dev saying it's the "chipset component", but the Northbridge would be on-die these days, so it should go up with the other temps, sadly it's just guesswork for me.

This basically adds up to the Nuvoton chip being entirely useless temperatures-wise. Now, I know you cannot rely 100% on these things and you need some sort of external probe etc. to properly measure temps, but having an estimation of at least the motherboard/VRM would be nice. Maybe I'm expecting too much from 140 Euro. Imagine how much the decent boards would cost here Sad

Also one question:

Possible to see the VRM temps on this board? I've seen much better boards with seemingly decent cooling getting real toasty on the VRM, so I suspect I should be worried a bit about my budget board with shitty VRM.

You have the PC details on the report. Thanks for taking a look!


Attached Files Thumbnail(s)
   

.htm   DESKTOP-MTS81IG.HTM (Size: 147.47 KB / Downloads: 3)
.dbg   HWiNFO64x.DBG (Size: 1.11 MB / Downloads: 1)
Reply
#2
"System" temperature is reporting the mainboard temperature (a sensor somewhere on mainboard, outside of hotspots), which should be correlating with ambient temperatures.
Auxiliary and AUXTIN values are invalid readouts that should be ignored.
PCH temperature is the Peripheral Controller Hub (sort of south bridge), which is a discrete chip on manboard. Its temperature is higher than ambient (depends on cooler) and doesn't depend on CPU load.
Your board is not capable of measuring VRM temperatures.
Reply
#3
(03-29-2018, 07:30 AM)Martin Wrote: "System" temperature is reporting the mainboard temperature (a sensor somewhere on mainboard, outside of hotspots), which should be correlating with ambient temperatures.
Auxiliary and AUXTIN values are invalid readouts that should be ignored.
PCH temperature is the Peripheral Controller Hub (sort of south bridge), which is a discrete chip on manboard. Its temperature is higher than ambient (depends on cooler) and doesn't depend on CPU load.
Your board is not capable of measuring VRM temperatures.

Thanks a lot. 

I suspected there won't be VRM readings. Still disappointed, that should be basic, I don't mind no RGB, no USB 3.1 type c, no WiFi, but extensive sensor data, that I want. Guess it will have to wait until z470 or w/e. I could actually get the PCH temp to go up by doing some CUDA/OPEN CL work, unsure if it was just the GPU heating up the motherboard or the chip itself, but yeah, it looks functional.

There's still a duplicate "CPU" reading, you can see it in the picture. The 2nd "CPU" reading is actually running the same values as "CPU (PECI)". There's also the AUXTIN3 1C temp reading which sometimes appears, sometimes it doesn't, don't know why. I know can remove them, I don't mind them, just thought it might be useful to you.

Thanks again for the clarification.
Reply
#4
(03-29-2018, 07:30 AM)Martin Wrote: "System" temperature is reporting the mainboard temperature (a sensor somewhere on mainboard, outside of hotspots), which should be correlating with ambient temperatures.
Auxiliary and AUXTIN values are invalid readouts that should be ignored.
PCH temperature is the Peripheral Controller Hub (sort of south bridge), which is a discrete chip on manboard. Its temperature is higher than ambient (depends on cooler) and doesn't depend on CPU load.
Your board is not capable of measuring VRM temperatures.

Hello again.

My apologies for reviving the old topic, but I do it pretty much to thank you.

It looks Beta added a "MOS" temperature for my motherboard. It is higher than the CPU (PECI) with about 15C and about the same as highest core temp (under prime95 avx load, stock 8700k). It goes down when idle.

Should I assume the app is now able to read VRM temps? Or is it a bug? I'd be incredibly happy if this was the VRM temp, as it's definitely manageable and not even close to worrisome.

PS: At idle the temps remain relatively high (45-50C) and don't decrease almost instantly like core temps, which would pretty much be what I expect from a passively cooled VRM. I don't see any temperature that matches MOS, so it's new and not a double.

Would be great if you could confirm this is VRM.

Thanks!
Reply
#5
Yes, HWiNFO added reporting of the MOS temperature for Z370 series, but apparently lots of Z370 models don't support this.
I'm still waiting for a better clarification from MSI and will then update accordingly.
Is that value comparable to highest Core Temperature at idle too? What is the difference?
Reply
#6
(10-25-2018, 07:16 AM)Martin Wrote: Yes, HWiNFO added reporting of the MOS temperature for Z370 series, but apparently lots of Z370 models don't support this.
I'm still waiting for a better clarification from MSI and will then update accordingly.
Is that value comparable to highest Core Temperature at idle too? What is the difference?

I will give you a few numbers. Specs are 8700K stock not delidded, 1070ti from Gigabyte, Z370 Gaming Plus, cheap 120mm Zalman CNPS10X Optima cooler, Fractal Design Meshify C case with 120mm fans front intake, 2 top exhaust, 1 back exhaust, Superflower Leadex II 650W PSU. In case they matter, so you get a proper picture.

1. 5 mins after boot. Ambient is around 22C. Idling while typing. Highest core temp 28C. MOS temp 30C. System 28C. PCH 31C. CPU(peci) 15C (I assume this temp is never accurate at idle).
2. starting prime 95 smallFFTs, allowing 5 mins. The BIOS has a long term power limit of 95W. 
Highest core temp 66C. MOS temp 63C. System 30C. PCH 33C. CPU(peci) 53C 
3. 4.9GHz with power limits removed, 1.32V, prime95 non-AVX (145-150W CPU package power)
Highest core temp 98C. MOS temp 90--->96CC. System 28C. PCH 31C. CPU(peci) 84C

MOS temps seem to be steadily climbing. This is the most I can extract from this cooling setup without delidding/buying better cooling.

Seems to me it's kind of reasonable to how the VRM temps should act?
Reply
#7
Thanks for the data. How quick does the MOS temp ramp up or go down? If it's much slower that the CPU temp, my guess would be that it's indeed the MOS temperature, because that's exactly the way how they behave.
Anyway, let's wait for the response from MSI too...
Reply
#8
(10-25-2018, 03:13 PM)Martin Wrote: Thanks for the data. How quick does the MOS temp ramp up or go down? If it's much slower that the CPU temp, my guess would be that it's indeed the MOS temperature, because that's exactly the way how they behave.
Anyway, let's wait for the response from MSI too...

I edited the post above. The MOS temps go down VERY slow. I shut down prime95 and now after 1-2 mins MOS is 70C, PECi is 25C, core max 37C.

If this is truly the VRM, then it's good news for me, it should allow me to get 5GHz when then warranty expires and I buy a new videocard that requires such speeds, delid and buy the CPU a new cooler. I bought this in desperation with Bitcoin when it was super high in Dec 2017 and I fit everything I could in the budget, better VRM was sadly not possible, and I was always reluctant because of the flimsy 4+1 phase or whatever it is. But considering the relatively low 150W at 4.9GHz, who knows.
Reply
#9
To me this looks pretty like the VRM temperature Smile
You might also try to blow a small fan at the MOS and see how it changes Wink
Reply
#10
(10-25-2018, 04:05 PM)Martin Wrote: To me this looks pretty like the VRM temperature Smile
You might also try to blow a small fan at the MOS and see how it changes Wink

Unsure I have another fan in the house. I would have to unscrew one of the case fans which I might be a bit too lazy to do Tongue If I find another fan I'll open the case and blow it on the VRM area, and return with numbers.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)