You are a lifesaver my friend. I cant believe it was that simple. I got 10bit block size now. Cheers
very nice
If you find some summer settings for the c mode please share.
must have cooler mems when at work, otherwise it will be 50 degrees i the room
outside temp 30 now, room 30
fan 80…core 1300…VDD 900…Mem 900
Give 80 hash 157w 70 degrees on mem
fan 57…core 1300…VDD 850…Mem 850
Give 80 hash 135w 69 degrees on mem
They’re machines, not humans that get damage with time. I’ve read that If temperatures and voltages were too high, it could potentialy damage the silicon structure. But doesn’t seem to be this case.
Since Nov 2019 to 2021 at 93Mhs
Since May 2021 til now at 115Mhs
So far so good
dmesg | grep ‘drm.*block size’
[ 17.675072] [drm] vm size is 262144 GB, 4 levels, block size is 9-bit, fragment size is 9-bit
[ 18.291012] [drm] vm size is 262144 GB, 4 levels, block size is 9-bit, fragment size is 9-bit
thats my grub, config
GRUB_CMDLINE_LINUX_DEFAULT=“text consoleblank=0 intel_pstate=disable net.ifnames=0 ipv6.disable=1 pci=noaer iommu=soft usbcore.autosuspend=-1 radeon.si_support=0 radeon.cik_support=0 amdgpu.vm_fragment_size=9 amdgpu.si_support=1 amdgpu.cik_support=1 amdgpu.ppfeaturemask=0xffff7fff amdgpu.runpm=0 amdgpu.gpu_recovery=0 noibrs noibpb nopti nospectre_v2 nospectre_v1 l1tf=off nospec_store_bypass_disable no_stf_barrier mds=off mitigations=off amdgpu.vm_block_size=10 amdgpu.vm_size=1024”
not sure why does not say 10-bit followed the instructions.
Has anyone had issues with their cards after running in c mode?
yes i only get 8-9mh so i dont use cmode. using 20.40 drivers
Wow, something must be set incorrectly.
Заметил очень сданную особенность. Когда проверял карты по 1 шт они выдавали по 102 хэша. как собрал их в кучу некоторые 99,56 стали выдавать некоторые 101 некоторые 100,1… память hynix на всех…
настройки на всех одинаковые
1675
950/0/0
1125
amdmemtweak --RC 38 --RP 13 --FAW 12 --RAS 25 --REF 7500 --RTP 6 --RRDS 3 --RCDRD 13
подскажите с чем это связано?
While I don’t have the answer for you @TESLA78 , I see some GPU performance variation between single GPU, in rigs, and even dependencies among slots in rigs. These are all riserless examples.
Can only speculate some of it is related to the chipsets, load balancing of work, etc. Wish I knew
Tweaking a Hynix Memory VII down a notch?
I have one GPU in a multi VII rig which will “choke” in TeamRedMiner after a few hours running with the baseline amdmemtweak below:
“/hive/sbin/amdmemtweak --i 0 --ref 7500 --rtp 6 --rrds 3 --faw 12 --ras 19 --rc 30 --rcdrd 11 --rp 11”
Any suggestions on which settings would be best to try and back down a notch?
Note: running this as baseline across all my air cooled cards and 10bit is successfully enabled:
“In HiveOS make sure your overclocks are 1600 Core @900mv and max 1000 mem clock.”
double-check the test file its prob an extra " some here or missing one.
Did you manage to solve it?
198 did not work for me i had to got to 195.
c mode should automatically on when 10bit is enabled.
Hi. Where to enable 10bit? I always get error, when i try Radeon VII C-MODE GUIDE!** tnx
i did once, its got to do with the hiveos version not all versions will allow c mode for these cards, i am back to B mode.
Ok, tnx. And which version of HiveOs is that?
Cant update-grub2 keeps saying syntax error line 54
just got it working, after to do the text file update you need to run the update-grub2
command and then reboot and it should just come up in cmode, the version i am using is 5.4.0-hiveos · H 0.6-198 · A 20.40
can i install a radeon vii along with other nvidia cards?