Sorry if this is a stupid request, but would it be possible to add a feature for naming gpu's? I know you can re-number them, but custom alphanumeric naming might help me take out some frustration on poorly performing cards.
Not a stupid request at all. We are going to add such feature in one of the later releases.
have been trying to use Phoenix on one of my small rigs -- keep getting a variety of cuda errors sometimes including illegal instruction -- with the miner crashing right after start -- it also now has decided not to write logfiles so i don't know what the problem is now -- running win 10 with 4gig ddr3 with 2-1060 3gb 2-1060 6gb 2 r9 280x and 1 rx570 8gb -- currently using claymore with no issues. Have increased the virtual memory to 60 gig from 30 gig added the -log 2 commands -- still crashes right away -- any suggestions would be appreciated. The rig is really just a spare parts rig -- but hashes quite well.
ok turns out part of this is win 10 -- lets just say it sucks .
The cuda errors are still happening will run for 20 minutes and then will crash -- running on ubq so dag shoud not be an issue -- cards are overclocked but not by much -- did try it with stock and same thing happens
Please try to specify
-nvf 3 and/or
-nvnew 0 to fix the problem.
Possible feature request.
How does PM handle the auto-tune settings? Are they “remembered” in any way? Or do you waste 5-10mins of sub optimal hashrates at the beginning of every time you start the miner?
If they aren’t remembered, maybe you could have the miner auto-append the config.txt or the start.bat file with the optimized gt values? That way you won’t have to auto-tune when simply re-running the miner.
Generally auto-tuning is supposed to be used only in the period of initial tuning of the rig, or after the OC settings are changed, or after other significant changes (e.g. new kernels, trying out -clkernel 2 and so on). After that you can press the 's' key in the console or check the miner console output from the remote manager and see the auto-tuned GT values, and copy them into your config.txt for permanent usage. However we can add a command-line option to append the found best -gt values to the config.txt file in a future version.
hI!
Anybody know how to mining Daggerhashimoto in Nicehash? Because if Nicehash "send" an small DAG it works at 22-23Mh but when goes to mining ethereum, drop to 8-9Mh.
Thanks.
Make sure that the AMD Compute mode is turned on. PhoenixMiner will print a warning on startup if a card supports Compute mode but it is not turned on (or you are using too old drivers that doesn't support Compute mode). Another possibility is that you are using older cards like 280X or 970, which have a steep drop off in hashrate on bigger DAG and there is no workaround for this. In such case you are better off mining Pirl or some other ethash coin with small DAG instead of using Nicehash.
Wanted to share the stats with the forum. These are the current stats of one of my rigs running on the PhoenixMiner 2.9e:
*** 1142:28 ***************************************************
Eth: Mining ETH on eu1.ethermine.org:4444 for 80:00
Available GPUs for mining:
GPU1: Radeon RX 580 Series (pcie 1), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU2: Radeon RX 580 Series (pcie 2), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU3: Radeon RX 580 Series (pcie 3), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU4: Radeon RX 580 Series (pcie 7), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU5: Radeon RX 580 Series (pcie 9), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU6: Radeon RX 580 Series (pcie 10), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU7: Radeon RX 580 Series (pcie 12), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU8: Radeon RX 580 Series (pcie 13), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU9: Radeon RX 580 Series (pcie 15), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU10: Radeon RX 580 Series (pcie 16), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU11: Radeon RX 580 Series (pcie 17), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU12: Radeon RX 580 Series (pcie 18), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU13: Radeon RX 580 Series (pcie 19), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU1: 54C 92%, GPU2: 57C 86%, GPU3: 57C 85%, GPU4: 56C 85%, GPU5: 54C 79%, GPU6: 52C 44%, GPU7: 52C 41%, GPU8: 58C 86%, GPU9: 54C 84%, GPU10: 54C 99%, GPU11: 52C 43%, GPU12: 62C 81%, GPU13: 49C 0%
Eth: Accepted shares 407093 (40 stales), rejected shares 3 (0 stales)
Eth: Incorrect shares 169 (0.04%), est. stales percentage 0.01%
Eth: Maximum difficulty of found share: 12954.2 TH (!!!)
Eth: Average speed (5 min): 407.232 MH/s
Eth: Effective speed: 395.92 MH/s; at pool: 395.92 MH/s
PhoenixMiner: If you ever have a moment of doubt about your work... get back to this post. THANKS!
Thanks a lot, much appreciated
After more tests, it seems auto-tuning IS working ...
But I can't make the -cvddc switch to work
If I run Claymore for a minute, it sets correct core voltage, I stop it and run Phoenix - doesn't matter if I use cvddc or not - it will use what Claymore had set. Same if set with Afterburner.
Are there some requirements for OS/Driver combo in order to use this switch? I tried on 7 and 10, relatively older drivers because I don't update what runs ok.
Yes, the OC settings work reliably only on 18.x.x drivers.
Moved from 2.9e to 3.0,
Version 3 is ok for my 470s, but crashed for my 1080ti's every couple of 10min with same command line. didnt disable the autotune yet
Tested with Clay 11.7 and Phoenix 2.9e with no issues
Auto-tune has no effect on Nvidia cards. Could you please send us logs from the crashes on 1080ti?
I tested v3.0c, some cards dropped 0 hashrate and restart, back to 2.9e.
Can you tell us what cards are having problems and what drivers you are using?
Hello, im having serius issues with connecting to ethermine pool for devfee shares. Do you have any solution to this?
This is highly unusual, perhaps some temporary problem with the pool. Please send us a log to be able to diagnose the problem any further.
Hi,
Any plans for linux (hiveos) version?
Yes, at least a generic Linux version should be ready in about a month.