NiceHashSupport
|
|
October 24, 2015, 01:55:30 PM |
|
Checking into that. Are you a miner among them with high stale rate? Could you provide us a ping report to USA server? Check FAQ how to do this. Looks like you already fix the problem. Thanks We did nothing to fix it. But here is what was happening; at time of high rejects rates, some customers had very wild orders that were constantly switching jobs and diffs. This caused quite some rejects being produced. Because we drop old jobs and don't keep information about them (thus we don't know diff of stale shares and use current diff for speed calculation), stale rejected speed when switching from low diff job to high diff job is highly exaggerated. Additionally to that, we reward miners with extra virtual shares when fast job switching is happening. True that you may have high reject rate, but if you check your accepted speed, it should be normal as any other time.
|
|
|
|
Icon
|
|
October 24, 2015, 04:08:27 PM Last edit: October 24, 2015, 04:22:53 PM by Icon |
|
Might just be me but, using latest nvidia drivers running win7 64 bit on a (nvidia3x) Zotac GTX 650, and latest nicehash miner 1.1.0.4 and still stuck on neoscrypt while running standard benchmark running ~ 5 mins now and still sitting there.
Did notice while benchmarking while running gpu-z the gpu is not even being used @ 0% while benchmarking neoscrypt.
Question, do we need any version of Microsoft's C++ runtime installed while running nicehash miner? I know while running standard CUDA miner package 2008/2011 sdk is needed in order for it to run.
Icon
|
|
|
|
NiceHashSupport
|
|
October 24, 2015, 05:26:07 PM |
|
Might just be me but, using latest nvidia drivers running win7 64 bit on a (nvidia3x) Zotac GTX 650, and latest nicehash miner 1.1.0.4 and still stuck on neoscrypt while running standard benchmark running ~ 5 mins now and still sitting there.
Did notice while benchmarking while running gpu-z the gpu is not even being used @ 0% while benchmarking neoscrypt.
Question, do we need any version of Microsoft's C++ runtime installed while running nicehash miner? I know while running standard CUDA miner package 2008/2011 sdk is needed in order for it to run.
Icon
No, nothing more is needed. Is the process ccminer_tpruvot.exe running? Manually run benchmark: ccminer_tpruvot.exe --algo=neoscrypt --benchmark --time-limit 60 Paste here what happens.
|
|
|
|
Icon
|
|
October 24, 2015, 08:17:10 PM |
|
D:\NiceHash\bin>ccminer_tpruvot.exe --algo=neoscrypt --benchmark --time-limit 60
*** ccminer 1.7-dev for nVidia GPUs by tpruvot@github *** Built with VC++ 2013 and nVidia CUDA SDK 6.5
Originally based on Christian Buchner and Christian H. project Include some of the work of djm34, sp, tsiv and klausT.
BTC donation address: 1AJdfCpLWPNoAMDfHF1wD5y8VgKSSTHxPo (tpruvot)
[2015-10-24 16:14:09] NVAPI GPU monitoring enabled. [2015-10-24 16:14:09] 1 miner thread started, using 'neoscrypt' algorithm. [2015-10-24 16:14:09] Sorry neoscrypt is not supported on SM 3.0 devices
D:\NiceHash\bin>
Seems my 650 cant run it... using nvidia drivers 358.50 and Physx 9.15.0428
Icon
|
|
|
|
NiceHashSupport
|
|
October 24, 2015, 09:34:44 PM |
|
D:\NiceHash\bin>ccminer_tpruvot.exe --algo=neoscrypt --benchmark --time-limit 60
*** ccminer 1.7-dev for nVidia GPUs by tpruvot@github *** Built with VC++ 2013 and nVidia CUDA SDK 6.5
Originally based on Christian Buchner and Christian H. project Include some of the work of djm34, sp, tsiv and klausT.
BTC donation address: 1AJdfCpLWPNoAMDfHF1wD5y8VgKSSTHxPo (tpruvot)
[2015-10-24 16:14:09] NVAPI GPU monitoring enabled. [2015-10-24 16:14:09] 1 miner thread started, using 'neoscrypt' algorithm. [2015-10-24 16:14:09] Sorry neoscrypt is not supported on SM 3.0 devices
D:\NiceHash\bin>
Seems my 650 cant run it... using nvidia drivers 358.50 and Physx 9.15.0428
Icon
In upcoming 1.1.0.5, this will be fixed. For now, you can manually set speeds in config file.
|
|
|
|
blue-yu
Newbie
Offline
Activity: 58
Merit: 0
|
|
October 25, 2015, 06:29:15 AM |
|
ScryptJaneNf16.
Is there any chance to mine Jane on Centos? Right now I'm mining Neoscrypt, but every attempt to compile something under linux has failed
|
|
|
|
PeaMine
|
|
October 25, 2015, 06:57:04 AM Last edit: October 25, 2015, 07:20:17 AM by PeaMine |
|
ScryptJaneNf16.
Is there any chance to mine Jane on Centos? Right now I'm mining Neoscrypt, but every attempt to compile something under linux has failed Was wondering this as well for Ubuntu 14.04 using AVX2 E5 processors. Got it to work, but it's extremely slow compared to even a E3-1230v3 using just 4 cores on Windows using the NiceHashMiner.exe
|
Datacenter Technician and Electrician. If you have any questions feel free to ask me as I am generally bored looking at logs and happy to help during free time.
|
|
|
NiceHashSupport
|
|
October 25, 2015, 07:59:05 AM |
|
ScryptJaneNf16.
Is there any chance to mine Jane on Centos? Right now I'm mining Neoscrypt, but every attempt to compile something under linux has failed Was wondering this as well for Ubuntu 14.04 using AVX2 E5 processors. Got it to work, but it's extremely slow compared to even a E3-1230v3 using just 4 cores on Windows using the NiceHashMiner.exe You need to set march=native when compiling
|
|
|
|
|
Bananana
|
|
October 26, 2015, 02:44:11 AM |
|
Checking into that. Are you a miner among them with high stale rate? Could you provide us a ping report to USA server? Check FAQ how to do this. Looks like you already fix the problem. Thanks We did nothing to fix it. But here is what was happening; at time of high rejects rates, some customers had very wild orders that were constantly switching jobs and diffs. This caused quite some rejects being produced. Because we drop old jobs and don't keep information about them (thus we don't know diff of stale shares and use current diff for speed calculation), stale rejected speed when switching from low diff job to high diff job is highly exaggerated. Additionally to that, we reward miners with extra virtual shares when fast job switching is happening. True that you may have high reject rate, but if you check your accepted speed, it should be normal as any other time. Thanks for the quick reply, and yes you are right the accepted speed is the same . More question: Is there a way to change the payout to every 24 hours instead of every 6 hours?
|
|
|
|
PeaMine
|
|
October 26, 2015, 04:17:52 AM |
|
ScryptJaneNf16.
Is there any chance to mine Jane on Centos? Right now I'm mining Neoscrypt, but every attempt to compile something under linux has failed Was wondering this as well for Ubuntu 14.04 using AVX2 E5 processors. Got it to work, but it's extremely slow compared to even a E3-1230v3 using just 4 cores on Windows using the NiceHashMiner.exe You need to set march=native when compiling Thanks, after using ./autogen, ./configure CFLAGS="-march=native" , ./build still getting 170H max using 7-16 cores of a http://ark.intel.com/products/81706/Intel-Xeon-Processor-E5-2660-v3-25M-Cache-2_60-GHz compared to E3-1230v3 4 cores in Windows which is 124H. Seems like it doesn't scale with the additional cores, 7 cores being the best for the 2660 but only slightly faster than just 4 cores of the E3. Using 7 cores on the E3 I'm able to reach around 150~h but it gets some rejects for some reason. Either way thanks for the help!
|
Datacenter Technician and Electrician. If you have any questions feel free to ask me as I am generally bored looking at logs and happy to help during free time.
|
|
|
tbearhere
Legendary
Offline
Activity: 3164
Merit: 1003
|
|
October 26, 2015, 07:20:58 AM |
|
Nicehash ... on nicehashminer for nvidia what version of ccminer does it use..why does it mine with two cmd windows ....why is it mining a worthless algo.. why 20 more questions but don't have the time to ask. Theres gotta be a better miner control.
What kind of GPUs you have inside? It will use tpruvots for non-maxwell cards and sp for maxwell cards. Make sure to complete benchmark correctly and that values reported in benchmark are what you expect. Nvidia.... why does it mine with two cmd windows ? And I have sp ccminer latest at higher hashrates. Some of the benchmarks are good.
|
|
|
|
NiceHashSupport
|
|
October 26, 2015, 07:23:35 AM |
|
ScryptJaneNf16.
