dingdongtobias
Newbie
Offline
Activity: 156
Merit: 0
|
|
November 08, 2018, 10:16:56 AM |
|
when new version? monero mining slow on vega 56/64 with examples you give to me
|
|
|
|
claypaky
Jr. Member
Offline
Activity: 154
Merit: 1
|
|
November 08, 2018, 12:36:24 PM |
|
new version (033b5) very unstable for me
032q better than 033b5... for now
|
|
|
|
dnkn
Newbie
Offline
Activity: 11
Merit: 0
|
|
November 08, 2018, 02:31:29 PM |
|
You should try 0.33b4, it has some heavy optimization and it's stable for me. 0.33b5 should be avoided for now. It has increased fee but has very unstable heavy hashrate (especially the effective one). I'm a big fan of this miner and been using it for several months and I'm sure the creator will fix the problems but for now I would recommend to use previous versions or other miners. As for V8, I couldn't get close to teamred results, so I don't quite understand the new fee with explanation like "I only have teamred as a competitor, thus such fee". Also it looks like it would be fair to take that fee accordingly to the algo mined. For example jce is not better then others in cn-lite,xtl... So I would mine with other miners to save extra 1.5% of hashrate but I would definitely use jce for heavy variants with it's brilliant implementation (when it's fixed for stability).
|
|
|
|
io8621
Member
Offline
Activity: 149
Merit: 11
|
|
November 08, 2018, 02:39:04 PM |
|
Ok, after some test:
- old card like 370 2gb work fine, best hashrate compared to other miner, constant > to 510 hs on my 370 2gb
- rx470 and similar with 4 gb ok, better than srbminer my sapphire 470 reference working at 980 without problem.
- rx580 with 8 gb HYNIX after some day of test make new bios with this strap
999000000000000022559D0052626C48A0551214BC0D460B0048C4007D0714204A8900A00200712 419123138B42D3D17
and can get good speed, tested with multihash from 1136 to 1888 with 16 increase from test to test.
But is no stable, get from 1130 hs max to 1030 to 930, average 970 hs.
With srbminer can get 1010 stable with 1250 (950mv) 2200 (900mv).
My card is sapphire pulse 580 8gb hynix, only hynix chip not model with samsung + hynix in the same board.
Look forward at a new version to make new test, congratulations to the dev.
Bye.
|
|
|
|
Lermite
Newbie
Offline
Activity: 54
Merit: 0
|
|
November 08, 2018, 02:46:29 PM |
|
The hashrate instability issue doesn't affect only CN V8 and Heavy.
It also concerns CN Fast (MSR) with the 0.33b4. One of my RX 570 hashratr went from ~1200 to ~900 h/s.
That made me to switch to the 0.33b3 which seems to work fine so far but I yet have to compare its hashrates with SRB's.
|
|
|
|
UnclWish
|
|
November 08, 2018, 02:50:52 PM |
|
You should try 0.33b4, it has some heavy optimization and it's stable for me. 0.33b5 should be avoided for now. It has increased fee but has very unstable heavy hashrate (especially the effective one). I'm a big fan of this miner and been using it for several months and I'm sure the creator will fix the problems but for now I would recommend to use previous versions or other miners. As for V8, I couldn't get close to teamred results, so I don't quite understand the new fee with explanation like "I only have teamred as a competitor, thus such fee". Also it looks like it would be fair to take that fee accordingly to the algo mined. For example jce is not better then others in cn-lite,xtl... So I would mine with other miners to save extra 1.5% of hashrate but I would definitely use jce for heavy variants with it's brilliant implementation (when it's fixed for stability).
