Message boards : Number crunching : Overclocking your cpu/mb for faster U@H crunching
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5427 - Posted: 15 May 2022, 12:25:29 UTC
Last modified: 15 May 2022, 12:28:19 UTC

Hi,
It seems like there are two issues involved in OCing a motherboard.

1) What to set the parameters for in the MB bios.

And

2) Getting "instrumentation" to allow you to judge how hot your CPU is and how much voltage the CPU socket is using.

After experimenting with with Asus b450-F MB and my Ryzen 3950x as well as adding an AIO cpu cooler I decided to "move up" to X570 MB(s).

That is when I discovered the "Asus-wmi-sensors" from the Git Hub doesn't work on Amd's X570 MB's.

On another thread(s) at Einstein@Home I have been discussing this. Unfortunately the E@H message website is now being afflicted with the "Pentathlon" project slowdown/time out. And the actual project target is U@H. So....

U@H seems to have recovered (at least in the message areas) from that.

So I thought I would bring the conversation over here.

Respectfully,
Tom M
A proud member of the OFA (Old Farts Assoc.)
ID: 5427 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5430 - Posted: 15 May 2022, 13:52:43 UTC - in response to Message 5427.  
Last modified: 15 May 2022, 14:04:12 UTC

<p>Keith Myers wrote:
</p>

<p>The X570 boards are not supported for asus-wmi-sensors since they do not have a WMI BIOS.</p>

<p>You are stuck with the standard nct6775 SIO driver which is of limited usefullness.</p>

<p>There is a new asus-ec-sensors driver in the upstream 5.17 kernel hwmon repository. It can be installed from its Github repo since George was able to successfully install it.</p>

<p><a href="https://github.com/zeule/asus-ec-sensors">asus-ec-sensors driver</a></p>

<p>Also you can use the zenpower3 driver to get correct voltages, temperatures&nbsp;and currents from the cpu.</p>

<p><a href="https://github.com/Ta180m/zenpower3">Zenpower3 kernel driver</a></p>

<p>And a GUI monitor that can be used instead of sensors in the command line.</p>

<p><a href="https://github.com/Ta180m/zenmonitor3">Zenmonitor3</a></p>



I am trying to bring the messages I want to respond to/post about over from Einstein@Home -> Generic CPU thread.
A proud member of the OFA (Old Farts Assoc.)
ID: 5430 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5431 - Posted: 15 May 2022, 14:05:37 UTC - in response to Message 5430.  

Here is another message from Keith:

modinfo nct6775

Will show the module is present in the system. All you have to do activate it.

sudo modprobe nct6775

And then follow that up with a lsmod to check that it is installed afterward.

lsmod

And finally a sensors command in the Teminal wil show the output of the nct6775 kernel module. It might be numbered a bit differently depending on just exactly which chip version your board uses.

A proud member of the OFA (Old Farts Assoc.)
ID: 5431 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5432 - Posted: 15 May 2022, 14:12:58 UTC - in response to Message 5431.  

Here is another message from Keith:

modinfo nct6775

Will show the module is present in the system. All you have to do activate it.

sudo modprobe nct6775

And then follow that up with a lsmod to check that it is installed afterward.

lsmod

And finally a sensors command in the Teminal wil show the output of the nct6775 kernel module. It might be numbered a bit differently depending on just exactly which chip version your board uses.


