OK, let's make this coming week the "Optimizations and New Algorithms Week!"
and please dont forget about vega(newest driver speed like blockchain driver speeD)
Topic: Gateless Gate Sharp 1.3.8: 30Mh/s (Ethash) on RX 480! - page 62. (Read 214431 times)
and please dont forget about vega(newest driver speed like blockchain driver speeD)
OK, let's make this coming week the "Optimizations and New Algorithms Week!"
Any estimate when Equihash support will be added?
It's twice more profitable on some cards.
It's twice more profitable on some cards.
Why eth speed is so low? Claymore gives 29-30M speed. GGS on default settings gives 27-28M... RX 580 8Gb
If you are using a modded BIOS for Claymore's, that's inevitable.
Also, I haven't done any optimizations for Ethash yet.
Maybe that's a good place to start with v1.3...
Why eth speed is so low? Claymore gives 29-30M speed. GGS on default settings gives 27-28M... RX 580 8Gb
4h runtime now,everything good so far. system is not laging anymore after longer runtime. i like it
Gateless Gate Sharp 1.2.17 beta
https://github.com/zawawawa/GatelessGateSharp/releases/tag/v1.2.17-beta
* Implemented a subset of sgminer's JSON RPC API for remote-monitoring.
* Fixed a number of bugs to increase stability.
I happened to have Awesome Miner and am using it for remote-monitoring, but GGS should work with other monitoring programs compatible with sgminer/cgminer. If everything goes well, I would like to make the first stable version out of this one. Please let me know if anything.
https://github.com/zawawawa/GatelessGateSharp/releases/tag/v1.2.17-beta
* Implemented a subset of sgminer's JSON RPC API for remote-monitoring.
* Fixed a number of bugs to increase stability.
I happened to have Awesome Miner and am using it for remote-monitoring, but GGS should work with other monitoring programs compatible with sgminer/cgminer. If everything goes well, I would like to make the first stable version out of this one. Please let me know if anything.
hi
the boost performance give power limit at 120% in my 570.doesn't touch the core clock neither memory clock.
in my case,no more perf.
the boost performance give power limit at 120% in my 570.doesn't touch the core clock neither memory clock.
in my case,no more perf.
Could I see a screenshot of the entire Dashboard?
hi
before and after
Ah, I see... Let me see what I can do. It may take some time, though.
hi
the boost performance give power limit at 120% in my 570.doesn't touch the core clock neither memory clock.
in my case,no more perf.
the boost performance give power limit at 120% in my 570.doesn't touch the core clock neither memory clock.
in my case,no more perf.
Could I see a screenshot of the entire Dashboard?
hi
before and after
I am actually pretty much done with the next version, ("v1.2.17 beta").
Along with all the bug fixes, I implemented a subset of sgminer's API for remote monitoring.
I am running some extra tests so that I won't blow up this release, but it will be available shortly.
Along with all the bug fixes, I implemented a subset of sgminer's API for remote monitoring.
I am running some extra tests so that I won't blow up this release, but it will be available shortly.
Miner works well on my 1080TI. About as fast as Claymore. But It eats about 30% CPU usage. Not good for power consumption. What's going on with that?
This is due to the infamous busy wait bug in NVIDIA's OpenCL drivers, and I haven't implemented a workaround in GGS yet.
I will get to it after I am done with the next stable version.
hi
the boost performance give power limit at 120% in my 570.doesn't touch the core clock neither memory clock.
in my case,no more perf.
the boost performance give power limit at 120% in my 570.doesn't touch the core clock neither memory clock.
in my case,no more perf.
Could I see a screenshot of the entire Dashboard?
Miner works well on my 1080TI. About as fast as Claymore. But It eats about 30% CPU usage. Not good for power consumption. What's going on with that?
Gateless Gate Sharp 1.2.16 alpha
https://github.com/zawawawa/GatelessGateSharp/releases/tag/v1.2.16-alpha
* Fixed a critical bug in logging.
Sorry, guys! This should fix the problem.
https://github.com/zawawawa/GatelessGateSharp/releases/tag/v1.2.16-alpha
* Fixed a critical bug in logging.
Sorry, guys! This should fix the problem.
miner restarted 2 times in 1 hour, 1.2.3 worked for days without 1 restart. i roll back.
i really dont want to hurt your feelings but i need to say this. Everytime you get to a good and stable version like 1.22, 1.23 you fuck it up in the next version with some stuff or changes nobody needs!
example
first versions after addng graphs where very instable
example
" Added the "Boost Performance" automatic overclocking/undervolting feature for RX 470/480/570/580, R9 Nano, and HD 7990."
why wasting working time with something like this? Real Miners know to config their cards, every card needs its own config to run stable, even if its from the same brand with same memory. this feature will only cause problems because people will try it and it will not work maybe proper on every system. on the other hand it will just support lazy fucktards who are to dumb and/or to lazy to do a simple biosflash, do some little research about settings for gpus.
example
planed on the fly memory mods
same as above for what? if somebody is to dumb or to lazy to do a biosflash, fuck him he deserves stock speed. you aint gonna earn more by supporting lazy fucktards.
i disagree. i dont use this (yet) i like a command line miners and know how to bios mod my amds and tune them for min power or max hash.. but i would like to have something for folks who may get into casual mining such as non technical friends and family. install, point to their wallet, and it does the rest. no bios modding, no afterburner with its profiles etc. and the more folks who are non techies that use this as opposed to not mining at all because its "too complicated," or using a simpler, less powerful miner that does no auto tuning, well the more zawawa will make. makes sense to me to have a very simple to use miner that can squeeze out max performance on its own. well as long as he keeps a command line in there somewhere
sure i want more algos and such but this is a beta and a work in progress. if a new version blows up, note the details post and zawawa will fix it. if its not the feature you want in the order you want it, well... just wait, the way zawawa is cranking out new versions it wont be long before new stuff comes along.
Hi guys!
People that uses overclock feature (we have started to use 2 days ago)... We found a same problem in 2 RIGs:
When using overclock setted equal to all GPUs, one of them, starts using lower memory, like 0.6g instead of 1.2g.
As result, the hash/s is lower than the another GPUs.
Sometimes, closing and reopening resolves the problem, sometimes no.
Disabling the GPU with the problem, another GPU starts with the same problem.
So, we tried to disable just the OC of the problematic GPU and setted the OC by Wattman.
Works fine, the problematic GPU works fast, but one of the GPUs with OC by GGS starts to work with lower memory and hashrate.
We guess be the rawintensity and threads are not be setting for the first (one of them) gpu with OC, using some default setting that uses just 0.6g.
At this time, to use all GPUs in the max hashrate we need to disable OC by GGS in all GPUs.
Should be just a little thing to resolve about the setting application from DB OC Data.
People that uses overclock feature (we have started to use 2 days ago)... We found a same problem in 2 RIGs:
When using overclock setted equal to all GPUs, one of them, starts using lower memory, like 0.6g instead of 1.2g.
As result, the hash/s is lower than the another GPUs.
Sometimes, closing and reopening resolves the problem, sometimes no.
Disabling the GPU with the problem, another GPU starts with the same problem.
So, we tried to disable just the OC of the problematic GPU and setted the OC by Wattman.
Works fine, the problematic GPU works fast, but one of the GPUs with OC by GGS starts to work with lower memory and hashrate.
We guess be the rawintensity and threads are not be setting for the first (one of them) gpu with OC, using some default setting that uses just 0.6g.
At this time, to use all GPUs in the max hashrate we need to disable OC by GGS in all GPUs.
Should be just a little thing to resolve about the setting application from DB OC Data.
hi
the boost performance give power limit at 120% in my 570.doesn't touch the core clock neither memory clock.
in my case,no more perf.
the boost performance give power limit at 120% in my 570.doesn't touch the core clock neither memory clock.
in my case,no more perf.
1-
I was having the overclock of the GPU 0 all times did incorrect by GGS (with a old value), I thought be another program doing this, but not finding anything, I tried to reset all settings with the reset buttons, and setted all data again... Well, so the overclock starts to use correct config to all cards.
I guess have some GUI option, like the OC setting for another algorithm, that writes in the same DB field, causing the error, and I can setted another algorithm by mistake sometime, but when corrected, the OC setting still valid to Neoscrypt.
I was having the overclock of the GPU 0 all times did incorrect by GGS (with a old value), I thought be another program doing this, but not finding anything, I tried to reset all settings with the reset buttons, and setted all data again... Well, so the overclock starts to use correct config to all cards.
I guess have some GUI option, like the OC setting for another algorithm, that writes in the same DB field, causing the error, and I can setted another algorithm by mistake sometime, but when corrected, the OC setting still valid to Neoscrypt.
Thank you so much! This was a really hard bug to find, buried deep inside GUI components.
UnclWish'es Pitcairn woes may be over soon...
.
2 bugs about DB:
1-
I was having the overclock of the GPU 0 all times did incorrect by GGS (with a old value), I thought be another program doing this, but not finding anything, I tried to reset all settings with the reset buttons, and setted all data again... Well, so the overclock starts to use correct config to all cards.
I guess have some GUI option, like the OC setting for another algorithm, that writes in the same DB field, causing the error, and I can setted another algorithm by mistake sometime, but when corrected, the OC setting still valid to Neoscrypt.
2-
All times, when de GGS goes to save the config data, I need to wait about 10 seconds with the program freezed to save. Saving in the roaming data folder or with save as option in documents folder. This problem occours in our 4 rigs in test with GGS, I thought be a same problem than many users here.
Best regards.
1-
I was having the overclock of the GPU 0 all times did incorrect by GGS (with a old value), I thought be another program doing this, but not finding anything, I tried to reset all settings with the reset buttons, and setted all data again... Well, so the overclock starts to use correct config to all cards.
I guess have some GUI option, like the OC setting for another algorithm, that writes in the same DB field, causing the error, and I can setted another algorithm by mistake sometime, but when corrected, the OC setting still valid to Neoscrypt.
2-
All times, when de GGS goes to save the config data, I need to wait about 10 seconds with the program freezed to save. Saving in the roaming data folder or with save as option in documents folder. This problem occours in our 4 rigs in test with GGS, I thought be a same problem than many users here.
Best regards.
zawawa, it's need to describe - what exactly "Boost" button do.
Jump to: