Hi
I use Google Translate
Sorry if it was a typo
Problem with motherboard asus b250 expert
I have 4 rigs with Asus motherboard
One with 4 graphics
One with 6 graphics
One with 8 graphics
And one with 11 graphics
The first series, when the rigs light up, everything is normal, both hashrate and cpu average
But if the system crashes for any reason, it will not work normally after the reboot, and the cpu average will go up and the hash will drop.
I connected the monitor, then the reboot gives this line “watchdog: BUG: soft lockup - CPU # 0 stuck”
The only solution is to turn off the power of the whole rig manually for 10 seconds and then turn it on for normal operation
I only have this problem with the Asus motherboard
No problem with Windows
I did all the solutions on the forum, such as how to set up and configure the BIOS, but the problem was not solved
Due to the fact that the motherboard is for Miner, it is used by a large number of users
I ask all users and forum administrators to think about this problem
I’m a fan of hive and I don’t want to leave it, but I have a problem
thank you
I’ve been using b250 boards for about 2 years now with hive and windows, and it was only this year when they became ‘sort of stable’ while using hiveos.
What I’ve learned is that while windows may ‘seem more stable’ with an identical setup it will not perform as well and over time issues will come up with windows. I.e. having to wait 15-30 minutes each time the GPU drivers have to reinstall because it detected something slightly different about the hardware setup is not fun on a 11 card rig or something stupid like forgetting to turn off software update and being stuck in an update loop that initiates in the middle of the night ; not being able to complete because of your custom bios settings.
Anyway here’s what I did to get it kind of stable (runs three to seven days without needing a reboot) My issues now are mainly with mining software initiated boots (phoenix miner) but that just started happening with 5.1c.
You’ll want the latest BIOS, this solved a lot of problems with stability.
Properly distribute PCI-X power among cards. (DO NOT USE SATA CONVERTERS/ADAPTORS) Good rule of thumb is you don’t want to share PCI-X power with two 1080’s but you can squidge by sharing it with one 1080 and a 1060 (but you’re pushing it.) But sharing one rail with two 1060’s is perfectly ok.
Use Molex style risers for lower-powered cards (like 1660ti/1060) if unavoidable
Try to use the same ‘type’ of riser card per each bank of connectors, and try to not mix riser types. (Do not mix version 6 with 6c/7, or if you have no other choice limit it to the same type per bank)
Use cation when mixing 1660ti’s (I’ve had a lot of problems with these cards that I was only able to fix using a specific bios/hardware configuration)
Overclock GPU’s one at a time.
Only load 50% capacity per PSU.
8GB of ram is recommended for more than 7 GPU’s, you may not use it but there may be a time when it gets used momentarily.
Disable watchdog until your rig is stable to avoid boot-loops. (Or disable entirely doesn’t seem to function anyway except to report its not working)
Calculate GPU power usage by the default powerdraw, take this into account as when your rig boots it’ll be running on stock settings until the overclock is applied… if something happens during the handshake process with the server your overclock (undervolt settings) won’t be triggered. If you do not balance your PSU power properly you will be unstable. In this reguard you cannot run cards like you can in windows, and linux will burn your sh*t.
Troubleshooting tips:
Check the power connectors at the PSU for any melted pins. This has happened to me on several occasions but still booted (after replacing cable/moving to a new port fixed the issue.) I’m still reusing the PSU but not on the ports that were ‘burned.’
Plug in gpu’s one at a time until the rig is stable to determine if its your card. If so attempt to re-arrange them until the problem disappears. Some issues arise from PCI-X address assignments and is difficult to predict/troubleshoot.
Move troublesome GPU’s to their own rig to minimize downtime while troubleshooting.
Repaste your old and ‘warmer’ running graphics cards. Fan control in hive is kinda funky. Maybe they’re working on it.
Hi dear friend🌹
I agree with all your words
I also had your experiences
The main problem is only with the asus 250 expert motherboard
After rebooting the system, it hangs and the power button must be turned on and off to solve the problem
