Search content
Sort by
Showing 3 of 3 results by Grumble Bumpkin
Re: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v11.7 (Windows/Linux)
I have a strange problem with my miner since Windows 10 1803 and I wonder if any of you have had similar issues that you've solved. The rig is a tiny hobby system: a B250 Mining Expert with four RX580's of varying brands. All cards have modded BIOSs, and I overclock using my start.bat options.
I looked back through the mining logs from Claymore (11.2 at the time) to see that before the update restart, all cards were going at near 31 Mh/s, a rate that it had been maintaining for months with no problem. After the update, all cards were running at 18.5 Mh/s.
When I noticed, I updated my AMD drivers to 18.5.1 and set it to compute mode, and updated Claymore to 11.7. At that point, one of my four cards was back to 30 Mh/s, all the rest remained at 18.5 Mh/s. I've checked my VM, which is at 48000 MB, so I doubt that's an issue. I haven't managed to locate anyone else on here reporting a similar issue, so I'm wondering if any of you have dealt with this and if so, how?
Thanks!
I looked back through the mining logs from Claymore (11.2 at the time) to see that before the update restart, all cards were going at near 31 Mh/s, a rate that it had been maintaining for months with no problem. After the update, all cards were running at 18.5 Mh/s.
When I noticed, I updated my AMD drivers to 18.5.1 and set it to compute mode, and updated Claymore to 11.7. At that point, one of my four cards was back to 30 Mh/s, all the rest remained at 18.5 Mh/s. I've checked my VM, which is at 48000 MB, so I doubt that's an issue. I haven't managed to locate anyone else on here reporting a similar issue, so I'm wondering if any of you have dealt with this and if so, how?
Thanks!
That's really interesting, I have 13 cards on same board, and I just added one of them to the system. Should be the best one for mining but it's running at 18 mh/s and all others 25+.
Love to know what's going on there, I'm in newest CM and newest driver as well, Win10 1709
Did you set all GPUs to compute? It is a setting in the registry under each GPU. With AMD settings software you have to set them one by one. CM with Y or AMD-Compute-Swithcer will set all installed GPUs to compute, but when you add GPUs after that, they will not be set to compute by default. Sorry if you already knew this.
Fantastic catch, Jonet. In looking back to the Radeon settings, it appears that "global" settings is only for each card individually, which is a bit misleading lol. Anyway, you are correct and all cards are back up to 31 Mh/s now! Thanks!
Re: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v11.7 (Windows/Linux)
I have a strange problem with my miner since Windows 10 1803 and I wonder if any of you have had similar issues that you've solved. The rig is a tiny hobby system: a B250 Mining Expert with four RX580's of varying brands. All cards have modded BIOSs, and I overclock using my start.bat options.
I looked back through the mining logs from Claymore (11.2 at the time) to see that before the update restart, all cards were going at near 31 Mh/s, a rate that it had been maintaining for months with no problem. After the update, all cards were running at 18.5 Mh/s.
When I noticed, I updated my AMD drivers to 18.5.1 and set it to compute mode, and updated Claymore to 11.7. At that point, one of my four cards was back to 30 Mh/s, all the rest remained at 18.5 Mh/s. I've checked my VM, which is at 48000 MB, so I doubt that's an issue. I haven't managed to locate anyone else on here reporting a similar issue, so I'm wondering if any of you have dealt with this and if so, how?
Thanks!
I looked back through the mining logs from Claymore (11.2 at the time) to see that before the update restart, all cards were going at near 31 Mh/s, a rate that it had been maintaining for months with no problem. After the update, all cards were running at 18.5 Mh/s.
When I noticed, I updated my AMD drivers to 18.5.1 and set it to compute mode, and updated Claymore to 11.7. At that point, one of my four cards was back to 30 Mh/s, all the rest remained at 18.5 Mh/s. I've checked my VM, which is at 48000 MB, so I doubt that's an issue. I haven't managed to locate anyone else on here reporting a similar issue, so I'm wondering if any of you have dealt with this and if so, how?
Thanks!
Re: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v10.2 (Windows/Linux)
Edit: Last night I found the solution. I hadn't installed the proper [Suspicious link removed]pute driver from AMD (17.30). All cards are now running 28.5+ without overclocking.
First of all, love the program, so many thanks to the developer! Now on to the cry for help
To make it short, I've had a rig running two Gigabyte Gaming 8GB RX580's for a while now and I'm getting a stable 24.5 Mh/s out of each. I do mod the BIOSs using the One Click method on the current Polaris software, I've changed the VRAM to 16GB, and I try to overclock using MSI Afterburner a bit, though I feel like I'm not very good at it. Hash rate's not super hot, but for the moment I'm content.
I recently added on a Sapphire Nitro+ 8GB RX580, which I know some get 30+ Mh/s out of, hoping to get the same result as other posters. After much dinking around, modding the BIOS (both of them), switching BIOSs, and attempts at overclocking, I'm managing to get a measly 18.4 Mh/s out of the card (even mining with that card solo), which is extremely disappointing. So my question is: why is a verified bad ass card so much worse than my other cards, and what can I do to improve? All help is appreciated!
One thing I did not do is install any extra drivers after adding the Sapphire. I figured using the AMD drivers I already had for my previous 580's would suffice, but I may be entirely wrong.
My rig, in case it helps:
Windows 10 Home OS
Asus B250 Mining Expert MOBO
2x Gigabyte Gaming 8GB RX580s on PCIE x1 risers
1x Sapphire Nitro+ 8GB RX580 on PCIE x1 riser
Thanks!
First of all, love the program, so many thanks to the developer! Now on to the cry for help
To make it short, I've had a rig running two Gigabyte Gaming 8GB RX580's for a while now and I'm getting a stable 24.5 Mh/s out of each. I do mod the BIOSs using the One Click method on the current Polaris software, I've changed the VRAM to 16GB, and I try to overclock using MSI Afterburner a bit, though I feel like I'm not very good at it. Hash rate's not super hot, but for the moment I'm content.
I recently added on a Sapphire Nitro+ 8GB RX580, which I know some get 30+ Mh/s out of, hoping to get the same result as other posters. After much dinking around, modding the BIOS (both of them), switching BIOSs, and attempts at overclocking, I'm managing to get a measly 18.4 Mh/s out of the card (even mining with that card solo), which is extremely disappointing. So my question is: why is a verified bad ass card so much worse than my other cards, and what can I do to improve? All help is appreciated!
One thing I did not do is install any extra drivers after adding the Sapphire. I figured using the AMD drivers I already had for my previous 580's would suffice, but I may be entirely wrong.
My rig, in case it helps:
Windows 10 Home OS
Asus B250 Mining Expert MOBO
2x Gigabyte Gaming 8GB RX580s on PCIE x1 risers
1x Sapphire Nitro+ 8GB RX580 on PCIE x1 riser
Thanks!