Maybe you forgot that the 2+% fee is only for heavy algo and heavy based algos. For v8 and all other algos based on v7 or v8 fee is still 0.9%.
|
|
|
|
Lermite
Newbie
Offline
Activity: 54
Merit: 0
|
|
November 08, 2018, 02:55:45 PM |
|
Maybe you forgot that the 2+% fee is only for heavy algo and heavy based algos. For v8 and all other algos based on v7 or v8 fee is still 0.9%. The 2.5% fee is applied to CN Saber too.
|
|
|
|
dnkn
Newbie
Offline
Activity: 11
Merit: 0
|
|
November 08, 2018, 03:05:36 PM |
|
You should try 0.33b4, it has some heavy optimization and it's stable for me. 0.33b5 should be avoided for now. It has increased fee but has very unstable heavy hashrate (especially the effective one). I'm a big fan of this miner and been using it for several months and I'm sure the creator will fix the problems but for now I would recommend to use previous versions or other miners. As for V8, I couldn't get close to teamred results, so I don't quite understand the new fee with explanation like "I only have teamred as a competitor, thus such fee". Also it looks like it would be fair to take that fee accordingly to the algo mined. For example jce is not better then others in cn-lite,xtl... So I would mine with other miners to save extra 1.5% of hashrate but I would definitely use jce for heavy variants with it's brilliant implementation (when it's fixed for stability).
Maybe you forgot that the 2+% fee is only for heavy algo and heavy based algos. For v8 and all other algos based on v7 or v8 fee is still 0.9%. You're right, it's increased only for heavy variants.
|
|
|
|
1033ruben
Newbie
Offline
Activity: 22
Merit: 0
|
|
November 08, 2018, 04:08:54 PM |
|
1033ruben exactly with JCE miner it is complitly simple: open the miner folder - find "start", click right mouse on it .edit. then just change: 1. pool address 2. port number 3. ssl or no ssl 4. wallet adress.payment id 5.chosse algo : in the end of the list write number 6. last line is the .bat ( if you want change some paramaters, please read the help, does mean start the help.bat so, for example my simple bat: jce_cn_gpu_miner64.exe --no-cpu -g 1,2,3,4 --bus-order --doublecheck --watchdog 10 --retrydelay 10 --keepalive --auto --any --forever --variation %FORK% --low -o %POOL%:%PORT% -u %WALLET% -p %PASSWORD% %SSL% %* THANKS FOR YOUR REPLY!!! I guess my question is how do I optimize it for the the cpus I recently mentioned the 8700K, RYYZEN5 1600 AND THE 6800K, and the 2 amd rigs one is 6 rx 560 and the other is 5 rx 570 THANKS IN ADVANCE RUBEN
|
|
|
|
Ultrasonik
Jr. Member
Offline
Activity: 103
Merit: 2
|
|
November 08, 2018, 05:32:34 PM |
|
what am I doing wrong? work only CPU but GPU error: Compilation of OpenCL kernels failed. Error: CL_BUILD_PROGRAM_FAILURE Code: O-2.10 (other miners work amd rx470 , recommended driver, compute mode) my BAT set FORK=13 jce_cn_gpu_miner64.exe --no-cpu --doublecheck --watchdog 10 --retrydelay 10 --keepalive --auto --any --forever --variation %FORK% --low -o %POOL%:%PORT% -u %WALLET% -p %PASSWORD% %SSL% %*
|
|
|
|
UnclWish
|
|
November 08, 2018, 06:03:43 PM |
|
what am I doing wrong? work only CPU but GPU error: Compilation of OpenCL kernels failed. Error: CL_BUILD_PROGRAM_FAILURE Code: O-2.10 (other miners work amd rx470 , recommended driver, compute mode) my BAT set FORK=13 jce_cn_gpu_miner64.exe --no-cpu --doublecheck --watchdog 10 --retrydelay 10 --keepalive --auto --any --forever --variation %FORK% --low -o %POOL%:%PORT% -u %WALLET% -p %PASSWORD% %SSL% %* Windows 7? It's not supported. GPU miner works only on Windows 8+.
|
|
|
|
BS0D
Newbie
Offline
Activity: 29
Merit: 0
|
|
November 08, 2018, 06:34:30 PM |
|
Didn't find any ready to use jce-monitor and tried to make it myself. Being not much strong in js I've stuck with json syntax error. Miner seems to send unvalid json with syntax like "hashrate": { instead of "hashrate": { and "gpu_status": [ { instead of "gpu_status": [{ Will try to rewrite it later. Maybe I've missed smth. Or I'll find the way to make json valid to parse strings for html normally.
|
|
|
|
Lermite
Newbie
Offline
Activity: 54
Merit: 0
|
|
November 08, 2018, 06:50:44 PM |
|
Are you sure you didn't mess with the GPU indexes in the config file?
|
|
|
|
UnclWish
|
|
November 08, 2018, 06:51:15 PM |
|
Compilation of OpenCL kernels failed. Error: CL_BUILD_PROGRAM_FAILURE Code: O-2.10 (other miners work amd rx470 , recommended driver, compute mode)
Windows 7? It's not supported. GPU miner works only on Windows 8+. of course i have Win 10 ! (SRB miner works without problems) [/quote] Did you have integrated Intel GPU? Check your devices indexes.
|
|
|
|
UnclWish
|
|
November 08, 2018, 07:05:25 PM |
|
my log
For Windows 64-bits Detecting OpenCL-capable GPUs... Found GPU 0, with: Vendor: AMD Processor: Ellesmere Device: 01:00 Compute-Units: 32 Cache Memory: 16 KB Local Memory: 32 KB Global Memory: 8192 MB Addressing: 64-bits
Preparing 2 Mining Threads...
+-- Thread 0 config ------------------------+ | Run on GPU: 0 | | Multi-hash: 864 | | Worksize: 8 | | Factor Alpha 64 | | Factor Beta 16 | +-------------------------------------------+
+-- Thread 1 config ------------------------+ | Run on GPU: 0 | | Multi-hash: 864 | | Worksize: 8 | | Factor Alpha 64 | | Factor Beta 16 | +-------------------------------------------+
Cryptonight Variation: Cryptonight-BitTube v2 of July-2018
Low intensity.
Starting GPU Thread 0, on GPU 0 Created OpenCL Context for GPU 0 at 000002a5cb731410 Created OpenCL Thread 0 Command-Queue for GPU 0 at 000002a5cb7cc200 Scratchpad Allocation success for OpenCL Thread 0 Allocating big 3456MB scratchpad for OpenCL Thread 0... Compiling kernels of OpenCL Thread 0... Compilation of OpenCL kernels failed. Error: CL_BUILD_PROGRAM_FAILURE Code: O-2.10
Starting GPU Thread 1, on GPU 0 Created OpenCL Thread 1 Command-Queue for GPU 0 at 000002a5d14944a0 Scratchpad Allocation success for OpenCL Thread 1 Allocating big 3456MB scratchpad for OpenCL Thread 1... Compiling kernels of OpenCL Thread 1... Compilation of OpenCL kernels failed. Error: CL_BUILD_PROGRAM_FAILURE Code: O-2.10 GPU Watchdog minimum hashrate: 10 Keep-Alive enabled Devfee for GPU is 2.3%
21:55:34 | OpenCL Thread 1 failed, Stop. 21:55:34 | BitTube (TUBE) Mining session starts!
21:55:34 | GPU Compute allocation starts at 80% and reaches 100% after ~5min, 21:55:34 | during this time, the hashrate may be unstable and inconsistent. 21:55:34 | Let the miner warm-up if you're tuning for performance. 21:55:34 | OpenCL Thread 0 failed, Stop. 21:55:34 | Unloaded OpenCL kernels of GPU Thread 1 21:55:34 | Unloaded OpenCL kernels of GPU Thread 0 21:55:34 | Released OpenCL Thread 1 Scratchpad at 000002a5d1a360b0 21:55:34 | Released OpenCL Thread 0 Scratchpad at 000002a5d198ddd0 21:55:34 | Released OpenCL Thread 1 Command-Queue of GPU 0 at 000002a5d14944a0 21:55:34 | Released OpenCL Thread 0 Command-Queue of GPU 0 at 000002a5cb7cc200 21:55:34 | Released OpenCL Context 000002a5cb731410 of GPU 0 21:55:35 | Connecting to mining pool bittube.miner.rocks:5555 ... 21:55:35 | Pool connection socket closed. 21:55:35 | Mining thread 0 stopped. 21:55:35 | Mining thread 1 stopped.
integrated GPU not avaiable
What about pagefile size? Did you have his size as amount of cards multiplied on 8Gb?
|
|
|
|
JCE-Miner (OP)
Member
Offline
Activity: 350
Merit: 22
|
|
November 08, 2018, 07:11:46 PM |
|
Hi all, JSON: is it really forbidden by the json standard to write rather than ? Otherwise, JCE provides a xrm-stak compatible mode, but with less info. Fees of 2.5% on CN-v8 are those of TeamRed, JCE fees on CN-v8 (and v7, stellite, all non-heavy...) are 0.9%But so far TeamRed is faster on the recent cards, i'm second, fees deduced. Fees included (mine are 1.6% lower) it should be a tie. I'm the fastest on older cards like HD7000 however. Compilation of OpenCL kernels failed. Error: CL_BUILD_PROGRAM_FAILURE Code: O-2.10 I finally reproduced that error thanks to your log. This is a critical (but not dangerous as it fails at the beginning) bug. I hope i can fix it quick, otherwise i'll release the b6 with the bug and fix after.
|
|
|
|
Ultrasonik
Jr. Member
Offline
Activity: 103
Merit: 2
|
|
November 08, 2018, 07:19:11 PM |
|
Compilation of OpenCL kernels failed. Error: CL_BUILD_PROGRAM_FAILURE Code: O-2.10 I finally reproduced that error thanks to your log. This is a critical (but not dangerous as it fails at the beginning) bug. I hope i can fix it quick, otherwise i'll release the b6 with the bug and fix after. thanks! if I can help, I will provide the necessary information!
|
|
|
|
JCE-Miner (OP)
Member
Offline
Activity: 350
Merit: 22
|
|
November 08, 2018, 07:22:12 PM |
|
Compilation of OpenCL kernels failed. Error: CL_BUILD_PROGRAM_FAILURE Code: O-2.10 I finally reproduced that error thanks to your log. This is a critical (but not dangerous as it fails at the beginning) bug. I hope i can fix it quick, otherwise i'll release the b6 with the bug and fix after. thanks! if I can help, I will provide the necessary information! Yes, can you give me the full path of the folder where you start the .bat from (the current dir)? i feel like non-standard pathes disturb the OpenCL init yeah i got it, using some crazy exotic chars in the path. It kills the OpenCL init. JCE, as a security, looks for any custom OpenCL dlls in its path, because this is a simple way to hack the OpenCL kernels. But when the dir is too exotic, so it just fails.
|
|
|
|
dingdongtobias
Newbie
Offline
Activity: 156
Merit: 0
|
|
November 08, 2018, 07:44:34 PM |
|
Hi all, JSON: is it really forbidden by the json standard to write rather than ? Otherwise, JCE provides a xrm-stak compatible mode, but with less info. Fees of 2.5% on CN-v8 are those of TeamRed, JCE fees on CN-v8 (and v7, stellite, all non-heavy...) are 0.9%But so far TeamRed is faster on the recent cards, i'm second, fees deduced. Fees included (mine are 1.6% lower) it should be a tie. I'm the fastest on older cards like HD7000 however. Compilation of OpenCL kernels failed. Error: CL_BUILD_PROGRAM_FAILURE Code: O-2.10 I finally reproduced that error thanks to your log. This is a critical (but not dangerous as it fails at the beginning) bug. I hope i can fix it quick, otherwise i'll release the b6 with the bug and fix after. where you get that you second ? my vegas slwer then on SRB
|
|
|
|
JCE-Miner (OP)
Member
Offline
Activity: 350
Merit: 22
|
|
November 08, 2018, 07:51:38 PM |
|
Well configured, i should provide performances between SRB and TeamRed on CN-v8. Look at the Github doc, at the end, there are some examples of good configurations. Overall the difference is tight between the three, so if one runs better on your rig, no problem. I'm talking about version b3 and later, on previous, my perf was pretty bad. Even about v7, some said i was faster, some slower, it probably depends on the precise configuration, drivers and cards settings. And a was talking about average recent cards, including RX550, 560..., not just your precise Vega. OpenCL bug: please confirm your path contained some non-English characters. I tried with some accented letters and it fails like you. Stupid i didn't test such a common case
|
|
|
|
|