Here is what I get when I follow the above commands:
=================
tlgalenson@MoonGlow-CPU:~$ sudo modinfo nct6775
[sudo] password for tlgalenson:
filename: /lib/modules/5.13.0-41-generic/kernel/drivers/hwmon/nct6775.ko
license: GPL
description: Driver for NCT6775F and compatible chips
author: Guenter Roeck <linux@roeck-us.net>
srcversion: 6A2450AC2B674DB39C45313
depends: hwmon-vid
retpoline: Y
intree: Y
name: nct6775
vermagic: 5.13.0-41-generic SMP mod_unload modversions
sig_id: PKCS#7
signer: Build time autogenerated kernel key
sig_key: 5D:E7:71:0F:BD:65:0F:59:B6:2A:8F:FB:B8:CC:3F:49:1C:39:33:89
sig_hashalgo: sha512
signature: 50:CE:DE:78:E8:A9:3D:98:87:0E:DF:6E:B7:38:CD:15:C9:DA:DA:3F:
46:BD:69:EF:82:3E:B7:AA:7E:76:12:72:8D:D7:22:8E:C1:2D:74:F2:
38:C5:F4:B1:A1:ED:34:F3:C2:2C:BE:6E:2A:35:A6:98:E6:30:49:44:
50:FD:6F:FC:95:46:24:67:12:2A:8D:68:14:0A:0A:90:25:D3:17:95:
C2:29:AD:B2:4C:8F:AE:23:BB:76:D1:83:51:A9:F9:9E:0A:36:EC:6F:
F3:B0:75:D6:49:0A:77:C2:2A:62:30:4D:44:6C:E6:27:F3:17:E2:3D:
96:9E:62:74:E4:1E:F4:0B:7C:65:CD:08:60:F6:9E:C0:20:E7:B9:ED:
19:36:2F:FE:B7:6F:A7:73:94:00:89:EB:97:3C:FA:71:D2:54:36:AD:
0C:9C:1F:8D:02:3E:E8:D8:9E:14:BC:34:32:DD:72:68:57:3C:6C:A9:
43:BF:E2:42:83:26:ED:B6:1D:55:90:8D:AE:8F:29:3D:3B:75:BF:7A:
B5:DD:77:BA:89:15:5F:9E:80:A0:69:FD:B3:DE:1C:4A:99:E6:3E:58:
73:11:75:E8:64:72:24:3A:95:1B:49:EB:E4:08:FA:E7:6B:83:54:D6:
6E:C0:0C:DB:F3:8D:C8:5B:CA:D3:79:83:ED:58:5D:B4:3E:76:F7:6C:
CD:39:BE:8F:5E:DF:09:2E:6E:45:BD:51:8E:22:85:5F:AA:9F:4D:3D:
3A:1F:13:C0:8D:5B:D0:66:78:3B:37:23:5A:C0:D5:B7:80:4D:D6:B2:
A6:2D:E5:7E:22:15:81:E0:8F:1C:EB:40:4F:8F:FE:5B:18:8E:DC:8A:
34:38:5E:92:A7:90:7D:85:17:AD:F8:22:45:DF:01:5A:00:D2:89:20:
2A:59:24:13:B1:41:E9:D7:C0:33:CA:33:22:DD:94:7A:C1:FB:CF:99:
DE:36:4C:49:B2:54:95:E4:4A:C0:4C:85:EF:BD:91:27:AA:3A:A1:31:
E6:40:DF:EE:30:E6:4A:7A:DE:30:DD:75:0C:6F:70:3D:68:74:E5:DE:
FB:33:C2:AA:55:22:6D:2E:36:A2:71:68:41:8B:06:E6:56:12:FA:61:
FE:B3:F1:7B:6C:62:C4:6D:20:B6:EE:1C:EF:BF:DC:EA:98:FE:C5:86:
1D:88:BF:34:D8:A5:96:A8:AB:1C:44:6A:A0:BC:85:3D:D7:94:64:09:
7B:2D:E2:AE:91:62:97:B6:E9:8D:0C:C5:19:C7:52:9E:E9:B3:E4:55:
6C:12:77:8A:65:94:14:28:53:BC:DD:43:61:A1:44:5A:D1:B5:32:54:
E7:10:26:F8:7F:05:67:69:FC:9D:71:D1
parm: force_id:Override the detected device ID (ushort)
parm: fan_debounce:Enable debouncing for fan RPM signal (ushort)
tlgalenson@MoonGlow-CPU:~$ sudo modprobe nct6775
tlgalenson@MoonGlow-CPU:~$
tlgalenson@MoonGlow-CPU:~$
tlgalenson@MoonGlow-CPU:~$ lsmod
Module Size Used by
ccm 20480 6
joydev 28672 0
input_leds 16384 0
nvidia_uvm 1036288 4
intel_rapl_msr 20480 0
nvidia_drm 61440 7
intel_rapl_common 24576 1 intel_rapl_msr
nvidia_modeset 1200128 12 nvidia_drm
edac_mce_amd 32768 0
