I’ve been fighting this card for quite some time. It shows up fine in hive, I can set my OC settings, but it crashes the miner. I have isolated this card to a separate system and it works just fine in windows using TRM and SRB, it’s the weirdest thing.
I have another card, exact model works flawlessly in hive os. I’ve tried stock vbios, modded vbios(which is what I’m running in windows) it just keeps crashing the miner. I’ve tried older releases of different miners too.
Anyone have any advice besides keeping it running in windows?
I have my clocks set that way for economy, best hash per watt settings, I know I can crank it up and get around 13mh/s per card, but at an extra 20-25w per card, wasn’t economical for me.
I get 10.5 - 12 mh/s per card at 58-78 watts per card.
But even running stock vbios and clock/mem settings, card will not run in hiveos
You might want to try downgrading the HIVEOS to 06-0193.210831 using the upgrade/downgrade button at the top of HIVE. I had an RX 570-80 rig that was working great then I added 2 Vegas and nothing would run on the latest version. I took out all the cards and then added one by one. Turns out the Vegas which were XFX non reference Double Editions caused the problem. After trying ALL versions of HIVE, the only one that is working and stable is the 0193. Others below that would start but then crash after a bit. The version I am using now is solid for 24hrs so far.
Forgot to add used the reboot after downgrade/upgrade option each time. I know you get the msgs saying this or that feature is being removed, but I am still able to use Team Red without any issues - so far.
But it could be the OC for kawpow as well since what works in Win does not always work in HIVE without some extra tuning.
Since your miner states core is 950 and mem is 1750 you should put these in hiveos or higher. The miner is using these settings in windows despite the fact you set different settings.
So if I understand correctly, reset core, mem clocks and voltages to factory in windows, then move card over to hiveos? I do t know if I’ve tried that yet, I’ve tried stock settings in hiveos and it was no good, maybe I’ll have time later this week to try it out
no, settings doesnt matter when you transfer from one operating system to another because they are not saved in the GPU somewhere. they are implemented when the OS loads because they are in the OS. so here is what happens, you press the button and the PC starts to load. it runs all hardware on default settings until the OS boots and tells them otherwise. Meaning GPU starts with default settings and OC kicks in when the windows or hiveos app tells it now you need to run with the following parameters.
so when you transfer the GPU to hiveOS it will start with the stock settings, then you will have to OC it in hiveOS web. so standard settings for mining raven coin might be something like 1200-1300 core clock 875-925 core voltage, memory controller voltage 800 and memory clock 1750-2050.
That’s what I thought, but I’m a hobbyist and don’t do this for a living. So in that case, I’ve tried that with basic/stock clocks and voltages in hiveos and tried undervolting with lower clocks, just won’t work.
I keep this card in a dedicated tower, plugged directly into a PCIE slot and disconnect the SSD and boot hiveos on my flash drive and get the same error as when it was on a riser in my rig. It’s never mined in hiveos, always errors out the miner program, tried most of the popular ones. Works fine in windows though, so I’m at a loss, just wanted it in hive os for convenience, Remote Desktop tends to crash global wattman when I shutdown miner to make changes to clocks or coin I’m mining
start with standard OC settings and then test to lower them . first get it running.
try settings 1100 core clock and 2000 mem. for raven i dont thing you can reduce core clock and still mine. raven wants more core clock usually.
for raven i tested my card with 1125 core clock , 825 core voltage, 2050 memory clock and 800 memory controller voltage. test with these first.