Is there any chance to mine Jane on Centos? Right now I'm mining Neoscrypt, but every attempt to compile something under linux has failed Was wondering this as well for Ubuntu 14.04 using AVX2 E5 processors. Got it to work, but it's extremely slow compared to even a E3-1230v3 using just 4 cores on Windows using the NiceHashMiner.exe You need to set march=native when compiling Thanks, after using ./autogen, ./configure CFLAGS="-march=native" , ./build still getting 170H max using 7-16 cores of a http://ark.intel.com/products/81706/Intel-Xeon-Processor-E5-2660-v3-25M-Cache-2_60-GHz compared to E3-1230v3 4 cores in Windows which is 124H. Seems like it doesn't scale with the additional cores, 7 cores being the best for the 2660 but only slightly faster than just 4 cores of the E3. Using 7 cores on the E3 I'm able to reach around 150~h but it gets some rejects for some reason. Either way thanks for the help! I would suggest to debug this issue with tpruvot (Epsylon3 on this forum).
|
|
|
|
NiceHashSupport
|
|
October 26, 2015, 07:25:40 AM |
|
Nicehash ... on nicehashminer for nvidia what version of ccminer does it use..why does it mine with two cmd windows ....why is it mining a worthless algo.. why 20 more questions but don't have the time to ask. Theres gotta be a better miner control.
What kind of GPUs you have inside? It will use tpruvots for non-maxwell cards and sp for maxwell cards. Make sure to complete benchmark correctly and that values reported in benchmark are what you expect. Nvidia.... why does it mine with two cmd windows ? And I have sp ccminer latest at higher hashrates. Some of the benchmarks are good. Like being said already, it will mine with two, if you have a mix of maxwell and non-maxwell cards. You can insert in latest sp ccminer manually (bin folder).
|
|
|
|
tbearhere
Legendary
Offline
Activity: 3164
Merit: 1003
|
|
October 26, 2015, 07:31:42 AM |
|
Nicehash ... on nicehashminer for nvidia what version of ccminer does it use..why does it mine with two cmd windows ....why is it mining a worthless algo.. why 20 more questions but don't have the time to ask. Theres gotta be a better miner control.
What kind of GPUs you have inside? It will use tpruvots for non-maxwell cards and sp for maxwell cards. Make sure to complete benchmark correctly and that values reported in benchmark are what you expect. Nvidia.... why does it mine with two cmd windows ? And I have sp ccminer latest at higher hashrates. Some of the benchmarks are good. Like being said already, it will mine with two, if you have a mix of maxwell and non-maxwell cards. You can insert in latest sp ccminer manually (bin folder). Ok thank you....but if I open 2 cmd windows with my sp ccminer the haahrates drop. How come they don't with nicehash miner? And how can I get it to mine with only one algo at a time? thx
|
|
|
|
NiceHashSupport
|
|
October 26, 2015, 07:39:03 AM |
|
The hashrate does not drop, because sp miner is used on maxwell cards and tpruvot miner for 3.x cards, they don't interfere.
You already mine single algo - whichever is most profitable. If you ran benchmark previously, then NHM will autoselect best algorithm. Each ccminer instance is doing this - thus it can happen that sp miner is mining one algo and tpruvot another algo, because the ratio of speeds are not the same.
If you wish to mine with only 5.x devices, you can disable 3.x.
Support for disabling individual algorithms comes in 1.1.0.5.
|
|
|
|
tbearhere
Legendary
Offline
Activity: 3164
Merit: 1003
|
|
October 26, 2015, 07:49:31 AM Last edit: October 26, 2015, 08:17:38 AM by tbearhere |
|
The hashrate does not drop, because sp miner is used on maxwell cards and tpruvot miner for 3.x cards, they don't interfere.
You already mine single algo - whichever is most profitable. If you ran benchmark previously, then NHM will autoselect best algorithm. Each ccminer instance is doing this - thus it can happen that sp miner is mining one algo and tpruvot another algo, because the ratio of speeds are not the same.
If you wish to mine with only 5.x devices, you can disable 3.x.
Support for disabling individual algorithms comes in 1.1.0.5.