nvidia 35319808 822 nvidia_uvm,nvidia_modeset
snd_hda_codec_realtek 147456 1
snd_hda_codec_generic 81920 1 snd_hda_codec_realtek
ledtrig_audio 16384 1 snd_hda_codec_generic
snd_hda_codec_hdmi 61440 1
snd_hda_intel 53248 7
snd_intel_dspcfg 28672 1 snd_hda_intel
snd_intel_sdw_acpi 20480 1 snd_intel_dspcfg
nls_iso8859_1 16384 1
rtl8xxxu 135168 0
snd_hda_codec 147456 4 snd_hda_codec_generic,snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec_realtek
kvm 864256 0
mac80211 1028096 1 rtl8xxxu
snd_hda_core 94208 5 snd_hda_codec_generic,snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec,snd_hda_codec_realtek
snd_hwdep 16384 1 snd_hda_codec
crct10dif_pclmul 16384 1
ghash_clmulni_intel 16384 0
snd_pcm 114688 4 snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec,snd_hda_core
aesni_intel 376832 4
snd_seq_midi 20480 0
snd_seq_midi_event 16384 1 snd_seq_midi
crypto_simd 16384 1 aesni_intel
cryptd 24576 2 crypto_simd,ghash_clmulni_intel
snd_rawmidi 36864 1 snd_seq_midi
cfg80211 888832 2 mac80211,rtl8xxxu
snd_seq 73728 2 snd_seq_midi,snd_seq_midi_event
snd_seq_device 16384 3 snd_seq,snd_seq_midi,snd_rawmidi
libarc4 16384 1 mac80211
eeepc_wmi 16384 0
wmi_bmof 16384 0
rapl 20480 0
efi_pstore 16384 0
drm_kms_helper 253952 1 nvidia_drm
snd_timer 40960 2 snd_seq,snd_pcm
cec 53248 1 drm_kms_helper
rc_core 61440 1 cec
zenpower 16384 0
snd 94208 25 snd_hda_codec_generic,snd_seq,snd_seq_device,snd_hda_codec_hdmi,snd_hwdep,snd_hda_intel,snd_hda_codec,snd_hda_codec_realtek,snd_timer,snd_pcm,snd_rawmidi
fb_sys_fops 16384 1 drm_kms_helper
k10temp 16384 0
syscopyarea 16384 1 drm_kms_helper
sysfillrect 16384 1 drm_kms_helper
sysimgblt 16384 1 drm_kms_helper
soundcore 16384 1 snd
ccp 98304 0
mac_hid 16384 0
sch_fq_codel 20480 6
nct6775 69632 0
hwmon_vid 16384 1 nct6775
ipmi_devintf 20480 0
ipmi_msghandler 114688 1 ipmi_devintf
msr 16384 0
parport_pc 45056 0
ppdev 24576 0
lp 20480 0
drm 557056 11 drm_kms_helper,nvidia,nvidia_drm
parport 65536 3 parport_pc,lp,ppdev
ip_tables 32768 0
x_tables 49152 1 ip_tables
autofs4 45056 2
hid_generic 16384 0
usbhid 57344 0
hid 139264 2 usbhid,hid_generic
mfd_aaeon 16384 0
asus_wmi 36864 2 eeepc_wmi,mfd_aaeon
sparse_keymap 16384 1 asus_wmi
video 53248 1 asus_wmi
crc32_pclmul 16384 0
nvme 49152 2
r8169 77824 0
ahci 40960 0
xhci_pci 24576 0
realtek 32768 1
libahci 36864 1 ahci
xhci_pci_renesas 20480 1 xhci_pci
i2c_piix4 28672 0
nvme_core 126976 3 nvme
wmi 32768 3 asus_wmi,wmi_bmof,mfd_aaeon
tlgalenson@MoonGlow-CPU:~$ sensors
nvme-pci-0100
Adapter: PCI adapter
Composite: +58.9°C (low = -273.1°C, high = +84.8°C)
(crit = +84.8°C)
Sensor 1: +58.9°C (low = -273.1°C, high = +65261.8°C)
Sensor 2: +66.8°C (low = -273.1°C, high = +65261.8°C)

k10temp-pci-00c3
Adapter: PCI adapter
Tctl: +79.9°C
Tdie: +79.9°C
Tccd1: +80.0°C
Tccd2: +77.8°C

tlgalenson@MoonGlow-CPU:~$

==================
As far as I can tell the mode prob command is not returning "anything" like what I am used to.

Tom M
A proud member of the OFA (Old Farts Assoc.)
ID: 5432 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Keith Myers
Avatar

Send message
Joined: 10 May 20
Posts: 308
Credit: 4,733,484,700
RAC: 13,061
Message 5433 - Posted: 15 May 2022, 15:57:28 UTC

Did you reboot the host after the modprobe command? The kernel module won't be active until you do.
This is what you should see. The naming might be different depending on the SIO chip used on your board.
The values are mostly wrong as there is no scaling conf file in play.

keith@Rocinante:~$ sensors
nct6779-isa-0290
Adapter: ISA adapter
Vcore: 624.00 mV (min = +0.00 V, max = +1.74 V)
in1: 240.00 mV (min = +0.00 V, max = +0.00 V) ALARM
AVCC: 3.31 V (min = +2.98 V, max = +3.63 V)
+3.3V: 3.31 V (min = +2.98 V, max = +3.63 V)
in4: 1.85 V (min = +0.00 V, max = +0.00 V) ALARM
in5: 1.05 V (min = +0.00 V, max = +0.00 V) ALARM
in6: 848.00 mV (min = +0.00 V, max = +0.00 V) ALARM
3VSB: 3.39 V (min = +2.98 V, max = +3.63 V)
Vbat: 3.23 V (min = +2.70 V, max = +3.63 V)
in9: 0.00 V (min = +0.00 V, max = +0.00 V)
in10: 880.00 mV (min = +0.00 V, max = +0.00 V) ALARM
in11: 816.00 mV (min = +0.00 V, max = +0.00 V) ALARM
in12: 1.69 V (min = +0.00 V, max = +0.00 V) ALARM
in13: 912.00 mV (min = +0.00 V, max = +0.00 V) ALARM
in14: 760.00 mV (min = +0.00 V, max = +0.00 V) ALARM
fan1: 0 RPM (min = 0 RPM)
fan2: 0 RPM (min = 0 RPM)
fan3: 0 RPM (min = 0 RPM)
fan4: 0 RPM (min = 0 RPM)
fan5: 0 RPM (min = 0 RPM)
SYSTIN: +44.0°C (high = +0.0°C, hyst = +0.0°C) ALARM sensor = thermistor
CPUTIN: +43.5°C (high = +80.0°C, hyst = +75.0°C) sensor = thermistor
AUXTIN0: +34.5°C sensor = thermistor
AUXTIN1: +32.0°C sensor = thermistor
AUXTIN2: +36.0°C sensor = thermistor
AUXTIN3: +39.0°C sensor = thermistor
PCH_CHIP_CPU_MAX_TEMP: +0.0°C
PCH_CHIP_TEMP: +0.0°C
PCH_CPU_TEMP: +0.0°C
PCH_MCH_TEMP: +0.0°C
intrusion0: ALARM
intrusion1: ALARM
beep_enable: disabled

A proud member of the OFA (Old Farts Association)
ID: 5433 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5434 - Posted: 15 May 2022, 17:45:09 UTC - in response to Message 5430.  


<p>There is a new asus-ec-sensors driver in the upstream 5.17 kernel hwmon repository. It can be installed from its Github repo since George was able to successfully install it.</p>

<p><a href="https://github.com/zeule/asus-ec-sensors">asus-ec-sensors driver</a></p>



I got as far as decompiling the binary file.

But it makes other assumptions that I can't (yet) figure out.
A proud member of the OFA (Old Farts Assoc.)
ID: 5434 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5435 - Posted: 15 May 2022, 18:00:54 UTC - in response to Message 5433.  

Did you reboot the host after the modprobe command?


Yes and it didn't help. However someone mumbled around about
Hi, I was having a very similar problem and the solution was to add option "acpi_enforce_resources=lax" to parameter GRUB_CMDLINE_LINUX_DEFAULT= in /etc/default/grub and then run update-grub and reboot.


And when I added that and rebooted, I got:
 tlgalenson@MoonGlow-CPU:~$ sensors
k10temp-pci-00c3
Adapter: PCI adapter
Tctl:         +48.9°C  
Tdie:         +48.9°C  
Tccd1:        +37.8°C  
Tccd2:        +38.8°C  

nvme-pci-0100
Adapter: PCI adapter
Composite:    +54.9°C  (low  = -273.1°C, high = +84.8°C)
                       (crit = +84.8°C)
Sensor 1:     +54.9°C  (low  = -273.1°C, high = +65261.8°C)
Sensor 2:     +64.8°C  (low  = -273.1°C, high = +65261.8°C)

nct6798-isa-0290
Adapter: ISA adapter
in0:                        1.30 V  (min =  +0.00 V, max =  +1.74 V)
in1:                      1000.00 mV (min =  +0.00 V, max =  +0.00 V)  ALARM
in2:                        3.38 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in3:                        3.26 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in4:                        1.02 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in5:                      696.00 mV (min =  +0.00 V, max =  +0.00 V)  ALARM
in6:                      1000.00 mV (min =  +0.00 V, max =  +0.00 V)  ALARM
in7:                        3.38 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in8:                        3.28 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in9:                      904.00 mV (min =  +0.00 V, max =  +0.00 V)  ALARM
in10:                     224.00 mV (min =  +0.00 V, max =  +0.00 V)  ALARM
in11:                     648.00 mV (min =  +0.00 V, max =  +0.00 V)  ALARM
in12:                       1.02 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in13:                     992.00 mV (min =  +0.00 V, max =  +0.00 V)  ALARM
in14:                       1.01 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
fan1:                        0 RPM  (min =    0 RPM)
fan2:                     1148 RPM  (min =    0 RPM)
fan3:                        0 RPM  (min =    0 RPM)
fan4:                        0 RPM  (min =    0 RPM)
fan5:                     2934 RPM  (min =    0 RPM)
fan6:                     2368 RPM  (min =    0 RPM)
fan7:                        0 RPM  (min =    0 RPM)
SYSTIN:                    +43.0°C  (high = +80.0°C, hyst = +75.0°C)  sensor = thermistor
CPUTIN:                    +40.0°C  (high = +80.0°C, hyst = +75.0°C)  sensor = thermistor
AUXTIN0:                   +26.0°C    sensor = thermistor
AUXTIN1:                   +63.0°C    sensor = thermistor
AUXTIN2:                   +10.0°C    sensor = thermistor
AUXTIN3:                   +26.0°C    sensor = thermistor
PECI Agent 0 Calibration:  +40.5°C  
SMBUSMASTER 1:             +67.0°C  
PCH_CHIP_CPU_MAX_TEMP:      +0.0°C  
PCH_CHIP_TEMP:              +0.0°C  
intrusion0:               ALARM
intrusion1:               ALARM
beep_enable:              disabled

tlgalenson@MoonGlow-CPU:~$ ^C
tlgalenson@MoonGlow-CPU:~$ 



Which looks a lot like what you just posted.

Do any of the outputs map to CPU temperature? And/or Cpu volts?

Tom M
A proud member of the OFA (Old Farts Assoc.)
ID: 5435 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Keith Myers
Avatar

Send message
Joined: 10 May 20
Posts: 308
Credit: 4,733,484,700
RAC: 13,061
Message 5436 - Posted: 15 May 2022, 19:04:51 UTC

Oh, my mistake. I should have remembered that. Glad you got it working.

No, not really. The only outputs that can be believed are the fans speeds. Why I stated the utility of the driver is very low.

Without a sensors.conf scaling file in the sensors.d directory for the chipset, most of the values are meaningless.

Nuvoton makes the chips.

The problem is that the data sheet for the sensor SIO chip on the motherboards is proprietary. You have to pay for the license and sign an NDA to have access to the information so that you can scale the output values to correct values.

The Windows monitoring apps and the motherboard manufacturer monitoring apps do this so the values are correct.

That is never going to happen for Linux as the company will never release the datasheets to an OS that is open source and does not allow restriction of source code or NDA's.

The only way to get realistic values is to develop a sensors.conf scaling file through trial and error and reverse engineering by private individuals.

Why I suggested to get the zenpower3 kernel module built and installed. That provides the missing sensor information that the nct6775 driver does not believe in realistic and actual values.

If you want to take a crack at writing your own sensors.conf scaling file here is the man pages for sensors.

https://linux.die.net/man/5/sensors.conf
https://linux.die.net/man/3/libsensors

Or you can Google the hardware forums and search for someone that has figured out the scaling for your motherboard sensor and published their sensor.conf file so that you can copy it and use on your board.

A proud member of the OFA (Old Farts Association)
ID: 5436 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5439 - Posted: 15 May 2022, 21:27:07 UTC - in response to Message 5436.  
Last modified: 15 May 2022, 21:47:29 UTC


Why I suggested to get the zenpower3 kernel module built and installed. That provides the missing sensor information that the nct6775 driver does not believe in realistic and actual values.


========================================
tlgalenson@MoonGlow-CPU:~/Downloads/zenpower3-master$ sudo apt install dkms git build-essential linux-headers-$(uname -r)
[sudo] password for tlgalenson:
Reading package lists... Done
Building dependency tree
Reading state information... Done
git is already the newest version (1:2.25.1-1ubuntu3.4).
linux-headers-5.13.0-41-generic is already the newest version (5.13.0-41.46~20.04.1).
build-essential is already the newest version (12.8ubuntu1.1).
dkms is already the newest version (2.8.1-5ubuntu2).
The following package was automatically installed and is no longer required:
libfwupdplugin1
Use 'sudo apt autoremove' to remove it.
0 upgraded, 0 newly installed, 0 to remove and 1 not upgraded.
tlgalenson@MoonGlow-CPU:~/Downloads/zenpower3-master$ cd ~
tlgalenson@MoonGlow-CPU:~$ git clone https://github.com/Ta180m/zenpower3.git
fatal: destination path 'zenpower3' already exists and is not an empty directory.
tlgalenson@MoonGlow-CPU:~$ cd zenpower3
tlgalenson@MoonGlow-CPU:~/zenpower3$ sudo make dkms-install -i
dkms --version >> /dev/null
mkdir -p /usr/src/zenpower-0.2.0
cp /home/tlgalenson/zenpower3/dkms.conf /usr/src/zenpower-0.2.0
cp /home/tlgalenson/zenpower3/Makefile /usr/src/zenpower-0.2.0
cp /home/tlgalenson/zenpower3/zenpower.c /usr/src/zenpower-0.2.0
sed -e "s/@CFLGS@//" \
-e "s/@VERSION@/0.2.0/" \
-i /usr/src/zenpower-0.2.0/dkms.conf
dkms add zenpower/0.2.0
Error! DKMS tree already contains: zenpower-0.2.0
You cannot add the same module/version combo more than once.
make: [Makefile:41: dkms-install] Error 3 (ignored)
dkms build zenpower/0.2.0
Module zenpower/0.2.0 already built for kernel 5.13.0-41-generic/4
dkms install zenpower/0.2.0
Module zenpower/0.2.0 already installed on kernel 5.13.0-41-generic/x86_64
tlgalenson@MoonGlow-CPU:~/zenpower3$

