Pages:
Author

Topic: WildRig Multi 0.35.1 beta 2 multi-algo miner for AMD & NVIDIA - page 11. (Read 92264 times)

member
Activity: 724
Merit: 49
WildRig Multi 0.31.2
- improved heavyhash for AMD Polaris(~1.5%), Vega(~10%), RDNA/RDNA2(~3-4%)
- implemented correct message of unsupported iGPU's on newer AMD drivers(21.5.2+)
- implemented parameter --watchdog-script
- fixed monitoring AMD gpu's temperature/power/fan under Linux, and RDNA/RDNA2 on Windows
member
Activity: 724
Merit: 49
Hi

Interesting.

I have Vega 11 iGPU and some Compute 5 + 6 Nvidia GPU's. Miner starts when only chosing Nvidia Pascal and Maxwell (-d) with RTM but all GPU's don't hash. 0 hash but Miner runs with no error.

How to install OpenCL drivers? All newest Drivers installed after DDU.  Kiss
Hi, try --opencl-launch 256
Probably default one value is too high for your gpu's(especially Maxwell)
newbie
Activity: 42
Merit: 0
Hi

Interesting.

I have Vega 11 iGPU and some Compute 5 + 6 Nvidia GPU's. Miner starts when only chosing Nvidia Pascal and Maxwell (-d) with RTM but all GPU's don't hash. 0 hash but Miner runs with no error.

How to install OpenCL drivers? All newest Drivers installed after DDU.  Kiss
newbie
Activity: 8
Merit: 1
hello do you have plan to add neoscrypt algo?
newbie
Activity: 123
Merit: 0
any chance of adding cortex thats another one thats got crap mining software hardly works come on now i love wildrig.
jr. member
Activity: 133
Merit: 5
0.31.0 beta
- added default parameters for CMP 40/50/70/90/170 HX and RTX A4500/5000/6000
- improved heavyhash by ~80%(NVIDIA Pascal/Turing/Ampere) and 20%(AMD RDNA2)
- dev-fee set to 2% on heavyhash for nvidia rigs

0.31.1 beta
- improved heavyhash up to 10% on RDNA/RDNA2 gpu's with less power consumption
- fixed power jumps on NVIDIA gpu's


RX 6900XT ~900MH/s @ ~230W PPT.
RX 6900XT ~920MH/s @ ~240W PPT.

Some nice efficiency increases there!
member
Activity: 724
Merit: 49
0.31.0 beta
- added default parameters for CMP 40/50/70/90/170 HX and RTX A4500/5000/6000
- improved heavyhash by ~80%(NVIDIA Pascal/Turing/Ampere) and 20%(AMD RDNA2)
- dev-fee set to 2% on heavyhash for nvidia rigs

0.31.1 beta
- improved heavyhash up to 10% on RDNA/RDNA2 gpu's with less power consumption
- fixed power jumps on NVIDIA gpu's



Hi,

me again with the Vega64 and heavyhash problem. It works round about 4min and then... (on two different rigs with Vega's)

[20:55:18] new job from obtc.suprnova.cc:4075 diff 10.89G/473.13T
[20:55:18] block: 38,630        job target: 0x0000000064ef83aa
[20:55:35] accepted(38/0)       diff 25.96G     GPU#0   (41 ms)
[20:55:40] Uptime: 0 days 00:03:59
[20:55:40] GPU #0[T:56C F:99%]: 10s: 144.36 60s: 144.09 15m: n/a MH/s
[20:55:40] GPU #1[T:38C F:55%]: 10s: 345.03 60s: 344.84 15m: n/a MH/s
[20:55:40] hashrate: 10s: 489.39 60s: 488.94 15m: n/a MH/s max: 489.57 MH/s
[20:55:43] accepted(39/0)       diff 14.20G     GPU#1   (41 ms)
[20:55:45] new job from obtc.suprnova.cc:4075 diff 10.89G/473.13T
[20:55:45] block: 38,631        job target: 0x0000000064ef83aa
[20:55:45] rejected(39/1)       diff 184.89G    GPU#1   (45 ms)
[20:55:45] reason: job not found
[20:55:51] rejected(39/2)       diff 70.92G     GPU#1   (41 ms)
[20:55:51] reason: low difficulty share of 2.7181934694157165e-10
[20:55:52] rejected(39/3)       diff 16.74G     GPU#1   (42 ms)

i hope it can help you for a future fix, at moment i will use version 0.30.6.

Thanks.



Yeah, already got a report... made another mistake. Happens on new block.



Hi,

me again with the Vega64 and heavyhash problem. It works round about 4min and then... (on two different rigs with Vega's)

[20:55:18] new job from obtc.suprnova.cc:4075 diff 10.89G/473.13T
[20:55:18] block: 38,630        job target: 0x0000000064ef83aa
[20:55:35] accepted(38/0)       diff 25.96G     GPU#0   (41 ms)
[20:55:40] Uptime: 0 days 00:03:59
[20:55:40] GPU #0[T:56C F:99%]: 10s: 144.36 60s: 144.09 15m: n/a MH/s
[20:55:40] GPU #1[T:38C F:55%]: 10s: 345.03 60s: 344.84 15m: n/a MH/s
[20:55:40] hashrate: 10s: 489.39 60s: 488.94 15m: n/a MH/s max: 489.57 MH/s
[20:55:43] accepted(39/0)       diff 14.20G     GPU#1   (41 ms)
[20:55:45] new job from obtc.suprnova.cc:4075 diff 10.89G/473.13T
[20:55:45] block: 38,631        job target: 0x0000000064ef83aa
[20:55:45] rejected(39/1)       diff 184.89G    GPU#1   (45 ms)
[20:55:45] reason: job not found
[20:55:51] rejected(39/2)       diff 70.92G     GPU#1   (41 ms)
[20:55:51] reason: low difficulty share of 2.7181934694157165e-10
[20:55:52] rejected(39/3)       diff 16.74G     GPU#1   (42 ms)

i hope it can help you for a future fix, at moment i will use version 0.30.6.

Thanks.



Uploaded new version with a fix. Unfortunately github is dead, so only mirror links are working.
newbie
Activity: 44
Merit: 0
Hi,

Vega64 works again with 0.30.8.1 (heavyhash), Nice job. You are great.

Thanks.

Hi,

me again with the Vega64 and heavyhash problem. It works round about 4min and then... (on two different rigs with Vega's)

[20:55:18] new job from obtc.suprnova.cc:4075 diff 10.89G/473.13T
[20:55:18] block: 38,630        job target: 0x0000000064ef83aa
[20:55:35] accepted(38/0)       diff 25.96G     GPU#0   (41 ms)
[20:55:40] Uptime: 0 days 00:03:59
[20:55:40] GPU #0[T:56C F:99%]: 10s: 144.36 60s: 144.09 15m: n/a MH/s
[20:55:40] GPU #1[T:38C F:55%]: 10s: 345.03 60s: 344.84 15m: n/a MH/s
[20:55:40] hashrate: 10s: 489.39 60s: 488.94 15m: n/a MH/s max: 489.57 MH/s
[20:55:43] accepted(39/0)       diff 14.20G     GPU#1   (41 ms)
[20:55:45] new job from obtc.suprnova.cc:4075 diff 10.89G/473.13T
[20:55:45] block: 38,631        job target: 0x0000000064ef83aa
[20:55:45] rejected(39/1)       diff 184.89G    GPU#1   (45 ms)
[20:55:45] reason: job not found
[20:55:51] rejected(39/2)       diff 70.92G     GPU#1   (41 ms)
[20:55:51] reason: low difficulty share of 2.7181934694157165e-10
[20:55:52] rejected(39/3)       diff 16.74G     GPU#1   (42 ms)

i hope it can help you for a future fix, at moment i will use version 0.30.6.

Thanks.




member
Activity: 724
Merit: 49
Hi,

Version  0.30.8 doesn't work with my AMD Vega64. A second RX480/8GB is also in the rig. Problem is -> low difficulty and rejecting from server.

I checked it with driver 20.3.x and 21.11.3. (Windows 10).


Difficulty by the Vega is 27 and the RX480 23 (auto setting).


All works fine with 0.30.6.


I hope it no big problem for the next versions.


Thanks.

Updated links to version 0.30.8.1, it should work now.
member
Activity: 724
Merit: 49
Hi,

Version  0.30.8 doesn't work with my AMD Vega64. A second RX480/8GB is also in the rig. Problem is -> low difficulty and rejecting from server.

I checked it with driver 20.3.x and 21.11.3. (Windows 10).


Difficulty by the Vega is 27 and the RX480 23 (auto setting).


All works fine with 0.30.6.


I hope it no big problem for the next versions.


Thanks.

Hi, what algorithm?

Hi,

Version  0.30.8 doesn't work with my AMD Vega64. A second RX480/8GB is also in the rig. Problem is -> low difficulty and rejecting from server.

I checked it with driver 20.3.x and 21.11.3. (Windows 10).


Difficulty by the Vega is 27 and the RX480 23 (auto setting).


All works fine with 0.30.6.


I hope it no big problem for the next versions.


Thanks.

Hi, what algorithm?
Damn, looks like heavyhash. I got a very rare bug with compiler %) In Debug mode it works like a charm, release - shift in 1 byte  Angry Will fix it, crazy days...
newbie
Activity: 44
Merit: 0
Hi,

Version  0.30.8 doesn't work with my AMD Vega64. A second RX480/8GB is also in the rig. Problem is -> low difficulty and rejecting from server.

I checked it with driver 20.3.x and 21.11.3. (Windows 10).


Difficulty by the Vega is 27 and the RX480 23 (auto setting).


All works fine with 0.30.6.


I hope it no big problem for the next versions.


Thanks.
member
Activity: 724
Merit: 49
name one that does it for amd and nvidia then at same time in one program?Huh

Wildrig also not work when u have mixed rig (nvidia+amd). For exmaple, i have AMD rx570 and nvidia gtx1660 in same PC
and wildrig can see only AMD gpu:
Code:
PLATFORMS:
#0: Advanced Micro Devices, Inc.

DEVICES:
GPU #0: Radeon RX 570 Series    (busID: 3, arch: Ellesmere, cu: 32, mem: 4096Mb)
Press any key to continue . . .
Problem is OpenCL drivers for NVIDIA. You need to fix that or use mining OS with properly pre-installed drivers. Not much can be done on my side, except using CUDA, but I don't want to Smiley
newbie
Activity: 315
Merit: 0
name one that does it for amd and nvidia then at same time in one program?Huh

Wildrig also not work when u have mixed rig (nvidia+amd). For exmaple, i have AMD rx570 and nvidia gtx1660 in same PC
and wildrig can see only AMD gpu:
Code:
PLATFORMS:
#0: Advanced Micro Devices, Inc.

DEVICES:
GPU #0: Radeon RX 570 Series    (busID: 3, arch: Ellesmere, cu: 32, mem: 4096Mb)
Press any key to continue . . .
member
Activity: 724
Merit: 49
0.30.8 beta
-  fixed low hashrate on heavyhash for AMD gpu's with old drivers

ps: removed 0.30.7 as a buggy one
member
Activity: 724
Merit: 49
name one that does it for amd and nvidia then at same time in one program?Huh
I think nanominer, but not sure. Anyway, Octopus is highly prefer NVIDIA gpu's over AMD, so almost no sense.
newbie
Activity: 123
Merit: 0
name one that does it for amd and nvidia then at same time in one program?Huh
member
Activity: 724
Merit: 49
any chance you can add octopus for cfx as there aint a good miner that does both amd or nvidia at same time many thanks.
Sorry, I don't do mainstream algorithms in most cases, and there plenty of solutions out there doing their job pretty good for Octopus Smiley
newbie
Activity: 123
Merit: 0
any chance you can add octopus for cfx as there aint a good miner that does both amd or nvidia at same time many thanks.
member
Activity: 724
Merit: 49
0.30.6.1 beta
- IMPORTANT removed auto-detection of ssl to avoid problems with some pools(no mining after dev-fee, no reconnect and so on)
jr. member
Activity: 133
Merit: 5
EDIT:
Potential issue: heavyhash has a tendency to crash the AMD drivers when the Donation Mining starts. (driver timeout)
Resulting in invalid shares after. (need to restart miner)
Hmmm, it might be, will take a look and change some things. Can you try 0.30.6?

RX 6900XT saw a ~5% extra performance on the 0.30.6 build.
RX 6900XT ~650MH/s for 225W PPT.

Driver timeout might have been because I tried to run --mutiple-instance with dual algorithm on the gpu's.
Could be the Vega 64 wasn't liking the temperatures, was running quite high on the Core Hotspot temp. I tried a lower PPT and issue didn't come back running only heavyhash alone.
Pages:
Jump to: