mbc0 Posted March 6, 2019 Share Posted March 6, 2019 Thanks so much for the reply! As you can see they are all quite different :-S which is correct? Quote Link to comment
bastl Posted March 6, 2019 Share Posted March 6, 2019 @mbc0 I would bet that the CPU temp showing with the it87 driver is the Tdie temperature now. "CPU Temp" + "27°C Offset" = "Tctl" Have an eye on the sensors with "watch sensors" inside a terminal and put some load on the CPU. Depending on the load the temps should jump up. You can also use the docker "Netdata" to monitor your server and check the temps with a graphs if you like. For me it sits around 40-45°C with browsing the web inside a VM with an Noctua NH U14S TR4-SP3 overclooked to 4GHz. It sometimes jumps to 50 and back down. For heavy load, rendering or gaming I see temperatures around 55. Quote Link to comment
mbc0 Posted March 6, 2019 Share Posted March 6, 2019 Hi, Thanks for the tip on netdata! here is what it is showing when encoding a video using the handbrake docker although it unfortunately barely stresses the cpu? but hopefully k10temp - CPU Temp is real? Quote Link to comment
bastl Posted March 6, 2019 Share Posted March 6, 2019 I think the temps where the CPU starts to throttle is at 68°C for Tdie. This is what you see here. The CPU is getting hot with that small Noctua NH-U9 you are using. Always keep in mind what kind of a beast of a CPU that thing is and you're using one of the smallest cooling solutions for that. If you have no option for upgrading the cooler let's say for hight reasons you have to life with it. If you have the space I would advice you to upgrade it to an U14. Quote Link to comment
mbc0 Posted March 6, 2019 Share Posted March 6, 2019 Hi, If it is true and I am getting below 70 under stress I am happy with that, I use a 24 Disk U4 Rackmount so 150 is the highest I can get in there I just don't seem to "Know For Sure" what my temps are which I find un-nerving! Quote Link to comment
bastl Posted March 6, 2019 Share Posted March 6, 2019 Btw, are you running your CPU at stock speeds? @mbc0 Quote Link to comment
mbc0 Posted March 6, 2019 Share Posted March 6, 2019 5 minutes ago, bastl said: Btw, are you running your CPU at stock speeds? @mbc0 I originally tried 4.00Ghz but unraid locked as soon as it booted so I dropped down to 3.70Ghz all seems well so far! I didn't mess with any voltages I just changed the multiplier so maybe if I spent a bit more time I could easily get the 4.00 Ghz but am looking into a small AIO Cooler first but not sure if it will be much different as the airflow from the 3 turbo fans going through the rackmount is incredible. Quote Link to comment
bastl Posted March 6, 2019 Share Posted March 6, 2019 Quickly checked the dimensions of the U12 and the U14. Both won't fit into your case. U12 hight is 158mm and U14 is 165mm without cooler. I guess the usual workloads like some gaming in VMs or a reencode with handbrake from time to time on stock speeds the U9 is OK'ish. I wouldn't push it above the stock speeds with that small cooler. There isn't really much headroom for cooling. One example review talking about the cooling capabilities, most will tell you the same, don't OC on the U9. If you plan for going for an AIO DON'T USE ONE OF THE STANDARD ASETEK ONES. Sure you have an adapter to mount it on an TR4 but they won't cover the whole IHS. You have to buy one with an IHS that covers the complete die like the ones from Enermax. I know the temperature offset AMD introduced with the AM4/TR4 chips is kinda confusing. Tdie is the actual CPU temperature and as I understand it the offset is there to have some headroom for the auto overclock boost clock speeds to work as intended. Quote Link to comment
mbc0 Posted March 6, 2019 Share Posted March 6, 2019 Hi, Many thanks for your time mate I really appreciate it! I have just increased the fan speed on the centre case fan which blows directly at the CPU cooler and it seems to have made quite a difference, not sure I am going to get any better with AIO? I am idling at 30ish and under stress I am now in the low 50's Quote Link to comment
bastl Posted March 6, 2019 Share Posted March 6, 2019 Looks good to me, if you can life with the higher noise from the fan I wouldn't change anything. If not an AIO would surely help to reduce the noise at the same temps. 1 Quote Link to comment