====================

Now what?
===
Sorry still reading dirs. Be back.
====
tlgalenson@MoonGlow-CPU:~$ sudo modprobe zenpower
tlgalenson@MoonGlow-CPU:~$ sensors
zenpower-pci-00c3
Adapter: PCI adapter
SVI2_Core: 1.28 V
SVI2_SoC: 1.11 V
Tdie: +76.8°C (high = +95.0°C)
Tctl: +76.8°C
Tccd1: +77.0°C
Tccd2: +74.0°C
SVI2_P_Core: 97.13 W
SVI2_P_SoC: 13.36 W
SVI2_C_Core: 75.76 A
SVI2_C_SoC: 11.77 A

nvme-pci-0100
Adapter: PCI adapter
Composite: +61.9°C (low = -273.1°C, high = +84.8°C)
(crit = +84.8°C)
Sensor 1: +61.9°C (low = -273.1°C, high = +65261.8°C)
Sensor 2: +68.8°C (low = -273.1°C, high = +65261.8°C)

nct6798-isa-0290
Adapter: ISA adapter
in0: 1.31 V (min = +0.00 V, max = +1.74 V)
in1: 992.00 mV (min = +0.00 V, max = +0.00 V) ALARM
in2: 3.38 V (min = +0.00 V, max = +0.00 V) ALARM
in3: 3.25 V (min = +0.00 V, max = +0.00 V) ALARM
in4: 1.01 V (min = +0.00 V, max = +0.00 V) ALARM
in5: 712.00 mV (min = +0.00 V, max = +0.00 V) ALARM
in6: 1000.00 mV (min = +0.00 V, max = +0.00 V) ALARM
in7: 3.38 V (min = +0.00 V, max = +0.00 V) ALARM
in8: 3.28 V (min = +0.00 V, max = +0.00 V) ALARM
in9: 904.00 mV (min = +0.00 V, max = +0.00 V) ALARM
in10: 240.00 mV (min = +0.00 V, max = +0.00 V) ALARM
in11: 656.00 mV (min = +0.00 V, max = +0.00 V) ALARM
in12: 1.02 V (min = +0.00 V, max = +0.00 V) ALARM
in13: 984.00 mV (min = +0.00 V, max = +0.00 V) ALARM
in14: 1000.00 mV (min = +0.00 V, max = +0.00 V) ALARM
fan1: 0 RPM (min = 0 RPM)
fan2: 1824 RPM (min = 0 RPM)
fan3: 0 RPM (min = 0 RPM)
fan4: 0 RPM (min = 0 RPM)
fan5: 2641 RPM (min = 0 RPM)
fan6: 2356 RPM (min = 0 RPM)
fan7: 0 RPM (min = 0 RPM)
SYSTIN: +42.0°C (high = +80.0°C, hyst = +75.0°C) sensor = thermistor
CPUTIN: +53.0°C (high = +80.0°C, hyst = +75.0°C) sensor = thermistor
AUXTIN0: +26.0°C sensor = thermistor
AUXTIN1: +60.0°C sensor = thermistor
AUXTIN2: +10.0°C sensor = thermistor
AUXTIN3: +26.0°C sensor = thermistor
PECI Agent 0 Calibration: +65.5°C
SMBUSMASTER 1: +63.5°C
PCH_CHIP_CPU_MAX_TEMP: +0.0°C
PCH_CHIP_TEMP: +0.0°C
intrusion0: ALARM
intrusion1: ALARM
beep_enable: disabled

tlgalenson@MoonGlow-CPU:~$
A proud member of the OFA (Old Farts Assoc.)
ID: 5439 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5440 - Posted: 15 May 2022, 22:48:51 UTC - in response to Message 5439.  

It took some tinkering to get the Monitor program compiled and working.

--------
tlgalenson@MoonGlow-CPU:~/Downloads/zenmonitor3-master$ sudo apt-get install libgtk-3-dev
[sudo] password for tlgalenson:
Reading package lists... Done
Building dependency tree
Reading state information... Done
---etc-------------
tlgalenson@MoonGlow-CPU:~/Downloads/zenmonitor3-master$ make
cc -Isrc/include `pkg-config --cflags gtk+-3.0` src/ss/*.c src/sysfs.c src/zenmonitor-lib.c src/gui.c src/zenmonitor.c -o zenmonitor `pkg-config --libs gtk+-3.0` -lm -no-pie -Wall
tlgalenson@MoonGlow-CPU:~/Downloads/zenmonitor3-master$ sudo ./zenmonitor
A proud member of the OFA (Old Farts Assoc.)
ID: 5440 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5441 - Posted: 15 May 2022, 22:53:36 UTC - in response to Message 5440.  
Last modified: 15 May 2022, 22:55:23 UTC

Here is the current results. If I am seeing things correctly the voltage is under 1.30 volts. This is with the cpu voltage set on "auto" and all the LLC bios items available set to "extreme".



A proud member of the OFA (Old Farts Assoc.)
ID: 5441 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Keith Myers
Avatar

Send message
Joined: 10 May 20
Posts: 308
Credit: 4,733,484,700
RAC: 13,061
Message 5443 - Posted: 15 May 2022, 23:25:58 UTC

Ok, looks like you had to do a lot of gymnastics to get the dkms build to make and install.

If you fail on the first dkms install, it is always best to check the dkms status and if the build directory got installed in the failure, it is best to run the uninstall or remove the directory for a fresh start.

But you persevered and got it done I see. So now you have good sensor information to allow you to see the cpu voltages and temps to allow for lower voltage tweaking if you desire.

You can continue to use the terminal to see your cpu voltages or you can try and make and build the zenmonitor3 GUI application. You will have to install the libgtk-3-dev package first probably to satisfy the build dependencies.

I normally don't bother with git on github repos. I normally just use the package zip selection from the green CODE button and download the zip file of the package. Then unpack it in my Downloads folder and build there in the extracted directory. Up to you whatever way you choose.

Or just follow the standard instruction for Ubuntu from the readme.

sudo modprobe msr
sudo bash -c 'echo "msr" > /etc/modules-load.d/msr.conf'
sudo apt install build-essential libgtk-3-dev git
cd ~
git clone https://github.com/Ta180m/zenmonitor3
cd zenmonitor3
make
sudo make install
sudo make install-polkit

I just do this manually.

sudo apt install libgtk-3-dev

You also will need to install the msr driver to read the cpu registers.

sudo apt install msr-tools

The activate the msr driver

sudo modprobe msr

make

sudo make install

sudo make install-polkit

The polkit install makes a copy of the application with root user priviledges. So you end up with two copies in your Show Applications page.

The root version is handy for showing the effective clocks of each core but the most important one is the total package power which does not get shown with the normal user version of the application. You need the msr driver to read those registers and that requires sudo priviledges.

A proud member of the OFA (Old Farts Association)
ID: 5443 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Keith Myers
Avatar

Send message
Joined: 10 May 20
Posts: 308
Credit: 4,733,484,700
RAC: 13,061
Message 5444 - Posted: 15 May 2022, 23:31:33 UTC

Never mind. I see you built the monitor while I was typing up the how-to.

That extreme LLC setting is the tightest one. Only allows for very little droop. I'd be curious to know what the cpu voltage spikes to when you remove the BOINC load.
But the monitor probably can't catch that at its polling rate. You would need a scope to see the voltage spike.

But you now have the tools to try carving away at the cpu voltage off Auto to see what you need for absolute lowest voltage to maintain stability at your desired clock speed.

Good job on the driver and application builds.

A proud member of the OFA (Old Farts Association)
ID: 5444 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5445 - Posted: 16 May 2022, 0:48:08 UTC - in response to Message 5444.  

I have been pushing the CPU multiplier rather than trying to drop the CPU voltage.

So far the voltage has stayed the same, but I have reached around 80C with a 4.2Mhz cpu.

So I expect if I want to "go faster" I need to drop the CPU voltage to lower the CPU temperature?

Tom M
A proud member of the OFA (Old Farts Assoc.)
ID: 5445 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Keith Myers
Avatar

Send message
Joined: 10 May 20
Posts: 308
Credit: 4,733,484,700
RAC: 13,061
Message 5446 - Posted: 16 May 2022, 2:02:28 UTC - in response to Message 5445.  

No, if you want to go faster, then go faster. Just push the clock some more.

Remember you can go to 95°C.

This is on the 5950X, correct?

I do 4425Mhz all-core on my 5950X hosts, all three of them. All set to 0.20V offset and running 1.25 - 1.26V Vcpu under BOINC load. LLC3 load line calibration setting.

Running around 76 - 82° C. under custom cooling.

A proud member of the OFA (Old Farts Association)
ID: 5446 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Grant (SSSF)

Send message
Joined: 23 Apr 22
Posts: 156
Credit: 69,772,000
RAC: 582
Message 5447 - Posted: 16 May 2022, 5:16:04 UTC - in response to Message 5446.  

Remember you can go to 95°C.
...
Running around 76 - 82° C. under custom cooling.
For 24/7/365 and to last as long as the rest of the system i consider 70°c to be the limit. Maybe short periods up around 75°c (ie really hot days here & there).
But the higher you run over 70°c for extended periods, the sooner you're likely to have issues. The hotter a semi conductor is, the greater the risk of Electromigration
Grant
Darwin NT
ID: 5447 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Keith Myers
Avatar

Send message
Joined: 10 May 20
Posts: 308
Credit: 4,733,484,700
RAC: 13,061
Message 5449 - Posted: 16 May 2022, 5:49:03 UTC - in response to Message 5447.  

I've never burned out an AMD chip in over 25 years of use and extreme overclocking.

I have however killed a Intel chip from running it too fast and too hot. I still have the replacement 6850K on the shelf and never installed after Intel replaced it under the overclocking warranty I was smart enough to purchase.

I have seen no sign of electromigration in any AMD chip since the NexGen Nx586 I started with on my AMD journey.

I never felt the need to purchase the equivalent insurance when running AMD processors. They don't burn out.

A proud member of the OFA (Old Farts Association)
ID: 5449 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5450 - Posted: 16 May 2022, 9:15:56 UTC - in response to Message 5445.  

I have been pushing the CPU multiplier rather than trying to drop the CPU voltage.

So far the voltage has stayed the same, but I have reached around 80C with a 4.2Mhz cpu.

So I expect if I want to "go faster" I need to drop the CPU voltage to lower the CPU temperature?



Oops, it locked up. Just had to clear the CMOS to get it to boot again.

I have dialed it back to 4.150 since it has run 4.125 for days.

This is a 3950x cpu.

The Zen3 Monitor is really spiffy.

Tom M
A proud member of the OFA (Old Farts Assoc.)
ID: 5450 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Keith Myers
Avatar

Send message
Joined: 10 May 20
Posts: 308
Credit: 4,733,484,700
RAC: 13,061
Message 5467 - Posted: 16 May 2022, 18:52:56 UTC - in response to Message 5450.  

With normal behavior of cpus, to go faster requires more voltage. That is going to lead to more heat. No way around that.

Can you dump your BIOS settings via the BIOS profile menus for me.

Insert a USB stick to the motherboard rear ports and via the Tools menu >> Save Profile >> Alt-F2 dump the settings.txt file to the USB stick.

The copy and paste the settings.txt file via PM for me to examine.

A proud member of the OFA (Old Farts Association)
ID: 5467 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom M

Send message
Joined: 18 Jul 17
Posts: 134
Credit: 1,379,173,617
RAC: 7,900
Message 5468 - Posted: 16 May 2022, 20:12:22 UTC - in response to Message 5467.  

With normal behavior of cpus, to go faster requires more voltage. That is going to lead to more heat. No way around that.

Can you dump your BIOS settings via the BIOS profile menus for me.

Insert a USB stick to the motherboard rear ports and via the Tools menu >> Save Profile >> Alt-F2 dump the settings.txt file to the USB stick.

The copy and paste the settings.txt file via PM for me to examine.


Sigh, it looks like it rebooted at 4.150 GHz so I am back down to "auto".

So at the moment dumping you the bios settings is probably useless.

After a while I will try everything back up at 4.125 and see if it will stay up.

Since I now have available at least one much fancier Asus x570 MB that has all those granular settings, fiddling with this MB maybe wasted time on your part?

Tom M
A proud member of the OFA (Old Farts Assoc.)
ID: 5468 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
1 · 2 · Next

Message boards : Number crunching : Overclocking your cpu/mb for faster U@H crunching




Copyright © 2024 Copernicus Astronomical Centre of the Polish Academy of Sciences
Project server and website managed by Krzysztof 'krzyszp' Piszczek