|
core :1950, mem 4400. 90%% tdp 27mhash epoch 444
v1.10
|
|
|
|
|
v1.10 has reduced stale shares on NVIDA.
flexpool does't pay for stale shares so you need to play wity the ---xintensity parameter or change the pool
|
|
|
|
|
-y --retry-pause (Set retry pause between connection attempts if pool is not responding)
-t --timeout (Set timeout in seconds. Used for connecting and socket timeout) -a
|
|
|
|
TeamBlackMiner 1.10 was released yesterday that introduced API.
Yes, I looked into it earlier today but it didn't work (can be enabled, but doesn't respond to requests). I will give it a try on another system as well to see. It's just missing the http header, so wont work in browser or with curl. We added it into the linux build. Will be included in v1.11 |
|
|
|
I think there is a incompatibility with the AMD Stats or something, maybe?
you can try to run with --no-stats No, ETC on AMD have same problem with 16GB virtual memory settings.
Till TBM 1.06 I had some extra 2MH/s from AMD to 38MH/s from laptop RTX3060 (80W) on NiceHash ETH Can you try with --no-stats? Yes, it is working (1.09 - both devices on Nicehash ETH pool) but with no stats.  |
|
|
|
Device Number: 0
Device name: gfx90c
PCIe bus id: 16
Memory: 14GB
Internal graphics need a low value for --xintensity, but I think there is a hardcoded intensity to build the dagfile wich is causing the crash. Is it crashing before the dag file is generated? |
|
|
|
Thanks a lot, working great on Windows.
What is the ETA for linux version?
Today. Please wait a few hours |
|
|
|
|
use the cuda 11.4 build with the latest driver. If not use xintensity -1 and wait 5mins, or increase it.
|
|
|
|
When adding an API for TeamBlackMiner, please also include the PCI Bus ID for each of the GPUs you are reporting hashrate and other details for. This is how most mining management software identifies each GPU.
PCI Bus ID added. See api.txt for doc
Long line optimisation is needed. Are you sure "TYPE" column is really that useful? Everybody knows that AMD cards cant use CUDA and all Nvidia cards use CUDA for mining and not OpenCL.
Type is now moved up with the other stats  [moderator's note: consecutive posts merged] |
|
|
|
|
ssl doesn't work. Use the normal url / port
|
|
|
|
And do you need to enable SAM?
For the 6800XT on stock clocks slight power increase on higher intensity. Not sure about the 6800, but I expect similar results. --xintensity 1: 213W --xintensity 512: 230W v1.09 is producing abit high amount of stale shares when using high intensity. So you need to mine on a pool that pay for stales. v1.10 will have an improved dynamic intensity to reduce stales. And do you need to enable SAM?
Enable SAM for more speed increase. |
|
|
|
|
The Team Black Miner v1.09 wich is optimized for the latest cards, can do around 70MHASH on the rx 6800
|
|
|
|
I'm not getting any shares after running the miner for more than 20 minutes. Can you let me know how to fix it?
wich cards, pool and launchconfig?
1.10 is soon ready with API support and improved poolside hashrate on some pools.
The 12 hour test with the new dynamic intensity went good Poolside was a bit higher than reported, but still ~3% stales. --xintensity 64 is giving 4-5% stales. on hiveon.net [moderator's note: consecutive posts merged] |
|
|
|
1.10 is soon ready with API support and improved poolside hashrate on some pools. Testing on the https://hiveon.net/ pool  Will release tomorrow or Friday
Long line optimisation is needed. Are you sure "TYPE" column is really that useful? Everybody knows that AMD cards cant use CUDA and all Nvidia cards use CUDA for mining and not OpenCL.  We have moved the type column up to the other stats (mem/temp etc) in v1.10 [moderator's note: consecutive posts merged] |
|
|
|
I'm still testing , but this miner is giving me increased hashrates. Originally i also had alot more stale shares on ethermine.
Lowered xintensity to 48 and stale shares are almost non-existent now.
Nice find. --xintensity 48 seems to be abit better on etheremine. |
|
|
|
|
Show Posts