Ok I have only 5.x and 5.2x. and I need it to mine with the best algo only..1 cmd and on this computer only 2 of my 3 cards. Let me do the benchmark again because I think it did it with 3.x. The 750ti. is 5.x I think. EDIT:What I really need it to do is mine with only 2 of the 3 cards on this computer only. On my rigs it doesn't matter. Nice benchmark tester you have there.
|
|
|
|
NiceHashSupport
|
|
October 26, 2015, 08:16:21 AM |
|
The hashrate does not drop, because sp miner is used on maxwell cards and tpruvot miner for 3.x cards, they don't interfere.
You already mine single algo - whichever is most profitable. If you ran benchmark previously, then NHM will autoselect best algorithm. Each ccminer instance is doing this - thus it can happen that sp miner is mining one algo and tpruvot another algo, because the ratio of speeds are not the same.
If you wish to mine with only 5.x devices, you can disable 3.x.
Support for disabling individual algorithms comes in 1.1.0.5.
Ok I have only 5.x and 5.2x. and I need it to mine with the best algo only..1 cmd and on this computer only 2 of my 3 cards. Let me do the benchmark again because I think it did it with 3.x. The 750ti. is 5.x I think. When you open NHM, you can see how it qualifies devices. If some are recognized as NVIDIA3.x and you only have 5.x, then of course, this is a bug and has to be investigated why it happens.
|
|
|
|
tbearhere
Legendary
Offline
Activity: 3164
Merit: 1003
|
|
October 26, 2015, 08:20:34 AM |
|
The hashrate does not drop, because sp miner is used on maxwell cards and tpruvot miner for 3.x cards, they don't interfere.
You already mine single algo - whichever is most profitable. If you ran benchmark previously, then NHM will autoselect best algorithm. Each ccminer instance is doing this - thus it can happen that sp miner is mining one algo and tpruvot another algo, because the ratio of speeds are not the same.
If you wish to mine with only 5.x devices, you can disable 3.x.
Support for disabling individual algorithms comes in 1.1.0.5.
Ok I have only 5.x and 5.2x. and I need it to mine with the best algo only..1 cmd and on this computer only 2 of my 3 cards. Let me do the benchmark again because I think it did it with 3.x. The 750ti. is 5.x I think. When you open NHM, you can see how it qualifies devices. If some are recognized as NVIDIA3.x and you only have 5.x, then of course, this is a bug and has to be investigated why it happens. What I really need it to do is mine with only 2 of the 3 cards on this computer only. On my rigs it doesn't matter. Nice benchmark tester you have there. The 5x throw me off at first.. 5.0 = 750ti 5x = all 5. series I see. It did the benchmark properly with nvidia5x. Ok the worker name can be anything like x or just leave it at worker1.. correct?
|
|
|
|
NiceHashSupport
|
|
October 26, 2015, 08:31:33 AM |
|
The hashrate does not drop, because sp miner is used on maxwell cards and tpruvot miner for 3.x cards, they don't interfere.
You already mine single algo - whichever is most profitable. If you ran benchmark previously, then NHM will autoselect best algorithm. Each ccminer instance is doing this - thus it can happen that sp miner is mining one algo and tpruvot another algo, because the ratio of speeds are not the same.
If you wish to mine with only 5.x devices, you can disable 3.x.
Support for disabling individual algorithms comes in 1.1.0.5.
Ok I have only 5.x and 5.2x. and I need it to mine with the best algo only..1 cmd and on this computer only 2 of my 3 cards. Let me do the benchmark again because I think it did it with 3.x. The 750ti. is 5.x I think. When you open NHM, you can see how it qualifies devices. If some are recognized as NVIDIA3.x and you only have 5.x, then of course, this is a bug and has to be investigated why it happens. What I really need it to do is mine with only 2 of the 3 cards on this computer only. On my rigs it doesn't matter. Nice benchmark tester you have there. The 5x throw me off at first.. 5.0 = 750ti 5x = all 5. series I see. It did the benchmark properly with nvidia5x. Ok the worker name can be anything like x or just leave it at worker1.. correct? You have table where you can click on checkbox - if this checkbox is disabled, then it will not use this video card for mining. After that, you should run benchmark, so it benchmarks only cards that you want to use for mining. Worker can be anything, even empty.
|
|
|
|
|