carlo_0000
Legendary
Offline
Activity: 1281
Merit: 1003
|
|
June 13, 2016, 01:32:43 PM |
|
1.1.3 i m a bit confuse what s the last version? is it not the 0.9.41?
|
|
|
|
|
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
|
|
|
|
Genoil (OP)
|
|
June 13, 2016, 02:42:31 PM |
|
1.1.3 i m a bit confuse what s the last version? is it not the 0.9.41? 0.9.41 was the version of ethminer when I forked it. 1.1.3 is the release number of the fork.
|
ETH: 0xeb9310b185455f863f526dab3d245809f6854b4d BTC: 1Nu2fMCEBjmnLzqb8qUJpKgq5RoEWFhNcW
|
|
|
uncaer9
|
|
June 14, 2016, 05:03:07 AM |
|
Im use 2x R9 390 my OC : power limit +10% core clock 1100mhs memory 1700mhs fan speed auto (usage device) Gpu temperatur stable at 74
i get black scren(blank) after 2hours then i reduce the OC of memory to 1600mhz but i get blank screen again after 1hours.
anyone can help me ? im use genoil too.
just information, this is my first problem, before it im runing in 24hours without any problem like this
|
|
|
|
Bokelka
|
|
June 14, 2016, 02:55:23 PM |
|
Im use 2x R9 390 my OC : power limit +10% core clock 1100mhs memory 1700mhs fan speed auto (usage device) Gpu temperatur stable at 74
i get black scren(blank) after 2hours then i reduce the OC of memory to 1600mhz but i get blank screen again after 1hours.
anyone can help me ? im use genoil too.
just information, this is my first problem, before it im runing in 24hours without any problem like this
For the R9 390, When you run the core at 1100 MHz, the memory freqeuncy can be 1500 or 1250. 1600 MHz is too high.
|
|
|
|
FFI2013
|
|
June 14, 2016, 11:51:39 PM |
|
genoil I have two questions first I've been using your miner on ethermine.org and works great but I recently started renting on miningrigrentals and the last few rentals got nothing but rejects so I found out one was on dwarfpool is your miner supported there cause MRR support told me it is and was heavily tested and the last where can I find a list of commands for your miner Thank You
|
|
|
|
merc84
|
|
June 15, 2016, 09:40:28 AM |
|
I've got some rigs running at Win8 with 14.4 whql drivers. Miner 1.0.8 is working just fine (-S -G --farm-recheck 400). However, miner 1.1.3 does not accept/reject any shares.. it just shows [A0+0:R0:F0] all the time. It does show mining speed (pretty similar to 1.0.8 ) in the mining window.. Just not accepting shares. That's odd, haven't seen that before. What videocards are you using? I have seen this exact behavior with 280x and 14.4 drivers using stock etherminer, it is a problem with the drivers I suggest using 15.12 cl files. You can get the cl files from my dropbox for simplicity sake if u need too https://www.dropbox.com/s/7h19jw7xezpa2i7/Amd%2015.12%20dll.rar?dl=0 just extract them to the same folder as the miner.
|
|
|
|
bensam1231
Legendary
Offline
Activity: 1750
Merit: 1024
|
|
June 15, 2016, 06:36:16 PM |
|
SO I've figured out what's been causing some weird crashing on my rigs. Some rigs have smaller SSDs (32GB). With the DAG files from Ethash this is bumping right up against the disk space limit (1.0. . It seems as though ethminer does not gracefully handle running out of diskspace. When it does, some weird things start happening. The miner just hard locks with no output or sometimes the whole OS will lock. Since there is a small pagefile on the disk, this could be a combination of pagefile access, new dag file generation, and/or the OS trying to add to the pagefile when there is literally no diskspace left. Since when a new dag file is generated it essentially uses 3x one dag file, this can consume a lot of space. I originally thought this was a bad OC, but it doesn't appear to be the case. I have since fixed it by removing the pagefile on the disk, but since 1.1.3 isn't stable and has slightly lower hashrate I'm still running the older version. Genoil consider adding the streaming dag as a option or original dag file generation back in as a option in the newer versions till streaming dags get ironed out.
|
I buy private Nvidia miners. Send information and/or inquiries to my PM box.
|
|
|
jdc4429
Newbie
Offline
Activity: 13
Merit: 0
|
|
June 16, 2016, 12:27:19 AM |
|
I don't know where Claymore gets his additional 5-ish percent performance increase from, but I really can't keep up with the features he keeps adding to his miner. But mine is free I don't think Claymore's miner is faster, I think he gets the increase because his stratum protocol works allowing a direct connection to Dwarf. Supposedly this makes a big difference going direct rather then over eth-proxy... If you got stratum working with dwarf I think you would be just as fast. Great miner. Thank you for your work. Good way to verify would be testing solo mode with each as solo mode must use HTTP.
|
|
|
|
davembg
Sr. Member
Offline
Activity: 340
Merit: 251
Smell the glove.
|
|
June 16, 2016, 05:25:52 AM |
|
1.1.3 i m a bit confuse what s the last version? is it not the 0.9.41? 0.9.41 was the version of ethminer when I forked it. 1.1.3 is the release number of the fork. Is the stratum code working in the OpenCL build of 1.1.4? It builds fine, and runs great using network - but, I'm getting a missing symbol when using a -S (stratum) config with Ubuntu 15.10. Does it need a link to boost? or is that built-in? Anyone else running stratum with 1.1.4? If so, would you please share your config line. thanks!
|
|
|
|
Amph
Legendary
Offline
Activity: 3206
Merit: 1069
|
|
June 16, 2016, 05:41:15 AM |
|
I don't know where Claymore gets his additional 5-ish percent performance increase from, but I really can't keep up with the features he keeps adding to his miner. But mine is free I don't think Claymore's miner is faster, I think he gets the increase because his stratum protocol works allowing a direct connection to Dwarf. Supposedly this makes a big difference going direct rather then over eth-proxy... If you got stratum working with dwarf I think you would be just as fast. Great miner. Thank you for your work. Good way to verify would be testing solo mode with each as solo mode must use HTTP. Actually, while I can get 6-7-ish 10 percent right now using Genoil's miner, I don't think Claymore took my route. I think you're mostly right - it's the host code. 10% on amd only right? because if it is on the 970 too, that's what i call impressive
|
|
|
|
Genoil (OP)
|
|
June 16, 2016, 07:39:19 AM |
|
970 already has those 10% because of natively supported warp shuffle and better compiler.
|
ETH: 0xeb9310b185455f863f526dab3d245809f6854b4d BTC: 1Nu2fMCEBjmnLzqb8qUJpKgq5RoEWFhNcW
|
|
|
Genoil (OP)
|
|
June 16, 2016, 08:26:40 AM |
|
970 already has those 10% because of natively supported warp shuffle and better compiler.
Not entirely sure about that. I took some ghetto math shortcuts. I know and I should try hacking those into PTX if possible at all. But with the current CUDA kernel I can simply run the first keccak stage twice and not lose any performance, so I don't think it will do a great deal. Anyone using maxas here? https://github.com/NervanaSystems/maxas
|
ETH: 0xeb9310b185455f863f526dab3d245809f6854b4d BTC: 1Nu2fMCEBjmnLzqb8qUJpKgq5RoEWFhNcW
|
|
|
Genoil (OP)
|
|
June 16, 2016, 08:42:10 AM |
|
970 already has those 10% because of natively supported warp shuffle and better compiler.
Not entirely sure about that. I took some ghetto math shortcuts. I know and I should try hacking those into PTX if possible at all. But with the current CUDA kernel I can simply run the first keccak stage twice and not lose any performance, so I don't think it will do a great deal. Anyone using maxas here? https://github.com/NervanaSystems/maxasNo, no, I meant in the rest of the code. I haven't seen that so yes you could be right
|
ETH: 0xeb9310b185455f863f526dab3d245809f6854b4d BTC: 1Nu2fMCEBjmnLzqb8qUJpKgq5RoEWFhNcW
|
|
|
uncaer9
|
|
June 16, 2016, 08:57:04 AM |
|
Im use 2x R9 390 my OC : power limit +10% core clock 1100mhs memory 1700mhs fan speed auto (usage device) Gpu temperatur stable at 74
i get black scren(blank) after 2hours then i reduce the OC of memory to 1600mhz but i get blank screen again after 1hours.
anyone can help me ? im use genoil too.
just information, this is my first problem, before it im runing in 24hours without any problem like this
For the R9 390, When you run the core at 1100 MHz, the memory freqeuncy can be 1500 or 1250. 1600 MHz is too high. buying new fan to cooling my vga and now the temperature stable at 71 and i can mning prefectly without any problem
|
|
|
|
sp_
Legendary
Offline
Activity: 2898
Merit: 1087
Team Black developer
|
|
June 16, 2016, 09:35:09 AM |
|
The gtx 960 can go from 11 to 14MHASH with small modifications to the code. I run with --farm-recheck 150
|
|
|
|
moaddip
Newbie
Offline
Activity: 29
Merit: 0
|
|
June 16, 2016, 09:36:02 AM |
|
for 970 nvidia still better to use win7 , 347.52 driver with ethminer 1,1,4 -G option ?
|
|
|
|
sp_
Legendary
Offline
Activity: 2898
Merit: 1087
Team Black developer
|
|
June 16, 2016, 09:57:59 AM |
|
On 970 rigs yes. On 960 rigs no.
|
|
|
|
Amph
Legendary
Offline
Activity: 3206
Merit: 1069
|
|
June 16, 2016, 10:19:25 AM |
|
for 970 nvidia still better to use win7 , 347.52 driver with ethminer 1,1,4 -G option ?
are those the best driver? i changed to the last one, it may explain why i experienced a little loss in the hash, but not sure
|
|
|
|
Genoil (OP)
|
|
June 16, 2016, 10:27:35 AM |
|
The gtx 960 can go from 11 to 14MHASH with small modifications to the code. I run with --farm-recheck 150
Is that 960-only? Interesting, nice rate for a 128-bit bus. Haven't got one, so can't even start guessing.
|
ETH: 0xeb9310b185455f863f526dab3d245809f6854b4d BTC: 1Nu2fMCEBjmnLzqb8qUJpKgq5RoEWFhNcW
|
|
|
Genoil (OP)
|
|
June 16, 2016, 10:31:04 AM |
|
for 970 nvidia still better to use win7 , 347.52 driver with ethminer 1,1,4 -G option ?
For Nvidia use -U, not -G. When using win7, you can also use latest drivers. 347.52 (or 350.12, better) is only for gtx960,970,980 (950 too with some additonal tweaks) on Win10. This also requires a CUDA 6.5 build, which you'll need to do on your own.
|
ETH: 0xeb9310b185455f863f526dab3d245809f6854b4d BTC: 1Nu2fMCEBjmnLzqb8qUJpKgq5RoEWFhNcW
|
|
|
|