mbc0 Posted March 6, 2019 Share Posted March 6, 2019 The noise is really quite low I am still only running these fans at 30% and the airflow is amazing, I will look into liquid cooling (I have it all in my 8700K Gaming Rig with NVME and is damn near silent!) if after I put my rackmount into a case in the office to see how loud it is. You have been amazingly helpful, thank you so much! Quote Link to comment
skois Posted March 17, 2020 Share Posted March 17, 2020 (edited) Hello, sorry for waking up this thread again. But i can't figure out which sensor is which. I have the Gigabyte X399 Ex Designare and TR 2920x. and running 6.8.3 The cpu temps seams reasonable based on the workload. But motherboard temps seems off. MB temps are always near the cpu temp with only a couple of degrees difference. Eg. CPU 68C , MB 67C. After some time of heavy work load. CPU and MB temps are exactly the same. even on the decimals. At first i though that it was because i had many pci devices connected. But right now. Its only 1 USB Controller, 1 GPU and the disks. (4 HDD, 2SSD) But GPU and USB is off, only used when i start the VM. Sensors output under load I'm using Dark Rock Pro 4 TR4. I dont know if i should see that CPU temps. It gets at 68C even if using only 6 cores full load, Same with 12 cores full load Clocks remain at 3.9-4Ghz root@Prometheus:~# sensors it8792-isa-0a60 Adapter: ISA adapter in0: 1.33 V (min = +0.00 V, max = +2.78 V) in1: 1.49 V (min = +0.00 V, max = +2.78 V) in2: 1.04 V (min = +0.00 V, max = +2.78 V) in3: 1.98 V (min = +0.00 V, max = +2.78 V) in4: 1.78 V (min = +0.00 V, max = +2.78 V) in5: 1.49 V (min = +0.00 V, max = +2.78 V) in6: 2.78 V (min = +0.00 V, max = +2.78 V) ALARM 3VSB: 1.66 V (min = +0.00 V, max = +2.78 V) Vbat: 1.57 V Array Fan: 1318 RPM (min = 0 RPM) fan2: 1313 RPM (min = 0 RPM) fan3: 927 RPM (min = 0 RPM) temp1: +34.0°C (low = +127.0°C, high = +127.0°C) sensor = thermistor temp2: -55.0°C (low = +127.0°C, high = +127.0°C) sensor = Intel PECI temp3: +38.0°C (low = +127.0°C, high = +127.0°C) sensor = thermistor intrusion0: ALARM k10temp-pci-00c3 Adapter: PCI adapter MB Temp: +65.1°C (high = +70.0°C) Tctl: +92.1°C k10temp-pci-00cb Adapter: PCI adapter CPU Temp: +62.4°C (high = +70.0°C) Tctl: +89.4°C root@Prometheus:~# cat /proc/cpuinfo | grep "MHz". cpu MHz : 3274.140 cpu MHz : 2769.280 cpu MHz : 2933.898 cpu MHz : 2849.985 cpu MHz : 3059.644 cpu MHz : 2763.466 cpu MHz : 4141.714 cpu MHz : 3684.047 cpu MHz : 4141.711 cpu MHz : 3190.169 cpu MHz : 2993.078 cpu MHz : 3009.799 cpu MHz : 3108.950 cpu MHz : 2770.083 cpu MHz : 2937.507 cpu MHz : 2840.982 cpu MHz : 3067.074 cpu MHz : 2783.664 cpu MHz : 4141.702 cpu MHz : 3775.466 cpu MHz : 4141.487 cpu MHz : 3906.728 cpu MHz : 3011.004 cpu MHz : 2829.929 If anyone can help me please. i would appreciate. Edited March 17, 2020 by skois Quote Link to comment
bastl Posted March 18, 2020 Share Posted March 18, 2020 @skois I have 2 sensors for the 1950x on a AsRock x399 Fatality and an extra MB sensor. root@UNRAID:~# sensors nct6779-isa-0290 Adapter: ISA adapter Vcore: 640.00 mV (min = +0.00 V, max = +1.74 V) in1: 1.08 V (min = +0.00 V, max = +0.00 V) ALARM AVCC: 3.36 V (min = +0.00 V, max = +0.00 V) ALARM +3.3V: 3.36 V (min = +0.00 V, max = +0.00 V) ALARM in4: 1.84 V (min = +0.00 V, max = +0.00 V) ALARM in5: 1.14 V (min = +0.00 V, max = +0.00 V) ALARM in6: 1.36 V (min = +0.00 V, max = +0.00 V) ALARM 3VSB: 3.46 V (min = +0.00 V, max = +0.00 V) ALARM Vbat: 3.28 V (min = +0.00 V, max = +0.00 V) ALARM in9: 0.00 V (min = +0.00 V, max = +0.00 V) in10: 768.00 mV (min = +0.00 V, max = +0.00 V) ALARM in11: 824.00 mV (min = +0.00 V, max = +0.00 V) ALARM in12: 1.68 V (min = +0.00 V, max = +0.00 V) ALARM in13: 928.00 mV (min = +0.00 V, max = +0.00 V) ALARM in14: 752.00 mV (min = +0.00 V, max = +0.00 V) ALARM Array Fan: 1179 RPM (min = 0 RPM) fan3: 1380 RPM (min = 0 RPM) MB Temp: +39.0°C (high = +0.0°C, hyst = +0.0°C) sensor = thermistor CPUTIN: +41.5°C (high = +80.0°C, hyst = +75.0°C) sensor = thermistor AUXTIN0: +7.0°C sensor = thermistor AUXTIN1: +38.0°C sensor = thermistor AUXTIN2: +35.0°C sensor = thermistor AUXTIN3: +39.0°C sensor = thermistor SMBUSMASTER 0: +90.5°C intrusion0: ALARM intrusion1: ALARM beep_enable: disabled k10temp-pci-00c3 Adapter: PCI adapter Tdie: +61.9°C (high = +70.0°C) Tctl: +88.9°C k10temp-pci-00cb Adapter: PCI adapter CPU Temp: +61.9°C (high = +70.0°C) Tctl: +88.9°C I guess the 2 sensors you have selected are both from the CPU. The numbers are to close to each other to say one of them are from the MB. Mine is always sitting at 39°C. Not sure if it ever showed another number, but by stressing the CPU it won't change for me. Same temps as the CPU wouldn't make any sense. 12 hours ago, skois said: It gets at 68C even if using only 6 cores full load, Doesn't matter how much cores maxed out producing heat, the sensor will show it. Even with 1 or 2 cores at max the sensor will show high temps. Quote Link to comment
skois Posted March 18, 2020 Share Posted March 18, 2020 I think that the MB Temp listed on the plugin is the Tdie temp of the second chiplet. Thats why its always within a couple degrees off or even the same. So the main problem here is that i cant see anywhere the ChipSet temp. Do you know what driver does it uses? Quote Link to comment
bastl Posted March 18, 2020 Share Posted March 18, 2020 Tdie and Tctl are both from the CPU. First Gen Ryzens indroduced a offset varied by chips. TR4 1st gen had a 17°C offset. Ryzen itself had a different value. Second gen TR4 changed it to 27°C. Both of your sensors have a 27°C offset. What you see is 2 sensors, one for each die. Try to change the driver to "nct6775". Maybe that one is showing the right sensors Quote Link to comment
skois Posted March 18, 2020 Share Posted March 18, 2020 46 minutes ago, bastl said: Tdie and Tctl are both from the CPU. First Gen Ryzens indroduced a offset varied by chips. TR4 1st gen had a 17°C offset. Ryzen itself had a different value. Second gen TR4 changed it to 27°C. Both of your sensors have a 27°C offset. What you see is 2 sensors, one for each die. Try to change the driver to "nct6775". Maybe that one is showing the right sensors I tried typing the "nct6775" , hit save, and then load drivers. but nothing is happening. Maybe i dont have this on my board? Quote Link to comment
bastl Posted March 18, 2020 Share Posted March 18, 2020 @skois I guess you have to live with it how the it8792 driver reports it to you. As I said before my guesses are that both sensors (k10temp-pci-00c3, k10temp-pci-00cb) are reporting the temps for each die and the "MB Temp" labeling is wrong. To proof it check if the offset between the reported temps and the Tctl is 27°C and you will know it's a sensor in the CPU. Temp1 or Temp3 might be a sensor on the board somewhere. Quote Link to comment
skois Posted March 18, 2020 Share Posted March 18, 2020 1 minute ago, bastl said: @skois I guess you have to live with it how the it8792 driver reports it to you. As I said before my guesses are that both sensors (k10temp-pci-00c3, k10temp-pci-00cb) are reporting the temps for each die and the "MB Temp" labeling is wrong. To proof it check if the offset between the reported temps and the Tctl is 27°C and you will know it's a sensor in the CPU. Temp1 or Temp3 might be a sensor on the board somewhere. Temp 1,2,3 are never changing. my best guess is a placeholder value. The board have 2 sensor headers to external thermistors. Temp 2 is actualy minus 55 if i check it on bios. Hopefully a BIOS release or a kernel release will fix that. I'll try to get a disk to boot in windows to see how things are there. On ubuntu and manjaro could not also get mb temps. Quote Link to comment
bastl Posted March 18, 2020 Share Posted March 18, 2020 4 minutes ago, skois said: Temp 2 Might be the USB_Pump header for an AIO. Depending on where the board manufacturer placed the Temp1+3 sensors you might not see big changes at all. If it's near the chipset or the socket somewhere, sure, but somewhere else the temps won't really change. Quote Link to comment
skois Posted March 18, 2020 Share Posted March 18, 2020 3 minutes ago, bastl said: Might be the USB_Pump header for an AIO. Depending on where the board manufacturer placed the Temp1+3 sensors you might not see big changes at all. If it's near the chipset or the socket somewhere, sure, but somewhere else the temps won't really change. No, i mean it always T1-35C and T3-38C, on idle, on load, on bios. always. With ambient at 20C and with ambient at 30C. Thats why i though it's not a real value. Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.