Topic: CgWatcher throttling Hashpower every 2-3 seconds? (Read 774 times)
Start cgminer in "Realtime" priority
Have you set CGWatcher to use your batch file instead of pointing it to the executable to ensure all of the same settings are carried over? Once the miner is running, there is nothing that CGWatcher does other than request information via api from the miner. I don't believe CGWatcher uses enough resources to warrant a drop in hashrate... this is a relatively small amount of information and it happens instantly.
I suggest closing CGWatcher and launching the miner using your batch file. Wait about 10 minutes to ensure the hashrate has stabilized, then open CGWatcher. It should detect the miner and begin monitoring it. Watch for a change in hashrate once CGWatcher is running.
You can always go to the Tests tab and in the Debug Reports & Utilities drop-down list (middle of the window), select "Show last miner start configuration" and click Run. This will show you your original config settings and the exact settings used to launch the miner. If there are any differences, other than CGWatcher enabling the API and perhaps a couple additional settings listed but set to their default values, let me know.
The number of computers CGWatcher is running on displayed at cgwatcher.com is accurate and updated in real time. Out of the 43,000+ computers it is running on, only a handful of times has someone suggested it is affecting hashrate. This was almost always due to CGWatcher using the default gpu-threads value of 2, as was stated the default in cgminer's readme when it would actually run at 1 if not specified. I see this isn't the case here because you are setting gpu-threads to 2... but I'm sure there is an explanation somewhere.
I suggest closing CGWatcher and launching the miner using your batch file. Wait about 10 minutes to ensure the hashrate has stabilized, then open CGWatcher. It should detect the miner and begin monitoring it. Watch for a change in hashrate once CGWatcher is running.
You can always go to the Tests tab and in the Debug Reports & Utilities drop-down list (middle of the window), select "Show last miner start configuration" and click Run. This will show you your original config settings and the exact settings used to launch the miner. If there are any differences, other than CGWatcher enabling the API and perhaps a couple additional settings listed but set to their default values, let me know.
The number of computers CGWatcher is running on displayed at cgwatcher.com is accurate and updated in real time. Out of the 43,000+ computers it is running on, only a handful of times has someone suggested it is affecting hashrate. This was almost always due to CGWatcher using the default gpu-threads value of 2, as was stated the default in cgminer's readme when it would actually run at 1 if not specified. I see this isn't the case here because you are setting gpu-threads to 2... but I'm sure there is an explanation somewhere.
Hi guys,
I thought id ask for some advice with CGWatcher. Id love to be able to run this piece of software but for the past few weeks, whatever I try, I can't get it to not upset my GPUs whilst mining.
I run CGMiner 3.1.1 i find it stable enough for me. I know my cards inside out as to what their hash rates should be. I run 2x Rigs with 4x r9 280xs in each.
The rig in question has cards than run perfectly at 725khs. They fluctuate between 722-725 consistently all day everyday. I run the cgminer from my .bat file. I don't use a config file. On the .bat file, I don't have any auto GPU settings in there. I run the fans at 80% and that keeps the card at around 80 degrees which is fine for me.
Now, when I run CGWatcher... the cards run up to 725khs.. then drop to exactly 689 Everytime! Every 2-3 seconds too. Like I say, I have removed the auto temp control in CGWatcher.
Here is my config file...
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_USE_SYNC_OBJECTS 1
cgminer.exe --scrypt -o stratum+tcp://com:3333 -u 1XXXXXXXXXXXBJ -p x --failover-only -o stratum+tcp://.com:3333 -u XXXXXXXXXXXX -p x --lookup-gap 2 --thread-concurrency 8193,8193,8193,8193 -g 2 -I 13 -w 256 --gpu-fan 85,85,85,85 --gpu-memclock 1500,1500,1500,1500 --gpu-engine 1065,1055,1065,1065 --gpu-powertune -20,-20,-20,-20 --expiry 1 --scan-time 1 --queue 0
Any thoughts guys?
I thought id ask for some advice with CGWatcher. Id love to be able to run this piece of software but for the past few weeks, whatever I try, I can't get it to not upset my GPUs whilst mining.
I run CGMiner 3.1.1 i find it stable enough for me. I know my cards inside out as to what their hash rates should be. I run 2x Rigs with 4x r9 280xs in each.
The rig in question has cards than run perfectly at 725khs. They fluctuate between 722-725 consistently all day everyday. I run the cgminer from my .bat file. I don't use a config file. On the .bat file, I don't have any auto GPU settings in there. I run the fans at 80% and that keeps the card at around 80 degrees which is fine for me.
Now, when I run CGWatcher... the cards run up to 725khs.. then drop to exactly 689 Everytime! Every 2-3 seconds too. Like I say, I have removed the auto temp control in CGWatcher.
Here is my config file...
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_USE_SYNC_OBJECTS 1
cgminer.exe --scrypt -o stratum+tcp://com:3333 -u 1XXXXXXXXXXXBJ -p x --failover-only -o stratum+tcp://.com:3333 -u XXXXXXXXXXXX -p x --lookup-gap 2 --thread-concurrency 8193,8193,8193,8193 -g 2 -I 13 -w 256 --gpu-fan 85,85,85,85 --gpu-memclock 1500,1500,1500,1500 --gpu-engine 1065,1055,1065,1065 --gpu-powertune -20,-20,-20,-20 --expiry 1 --scan-time 1 --queue 0
Any thoughts guys?
Jump to: