Hi,
I have the following options in the config-file
But the fan speed stays at 20%
What am I doing wrong?
-tt 75
-powlim 50
-fanmin 70
-fanmax 80
GPU's are RX480's
Thanks for the help
regards,
Eric
Topic: Claymore's ZCash/BTG AMD GPU Miner v12.6 (Windows/Linux) - page 351. (Read 3839163 times)
I'm sorry, maybe something was misunderstood, but why 9.1 version still shows 90 seconds of work for the Creator? Like in the description clearly that it was 72 seconds. Real time is not measured, but was curious to measure. Another such great moment, I have not written backup pools, at some point my pool has stopped responding, and what do you think? The manner worked for the Creator for about an hour continuously! ) Very cool of course, I understand, but it's like that then you need to consider, but then after that, the miner operate without switching on Creator a couple of days.
OK, a bit more update.
Tried v9.1 on the same rig I did the earlier testing - same results as for v9.0
Tried 16.9.2 drivers - same results on each of v8 v9 and v9.1 as under 15.12 drivers - forget 16.12.1 the drivers are bad for mining.
Tried one of my "just arrived" RX 470 cards on that rig - v8 v9 and v9.1 ALL went into "crash as soon as the number of cards was reported" mode.
Gave up on that rig and moved it back to the HD 7870 and v8.
Built (rebuilt to be picky) a new rig with 1 x RX 470.
Catalyst 16.9.2, same Seasonic PS, 4GB RAM.
v8 works. v9 still flakes out. v9.1 works
One thing I notice in GPU-Z - memory use on the card is in excess of 2 GB, which makes me believe that part of the issues with the original testing rig was "lack of system RAM".
Add second card, get driver issues sorted out so both are working correctly.
v8 works. v9 still flakes out. v9.1 works
I don't know what changed between v9 and v9.1 but it appears that v9.1 fixed SOME sort of issue, and moving to a machine with 4GB of RAM fixed another (possibly related) issue.
Up side - v9.1 on the new rig DOES show ballpark 10% better hashrate than v8
Tried v9.1 on the same rig I did the earlier testing - same results as for v9.0
Tried 16.9.2 drivers - same results on each of v8 v9 and v9.1 as under 15.12 drivers - forget 16.12.1 the drivers are bad for mining.
Tried one of my "just arrived" RX 470 cards on that rig - v8 v9 and v9.1 ALL went into "crash as soon as the number of cards was reported" mode.
Gave up on that rig and moved it back to the HD 7870 and v8.
Built (rebuilt to be picky) a new rig with 1 x RX 470.
Catalyst 16.9.2, same Seasonic PS, 4GB RAM.
v8 works. v9 still flakes out. v9.1 works
One thing I notice in GPU-Z - memory use on the card is in excess of 2 GB, which makes me believe that part of the issues with the original testing rig was "lack of system RAM".
Add second card, get driver issues sorted out so both are working correctly.
v8 works. v9 still flakes out. v9.1 works
I don't know what changed between v9 and v9.1 but it appears that v9.1 fixed SOME sort of issue, and moving to a machine with 4GB of RAM fixed another (possibly related) issue.
Up side - v9.1 on the new rig DOES show ballpark 10% better hashrate than v8
You don't make based on the pool lmao except that you may not make based on the pool honesty. Coin mine and flypool are stealing shares plus a 1% plus of rejected shares.
If you're seeing 1% rejected shares on flypool, you have an issue with your ISP or with your mining rig(s).
I've NEVER seen that high a rejection rate on that pool (exception for my testing of Claymore v9 above, but that's an obvious issue with the software itself).
I normally get a little less than 1/2% rejects on any of my rigs that are running v8 (the single rig running v9.1 hasn't been running long enough to judge yet).
i heard the 8 gig cards usually have a better quality of ram when it comes to overclocking.
8 GB cards USUALLY have 8000 Mhz RAM - 4GB cards usually have 6600 or 7000.
when using dwarfpool for ZEC mining
how important is that port number for the difficulty/shares?
when i change it to say like 3334 instead of 3333 i seem to get errors from the client
i have 4 rigs one around 250, 480, 900 & 1.1k/hs
want to get the best out of them so should i do my best to set the right port number for the size of the rig?
how important is that port number for the difficulty/shares?
when i change it to say like 3334 instead of 3333 i seem to get errors from the client
i have 4 rigs one around 250, 480, 900 & 1.1k/hs
want to get the best out of them so should i do my best to set the right port number for the size of the rig?
Dwarfpool's vardiff (port 3333) works well. It adjusts the difficulty so that each worker is submitting approximately 200 shares per hour +/-10%.
Unless you have a more specific reason to prefer static difficulty, I'd suggest sticking with 3333 and forgetting about it.
nice, thanks for explaining
SSL was implemented long time ago for mining already but was removed again due to the increased network lag.
Lags can only create a pool if it spends a lot of time to encrypt/decrypt data, am I correct?
SSL connection is based on same sockets and it takes almost zero time on client side to encrypt/decrypt 1Kb of data.
I'm curious about server load, is it too expensive to use encrypted connections for you?
lol. fuck, this "delay" thingy is a total crap. I'm coming out of the heavily used encrypted pipe (being aes-128-ed on a 30$ wrt router;) with no delays whatsoever so why would I fucking care about few more processed BYTES on my end??? and in regard to the pool side - just watch and see and act accordingly.
also it seems sgminer-gm is improving
That is what I am saying, people blow it out of proportion. Initial connection SSL adds ~100ms for the handshake, but then after that it adds ~0.0000000000001ms per request, since encrypting like a 200 byte packet takes virtually no time, and is probably less than the packet size anyway, so essentially it adds 0 latency. Maybe it would matter on a 56K connection. My rejected rate on 3 different rigs is 0.08%, same as without SSL IIRC, and pool shows 0 stale.
Many sites use SSL, including all google sites, and any shopping site, or generally whenever you login to anything. There may be hundreds of kilobytes, or even megabytes to load a page, and everything is still really fast.
when using dwarfpool for ZEC mining
how important is that port number for the difficulty/shares?
when i change it to say like 3334 instead of 3333 i seem to get errors from the client
i have 4 rigs one around 250, 480, 900 & 1.1k/hs
want to get the best out of them so should i do my best to set the right port number for the size of the rig?
how important is that port number for the difficulty/shares?
when i change it to say like 3334 instead of 3333 i seem to get errors from the client
i have 4 rigs one around 250, 480, 900 & 1.1k/hs
want to get the best out of them so should i do my best to set the right port number for the size of the rig?
Dwarfpool's vardiff (port 3333) works well. It adjusts the difficulty so that each worker is submitting approximately 200 shares per hour +/-10%.
Unless you have a more specific reason to prefer static difficulty, I'd suggest sticking with 3333 and forgetting about it.
Can you guys help me. I have a problem with the miner. Yesterday my Zec total speed is 80-90H/s. When I restarted the PC it now becomes 12H/s.
How can I get back to 90H/s?
GPU: r7 260x -
I already tried v9.0, v9.1, v8.
How can I get back to 90H/s?
GPU: r7 260x -
I already tried v9.0, v9.1, v8.
use your intensity 2 or 3 or 4.
when using dwarfpool for ZEC mining
how important is that port number for the difficulty/shares?
when i change it to say like 3334 instead of 3333 i seem to get errors from the client
i have 4 rigs one around 250, 480, 900 & 1.1k/hs
want to get the best out of them so should i do my best to set the right port number for the size of the rig?
ohhh.... good to be back mining again after a long pause! roll of a super cold winter lol
how important is that port number for the difficulty/shares?
when i change it to say like 3334 instead of 3333 i seem to get errors from the client
i have 4 rigs one around 250, 480, 900 & 1.1k/hs
want to get the best out of them so should i do my best to set the right port number for the size of the rig?
ohhh.... good to be back mining again after a long pause! roll of a super cold winter lol
9.1 changes nothing than fee on SSL.
so go for SSL sites and you have more money
so go for SSL sites and you have more money
do u guys saw an increase in zec earnings after 24 hours? I'm getting the exact same amount. Maybe we should focus in power efficiency instead of hashing speed.
1
That's the way I see it also.
v9.1 crashes using anything above -i 3 with my furies, where as v9.0 was stable -i 5 with the same settings?
Can you give a few more details on your setup? I ran 6 Fury Nitros on -i 6 with v9.0 rock solid (15.18 drivers). I had tons of trouble when I went to v9.1, but I also updated my drivers to 16.12 at the same time so I haven't been able to pin-point the exact cause. I'm rolling the drivers back now to test v9.1 on 15.12 and 15.18.
I'm using 16.9.2 with a minor undervolt, -10 power limit, 1050/500.
Seems to crash at anything above I 3. Strangely enough, the same settings seems to work on my gaming PC with furies.
The mining rig has a 1300w evga g2 with roughly a 800w-900w load. My kill a watt died on me, so I'm not 100% sure
I assume the psu isnt the issue, even with the power spikes.
I have hi-end mainboard (Z87) with 7 PCI-E slots.
I can't use more than 6 videocards on it.
Theres's a lack of PCI-E lines.
I can't use more than 6 videocards on it.
Theres's a lack of PCI-E lines.
Z87 supports up to 8 PCI-E lanes. Check the spec here http://ark.intel.com/products/75013.
But in your case, there could have been a limitation in the motherboard itself in how it assigns resources to the various devices on it.
PCI-e Express 1X to 3 Port 1X card? - pointless, same reason.
I've read forum posts where it was possible to use the PCI-e Express 1X to 3 Port 1X card to run 8 GPUs on a 6 PIC-E slot motherboard. Search the Ethereum forum to see for yourself.
More than 6 - only on X79, X99 etc...
Any chipset that supports 8 lanes can theoretically support more than 6 GPUs. MSI Z97 Gaming 5 board can run 7 GPUs using its 7 PCI-E slots (as documented by Cryptomined on his YouTube channel), but people have found that it's generally not worth the effort. It's wise to save time and effort by running 6 cards on motherboards without too many bells and whistles.
Can you guys help me. I have a problem with the miner. Yesterday my Zec total speed is 80-90H/s. When I restarted the PC it now becomes 12H/s.
How can I get back to 90H/s?
GPU: r7 260x -
I already tried v9.0, v9.1, v8.
How can I get back to 90H/s?
GPU: r7 260x -
I already tried v9.0, v9.1, v8.
Hi,
I have couple of questions and it will be nice if someone can answer.
Does anybody use PCI-e Express 1X to 3 Port 1X card?
Does it has any effect on mining speed?
How many gpu we can have on Win10? Win7 and Win8 have limitation on number of gpu.
Does Claymore miner or OpenCL have some limitation on nuber of gpu?
BR
I have couple of questions and it will be nice if someone can answer.
Does anybody use PCI-e Express 1X to 3 Port 1X card?
Does it has any effect on mining speed?
How many gpu we can have on Win10? Win7 and Win8 have limitation on number of gpu.
Does Claymore miner or OpenCL have some limitation on nuber of gpu?
BR
I have hi-end mainboard (Z87) with 7 PCI-E slots.
I can't use more than 6 videocards on it.
Theres's a lack of PCI-E lines.
PCI-e Express 1X to 3 Port 1X card? - pointless, same reason.
More than 6 - only on X79, X99 etc...
i have z87 mainboard ( asrock fatality z87 ) and it works with 7 gpus ( gtx 1070 g1 ) , i use it with 8 gpus in windows 10 . 8th one is the integrated intel hd
Hi,
I have couple of questions and it will be nice if someone can answer.
Does anybody use PCI-e Express 1X to 3 Port 1X card?
Does it has any effect on mining speed?
How many gpu we can have on Win10? Win7 and Win8 have limitation on number of gpu.
Does Claymore miner or OpenCL have some limitation on nuber of gpu?
BR
I have couple of questions and it will be nice if someone can answer.
Does anybody use PCI-e Express 1X to 3 Port 1X card?
Does it has any effect on mining speed?
How many gpu we can have on Win10? Win7 and Win8 have limitation on number of gpu.
Does Claymore miner or OpenCL have some limitation on nuber of gpu?
BR
I tried the PCI-e Express 1X to 3 Port 1X card a couple of times in Win7 x64 but it didn't work. I read somewhere that you need Win10 to make it work. I doubt it would have any effects on mining speeds.
I have seen people report success running 8 GPUs in Windows 7 and 10, but it takes special circumstances and settings to make it work. Not worth the hassle is the general opinion. Seems it's safer to stick to 6 cards.
Claymore doesn't have limitation on number of GPUs (can use 10+ but switch on/off function and -di option only supports up to 9), nor does OpenCL. The limitation is usually in the driver and/or the OS.
In new version will up speed? Or it's all?
I'm guessing there is still a 30% total untapped potential in the hash rates / miner if I go by the power usage.
Ethereum+Dec or other 100% GPU taxing software uses about 30% more energy when I measure it using my Killawatt, vs the Claymore v9.x miners.
Load on the memory controller for an RX480 is around 64% with -i 4, I think there is potential for another 10-20% speedup.
In new version will up speed? Or it's all?
I'm guessing there is still a 30% total untapped potential in the hash rates / miner if I go by the power usage.
Ethereum+Dec or other 100% GPU taxing software uses about 30% more energy when I measure it using my Killawatt, vs the Claymore v9.x miners.
In new version will up speed? Or it's all?
Jump to: