Pages:
Author

Topic: Gateless Gate Sharp 1.3.8: 30Mh/s (Ethash) on RX 480! - page 84. (Read 214410 times)

sr. member
Activity: 1484
Merit: 253
I have been playing with phymem.sys for the last few days and confirmed that on-the-fly memory timing mods are possible with GGS, that is, I was able to manipulate parameters for memory timings such as ACTRD/ACTWR, TRRD, TRC, FAW/T32AW etc.  from a user-space.

Now, to make this feature useful, I need to write a BIOS parser, probably in C#, in order to access memory timing straps.
We will see how that goes.
It's very-very GOOD! Because rewriting bios and rebooting every changing timing to check - is a real hell!
just do it 1 time proper Cheesy
tried a little bit oc today, stable but to high powerdraw


I do proper, but every card have different quality and top limit of overclock/undervoltage.
full member
Activity: 364
Merit: 106
ONe Social Network.
I have been playing with phymem.sys for the last few days and confirmed that on-the-fly memory timing mods are possible with GGS, that is, I was able to manipulate parameters for memory timings such as ACTRD/ACTWR, TRRD, TRC, FAW/T32AW etc.  from a user-space.

Now, to make this feature useful, I need to write a BIOS parser, probably in C#, in order to access memory timing straps.
We will see how that goes.
It's very-very GOOD! Because rewriting bios and rebooting every changing timing to check - is a real hell!
just do it 1 time proper Cheesy
tried a little bit oc today, stable but to high powerdraw

sr. member
Activity: 1484
Merit: 253
I have been playing with phymem.sys for the last few days and confirmed that on-the-fly memory timing mods are possible with GGS, that is, I was able to manipulate parameters for memory timings such as ACTRD/ACTWR, TRRD, TRC, FAW/T32AW etc.  from a user-space.

Now, to make this feature useful, I need to write a BIOS parser, probably in C#, in order to access memory timing straps.
We will see how that goes.
It's very-very GOOD! Because rewriting bios and rebooting every changing timing to check - is a real hell!
full member
Activity: 364
Merit: 106
ONe Social Network.
I have been playing with phymem.sys for the last few days and confirmed that on-the-fly memory timing mods are possible with GGS, that is, I was able to manipulate parameters for memory timings such as ACTRD/ACTWR, TRRD, TRC, FAW/T32AW etc.  from a user-space.

Now, to make this feature useful, I need to write a BIOS parser, probably in C#, in order to access memory timing straps.
We will see how that goes.

idk if its worth it, it will just support lazy scumbags who are to lazy or retarded to do a bios flash

whats about this intensity thing for 2 different threads?^^
sr. member
Activity: 728
Merit: 304
Miner Developer
I have been playing with phymem.sys for the last few days and confirmed that on-the-fly memory timing mods are possible with GGS, that is, I was able to manipulate parameters for memory timings such as ACTRD/ACTWR, TRRD, TRC, FAW/T32AW etc.  from a user-space.

Now, to make this feature useful, I need to write a BIOS parser, probably in C#, in order to access memory timing straps.
We will see how that goes.
sr. member
Activity: 728
Merit: 304
Miner Developer
Having some issues with neoscrypt.

Running the 1.2.0 alpha on Win 10 with 6 AMD RX 570s with a custom pool.

Loads up fine, runs for <5 mins and exits.  The log file just shows a restart, no details given. The pool is seeing the miner (when it's up).

I trimmed the log down. Right after Share #24 accepted, it restarted. No other info given

2018-01-13 13:43:31.3814 [12] Share #24 accepted.
2018-01-13 13:43:41.9365 [1] Gateless Gate Sharp 1.2.0 alpha started.
2018-01-13 13:43:44.9796 [1] Number of Devices: 6
2018-01-13 13:43:45.6163 [1] Successfully initialized AMD Display Library.
2018-01-13 13:43:45.6163 [1] Number of ADL Adapters: 99
2018-01-13 13:43:47.3719 [1] Failed to initialize NVIDIA Management Library.
2018-01-13 13:43:53.1944 [1] Launching miner(s) for Custom Pool 0...
2018-01-13 13:43:53.1944 [1] Connecting to eu1.altminer.net:10001 as mywalletIDgoeshere...
2018-01-13 13:43:53.6496 [5] Unknown stratum method: {"id":null,"method":"client.show_message","params":["welcome on ALTMINER.NET - happy mining!"]}
2018-01-13 13:43:53.8421 [6] Miner thread for Device #0 started.
2018-01-13 13:43:53.8441 [6] Loaded Kernels\neoscrypt.cl for Device #0.
2018-01-13 13:43:53.9534 [5] Difficulty set to 64.
2018-01-13 13:44:05.4360 [6] Built NeoScrypt program for Device #0.
2018-01-13 13:44:05.4360 [6] Build options: -O5 -legacy -IKernels -DWORKSIZE=256
2018-01-13 13:44:05.7718 [7] Miner thread for Device #1 started.
2018-01-13 13:44:05.7769 [7] Loaded Kernels\neoscrypt.cl for Device #1.
2018-01-13 13:44:10.8544 [5] Received new job: 10
2018-01-13 13:44:14.8650 [6] Device #0 submitted a share.
2018-01-13 13:44:14.9773 [5] Share #4 accepted.
2018-01-13 13:44:15.3904 [6] Device #0 submitted a share.
2018-01-13 13:44:15.5007 [5] Share #5 accepted.
2018-01-13 13:44:15.7785 [6] Device #0 submitted a share.
2018-01-13 13:44:15.9028 [5] Share #6 accepted.
2018-01-13 13:44:17.1611 [7] Built NeoScrypt program for Device #1.
2018-01-13 13:44:17.1611 [7] Build options: -O5 -legacy -IKernels -DWORKSIZE=256
2018-01-13 13:44:17.4850 [8] Miner thread for Device #2 started.
2018-01-13 13:44:17.4880 [8] Loaded Kernels\neoscrypt.cl for Device #2.
2018-01-13 13:44:18.2209 [7] Device #1 submitted a share.
2018-01-13 13:44:18.3413 [5] Share #7 accepted.
2018-01-13 13:44:20.9743 [6] Device #0: 639.60 Kh/s (NeoScrypt)
2018-01-13 13:44:24.1327 [6] Device #0 submitted a share.
2018-01-13 13:44:24.2450 [5] Share #8 accepted.
2018-01-13 13:44:27.1798 [7] Device #1: 646.13 Kh/s (NeoScrypt)
2018-01-13 13:44:27.9318 [6] Device #0 submitted a share.
2018-01-13 13:44:28.0421 [5] Share #9 accepted.
2018-01-13 13:44:28.8542 [8] Built NeoScrypt program for Device #2.
2018-01-13 13:44:28.8542 [8] Build options: -O5 -legacy -IKernels -DWORKSIZE=256
2018-01-13 13:44:29.1871 [9] Miner thread for Device #3 started.
2018-01-13 13:44:29.1901 [9] Loaded Kernels\neoscrypt.cl for Device #3.


Any ideas?

I have the same issue and my Computer always reboot at this moment. I found out that this always happens when the temperature of a GPU reaches 75°. changing temperature settings on the GPU's doesn't effect this...

edit: When I mine with other algos I don't have this issue. Just with Neoscrypt...

Optimizations in my NeoScrypt kernel is very peculiar in that they make extensive use of conditionals in the order of hundreds IIRC.
So, I would imagine the kernel is very taxing on a tiny part of GPU's execution unit and therefore less tolerant of overclocking/BIOS mods/potentially faulty hardware than other kernels in GGS.
newbie
Activity: 6
Merit: 0
Having some issues with neoscrypt.

Running the 1.2.0 alpha on Win 10 with 6 AMD RX 570s with a custom pool.

Loads up fine, runs for <5 mins and exits.  The log file just shows a restart, no details given. The pool is seeing the miner (when it's up).

I trimmed the log down. Right after Share #24 accepted, it restarted. No other info given

2018-01-13 13:43:31.3814 [12] Share #24 accepted.
2018-01-13 13:43:41.9365 [1] Gateless Gate Sharp 1.2.0 alpha started.
2018-01-13 13:43:44.9796 [1] Number of Devices: 6
2018-01-13 13:43:45.6163 [1] Successfully initialized AMD Display Library.
2018-01-13 13:43:45.6163 [1] Number of ADL Adapters: 99
2018-01-13 13:43:47.3719 [1] Failed to initialize NVIDIA Management Library.
2018-01-13 13:43:53.1944 [1] Launching miner(s) for Custom Pool 0...
2018-01-13 13:43:53.1944 [1] Connecting to eu1.altminer.net:10001 as mywalletIDgoeshere...
2018-01-13 13:43:53.6496 [5] Unknown stratum method: {"id":null,"method":"client.show_message","params":["welcome on ALTMINER.NET - happy mining!"]}
2018-01-13 13:43:53.8421 [6] Miner thread for Device #0 started.
2018-01-13 13:43:53.8441 [6] Loaded Kernels\neoscrypt.cl for Device #0.
2018-01-13 13:43:53.9534 [5] Difficulty set to 64.
2018-01-13 13:44:05.4360 [6] Built NeoScrypt program for Device #0.
2018-01-13 13:44:05.4360 [6] Build options: -O5 -legacy -IKernels -DWORKSIZE=256
2018-01-13 13:44:05.7718 [7] Miner thread for Device #1 started.
2018-01-13 13:44:05.7769 [7] Loaded Kernels\neoscrypt.cl for Device #1.
2018-01-13 13:44:10.8544 [5] Received new job: 10
2018-01-13 13:44:14.8650 [6] Device #0 submitted a share.
2018-01-13 13:44:14.9773 [5] Share #4 accepted.
2018-01-13 13:44:15.3904 [6] Device #0 submitted a share.
2018-01-13 13:44:15.5007 [5] Share #5 accepted.
2018-01-13 13:44:15.7785 [6] Device #0 submitted a share.
2018-01-13 13:44:15.9028 [5] Share #6 accepted.
2018-01-13 13:44:17.1611 [7] Built NeoScrypt program for Device #1.
2018-01-13 13:44:17.1611 [7] Build options: -O5 -legacy -IKernels -DWORKSIZE=256
2018-01-13 13:44:17.4850 [8] Miner thread for Device #2 started.
2018-01-13 13:44:17.4880 [8] Loaded Kernels\neoscrypt.cl for Device #2.
2018-01-13 13:44:18.2209 [7] Device #1 submitted a share.
2018-01-13 13:44:18.3413 [5] Share #7 accepted.
2018-01-13 13:44:20.9743 [6] Device #0: 639.60 Kh/s (NeoScrypt)
2018-01-13 13:44:24.1327 [6] Device #0 submitted a share.
2018-01-13 13:44:24.2450 [5] Share #8 accepted.
2018-01-13 13:44:27.1798 [7] Device #1: 646.13 Kh/s (NeoScrypt)
2018-01-13 13:44:27.9318 [6] Device #0 submitted a share.
2018-01-13 13:44:28.0421 [5] Share #9 accepted.
2018-01-13 13:44:28.8542 [8] Built NeoScrypt program for Device #2.
2018-01-13 13:44:28.8542 [8] Build options: -O5 -legacy -IKernels -DWORKSIZE=256
2018-01-13 13:44:29.1871 [9] Miner thread for Device #3 started.
2018-01-13 13:44:29.1901 [9] Loaded Kernels\neoscrypt.cl for Device #3.


Any ideas?

I have the same issue and my Computer always reboot at this moment. I found out that this always happens when the temperature of a GPU reaches 75°. changing temperature settings on the GPU's doesn't effect this...

edit: When I mine with other algos I don't have this issue. Just with Neoscrypt...
newbie
Activity: 8
Merit: 0
Hi,any1 have similar problems with  ASUS STRIX RX 570 4GB, I cant go over 830h/s at cryptonight (electroneum) mining,ent-eu1.nanopool.org changed raw to 896 and moded bios and pushed in after burner to do max poslible settup(memory clock at 2020 and so on...)

https://prnt.sc/i0pke0
https://prnt.sc/i0pshf
sr. member
Activity: 1484
Merit: 253
RX570 Nitro+ 4Gb - Elpida/Hynix, RX570 Aorus 4Gb - Hynix, RX580 Armor 4Gb - Hynix, RX580 Strix 8Gb - Samsung. Used this tool for 1-click timings mod (1500 and up) - https://github.com/jaschaknack/PolarisBiosEditor
All cards 4Gb?
newbie
Activity: 24
Merit: 0
RX570 Nitro+ 4Gb - Elpida/Hynix, RX570 Aorus 4Gb - Hynix, RX580 Armor 4Gb - Hynix, RX580 Strix 8Gb - Samsung. Used this tool for 1-click timings mod (1500 and up) - https://github.com/jaschaknack/PolarisBiosEditor
sr. member
Activity: 1484
Merit: 253
What memory type your cards have?
newbie
Activity: 24
Merit: 0
UnclWish
Yes, I did. The thing is, I always check HWInfo64 for memory errors and for some cards it's 1875Mhz for others 2000Mhz stable with no errors. But none can do over 2000Mhz memory and I couldn't make any card faster than 30Mhs ETH, sadly. RX570 Nitro+, Aorus and RX580 Strix, Armor can only manage 27.7-29.5Mhs. I have a 13GPU modded RX570 Nitro+ rig with 1 PSU and 2 MBs, got 1650W from the wall of power usage mining only ETH, 2050W in DUAL and 1520W in Neoscrypt.

Anyways, at GG test Nitro 570 run at 26.7 where in Claymore it's 27.7
sr. member
Activity: 1484
Merit: 253
zawawa
Thanks for the hard work) Tried you miner today for the first time - neoscrypt AMD and so far so good! The hashrate is much better than nsgminer 750Khs vs 540Khs on my modded RX570|580 cards - http://prntscr.com/i08mrt
Also, the values are kinda way off, please put Intensity at 16 and work size at 64 (the same worked for nsgminer). All other values for all my rigs are either unstable or cause performance drops. This will be a good way to start for neoscrypt miners.

It will also be nice to add skunkhash algo in the future as well with default -i 20 Wink

Thanks

Out of curiosity, what are your mod settings?

Used Polaris bios mod master for AUTO tuning with timings. 1200 core clock with 1875-2000 memory (depending on card), 0.9v already flashed in the bios. Neoscrypt uses around 5-7% less power than ETH. Also, mining ETH with GatelessGate I loose around 1Mhs of speed, probably need some more tuning.
Did you tried to set memory clock 2100+?
newbie
Activity: 24
Merit: 0
zawawa
Thanks for the hard work) Tried you miner today for the first time - neoscrypt AMD and so far so good! The hashrate is much better than nsgminer 750Khs vs 540Khs on my modded RX570|580 cards - http://prntscr.com/i08mrt
Also, the values are kinda way off, please put Intensity at 16 and work size at 64 (the same worked for nsgminer). All other values for all my rigs are either unstable or cause performance drops. This will be a good way to start for neoscrypt miners.

It will also be nice to add skunkhash algo in the future as well with default -i 20 Wink

Thanks

Out of curiosity, what are your mod settings?

Used Polaris bios mod master for AUTO tuning with timings. 1200 core clock with 1875-2000 memory (depending on card), 0.9v already flashed in the bios. Neoscrypt uses around 5-7% less power than ETH. Also, mining ETH with GatelessGate I loose around 1Mhs of speed, probably need some more tuning.
newbie
Activity: 14
Merit: 0
zawawa
Thanks for the hard work) Tried you miner today for the first time - neoscrypt AMD and so far so good! The hashrate is much better than nsgminer 750Khs vs 540Khs on my modded RX570|580 cards - http://prntscr.com/i08mrt
Also, the values are kinda way off, please put Intensity at 16 and work size at 64 (the same worked for nsgminer). All other values for all my rigs are either unstable or cause performance drops. This will be a good way to start for neoscrypt miners.

It will also be nice to add skunkhash algo in the future as well with default -i 20 Wink



Thanks

Out of curiosity, what are your mod settings?
member
Activity: 170
Merit: 10
just remove the hooks in dashboard

the dashboard won't load, it crashes and exits on startup with that error.
full member
Activity: 364
Merit: 106
ONe Social Network.
just remove the hooks in dashboard
member
Activity: 170
Merit: 10
how do i disable nvidia or just do amd-only from cmd line or from config?  I dont see any documentation on cmd line parameters or cfg file format.  I have a multi gpu box (amd+nvidia) thats giving me a 'failed to initialize opencl devices'.  this system has integrated graphics (intel), amd cards, and nvidia cards, and I just want to run gatelessgate on the amd cards...
jr. member
Activity: 98
Merit: 2
Any way to add pci line number to identify the card? or device number?
Pages:
Jump to: