It seems to start out fairly accurate, but as time increases it becomes less accurate... Is there some sort of sliding window that is used to calculate that averaged value that isn't used in the "Shares Per Minute:" variable?
Took a look at the code, need to refactor abit here. we have a minute counter which is triggered by an event, but if the computer is lagging extra seconds/minutes are added to the timer. This causes the Poolspeed calculated in the console to be lower than it really is. but got a crash.
Might be a memleak in the lhr code. edit: memleak found. fixing
The memleak in lhr_unlock has been fixed. and new binaries for windows uploaded. Linux tomorrow. |
|
|
|
v1.26 (fixed / windows) The rejects seems to be gone on the hiveon.io pool. the LHR 3060 hashing 30.3 mhash @ 74 watt  but got a crash. Might be a memleak in the lhr code. edit: memleak found. fixing |
|
|
|
I have uploaded a new fixed version of 1.26 on windows. Just a minor bugfix with amd+nvidia cards in the same rig, and fix for ethereum classic at mph.
How accurate is the reported "Shares Per Minute/Pool Hashrate" on the miner?
It seems to start out fairly accurate, but as time increases it becomes less accurate... Is there some sort of sliding window that is used to calculate that averaged value that isn't used in the "Shares Per Minute:" variable?
Should be Shares / The Number of minutes. But the minutes is rounded to the closest, so if you mine for 59 minutes and 31 seconds it will round to 60 they use a 30 minute average to calculate the current rate and I assume that you're using the most recent hashrate or maybe average hashrate?
We are using the average of xxx last jobs sendt to the gpu. [moderator's note: consecutive posts merged] |
|
|
|
yes. The lhr-unlock might need some adjustments. Please test. edit: Ethereum classic on miningpoolhub seems to have issues after the stratum change, ethereum mining works fine. Strange..
I understand there are three different miner versions for different cuda devices (11.5, 11.4,11.2). How do I know which version to use for a particular device?
use the 11.5 version if you have the latest driver installed. Should be a little bit faster as well
edit: Ethereum classic on miningpoolhub seems to have issues after the stratum change, ethereum mining works fine. Strange..
will re-upload a fixed version soon. Found another index bug opencl+cuda mixed rig. [moderator's note: consecutive posts merged] |
|
|
|
The repository you are linking from is from 2018. I was in the team that made the first ethereum gpu miner. back in 2015. The Gateless Gate Sharp version seems to be a slightly modified version of the original kernel.  |
|
|
|
Hello,
I am on TBM 1.24, I got a rx 6800 and a gtx 1660 super. If I try to set different xintensities for the cards -E 8,224 I end up having 8 for both of them. What am i doing wrong ?
Use brackets around them -E [8,224] . 8 is too low. I think 24 is better |
|
|
|
|
We have fixed a duplicate share bug in multicard rigs on the hiveon pool and possible other ethereum 1.0 stratum pools. The ~2% rejected shares are now ~0% in version 1.26 and this will improve the poolside speed..
|
|
|
|
|
On high oc on 1080ti and the 1070 --kernel 0 can help. you can try with a low --kernel number
|
|
|
|
sorry but as i wrote above, i already tried dagintensity 1 and it didn't help =(
then try this: --lock-cclock [[1200,1200],[1200,1200],[1200,1200],[1200,1200]] it will undervolt and give more power to the memory controller. |
|
|
|
|
try with --dag-intensity 1 --lock-cclock [[1200,1200],[1200,1200]]
|
|
|
|
|
Looks like 60-70 in the wall, but still pretty good.
|
|
|
|
Overall, my combined rig hashrate has moved from about 468 MH/s average to 502 MH/s; which is amazing and shows all the hard work that has gone into this software has really paid off for the mining community (and I hope for the Devs too!).
Thanks for your support and testing. The fee is picking up, but we have a long way to go. Hopefully the new semi LHR unlock feature will move more ppl over to the fastest miner.  |
|
|
|
|
A working semi LHR unlock has now been implemented. rtx 3060 32MHASH@50watt. Abit slower than the other teams but use very little power.
This new feature needs more tuning and testing, but v1.26 will be released soon.
|
|
|
|
|
Not sure. Why don't you ask in the MMPos thread. Did you use the latest version of mmpos? rocm-smi is normally included with the driver.
|
|
|
|
Still no stats on AMD cards... (ver 1.25)
yes. Tested on the latest hiveos. We are using a library called rocm-smi. This is buildt into hiveos, but on other linux platforms might need to install it separately. |
|
|
|
Good Rtx 2060  yes. v1.25 is really fast. Spread the word.. |
|
|
|
|
I have users who ask for Team Black Miner integration into minerstat. Is this something doable?
|
|
|
|
This is unlike other miners that do not natively support proxies like teamredminer. main post.
The other miners seem to use the same codebase because they behave the same. More hits on the antivirus software as well.. The Team Black miner is written from scratch by opensource developers... Proxy is supported in zil+etc and zil+eth miningmode. Mine at rustpool or shardpool. |
|
|
|
not close... but it's beyond the EST. REWARD by minerstat and whattomine  Yeah. This software is pretty good, but only if you configure it correctly... |
|
|
|
|
Show Posts
