SRBMiner Cryptonight AMD GPU Miner V1.9.3 - native algo switching

[UnclWish]

/* This is just an example, edit it and remove comment lines (the slash and star) !! */
"gpu_conf" :
[
{ "id" : 0, "intensity" : 13, "double_threads" : true},
{ "id" : 1, "intensity" : 13, "double_threads" : true},
{ "id" : 2, "intensity" : 13, "double_threads" : true},
{ "id" : 3, "intensity" : 13, "double_threads" : true},
{ "id" : 4, "intensity" : 13, "double_threads" : true},
{ "id" : 5, "intensity" : 12, "double_threads" : true},
]

/* Make sure not to delete this last blank line */

this is how the end of the config looks like

For whom this post?

[1714299897]

1714299897

[1714299897]

1714299897

[UnclWish]

I guess it only works with GCN or newer cards like almost every other miner created

[2018-03-25 05:08:15] Error CL_INVALID_WORK_GROUP_SIZE when calling clEnqueueNDRangeKernel for kernel 4.

scrolling...

I think, that pre-GCN cards is not suitable for mining at all. Thats why it's needless to say...
Your error may be because of lack of vmemory. Try to lower intensity.

[Bulitt]

HD 6950 gets 200h/sec on monero in claymore 9.7
R9 270X gets 430h/sec on monero in claymore 9.7

shows a hashrate in the console but I doubt the results are valid

[2018-03-25 05:54:29] Pool accepted result 0x0000A903
[2018-03-25 05:54:30] Error CL_INVALID_WORK_GROUP_SIZE when calling clEnqueueNDR
angeKernel for kernel 3.
[2018-03-25 05:54:31] GPU0[T0]:     419.0 H/S
[2018-03-25 05:54:31] GPU1[T1]:     852.0 H/S
[2018-03-25 05:54:31] Total:        1271.0 H/S

it appears that the error is coming from the R9 270X and not the HD 6900... if kernel 3 means BUS_ID 3

[UnclWish]

HD 6950 gets 200h/sec on monero in claymore 9.7
R9 270X gets 430h/sec on monero in claymore 9.7

shows a hashrate in the console but I doubt the results are valid

[2018-03-25 05:54:29] Pool accepted result 0x0000A903
[2018-03-25 05:54:30] Error CL_INVALID_WORK_GROUP_SIZE when calling clEnqueueNDR
angeKernel for kernel 3.
[2018-03-25 05:54:31] GPU0[T0]:     419.0 H/S
[2018-03-25 05:54:31] GPU1[T1]:     852.0 H/S
[2018-03-25 05:54:31] Total:        1271.0 H/S

it appears that the error is coming from the R9 270X and not the HD 6900... if kernel 3 means BUS_ID 3

270X on Claymore 11.2 gives 500+ h/s.

[toptek]

- Low DevFee (0.85%) -> every ~2 hours 1 minute mining for the dev
- Non-agressive DevFee mining -> if miner can't connect to DevFee pool, no problem, switching back to user pool ASAP
 
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Smart move why can't the rest do it this way, more will use it with less complaints .  if anyone complains about this fee an the way it works they are ass holes .

I can see complaining about the 2 % fee it makes the DEV the ass holes imo.....an they  Usually are .......

[doktor83]

/* This is just an example, edit it and remove comment lines (the slash and star) !! */
"gpu_conf" :
[
{ "id" : 0, "intensity" : 13, "double_threads" : true},
{ "id" : 1, "intensity" : 13, "double_threads" : true},
{ "id" : 2, "intensity" : 13, "double_threads" : true},
{ "id" : 3, "intensity" : 13, "double_threads" : true},
{ "id" : 4, "intensity" : 13, "double_threads" : true},
{ "id" : 5, "intensity" : 12, "double_threads" : true},
]

/* Make sure not to delete this last blank line */

this is how the end of the config looks like

Try download this example config and run it :

https://mega.nz/#F!yJxTTZpK!ZA6WVs3kIfcw_slX3dgw1Q

[doktor83]

Thanks! Good job!

What about setting over/underclocking and under/overvoltaging? Maybe you add parameters to set clocks and voltages?

Temperature and fan control didn't work for R9 270X 4Gb.

Miner didn't work if i use gpu_conf section and left only gpu id#2 (270X).

Wrote "Error CL_INVALID_DEVICE when calling clCreateContext."

If I set gpu0 or gpu1 or all 3 cards miner starts to mine.

I can add gpu/mem clock parameters, that's not a problem, voltages we will see

can you put a screenshot how it looks when all three cards work, but 270x does not report temp/fans rpm ?

Of course, here

ScreenShot
How I can launch only on 270X?

Looks like it thinks your 270x gpu id is 47, not 2 LOL, now this is really strange.
Did you try leaving out the gpu_conf array and let the miner detect every card ?
Ex. set intensity:0 and leave double_threads off for testing.

[doktor83]

HD 6950 gets 200h/sec on monero in claymore 9.7
R9 270X gets 430h/sec on monero in claymore 9.7

shows a hashrate in the console but I doubt the results are valid

[2018-03-25 05:54:29] Pool accepted result 0x0000A903
[2018-03-25 05:54:30] Error CL_INVALID_WORK_GROUP_SIZE when calling clEnqueueNDR
angeKernel for kernel 3.
[2018-03-25 05:54:31] GPU0[T0]:     419.0 H/S
[2018-03-25 05:54:31] GPU1[T1]:     852.0 H/S
[2018-03-25 05:54:31] Total:        1271.0 H/S

it appears that the error is coming from the R9 270X and not the HD 6900... if kernel 3 means BUS_ID 3

No, kernel 3 has nothing to do with gpu id or bus id 3.
I will do a little more useful error messages so we can find out what is causing the problem.

[doktor83]

Ok, guys download V1.1.1 and let's try to solve misteries

https://mega.nz/#F!qVIgxAwB!kKmgCDICmQwbdVvMb-tAag

[Bulitt]

Same error but it says Thread 0 now which indicates it is actually the HD 6950.

Oh well I guess that card is going to have to be on neoscrypt now.

xmr-stak suffers the same problems and all he offered was some patches you could add an compile yourself but no binaries.

https://github.com/fireice-uk/xmr-stak-amd/issues/22

[xberg]

I'm not a troll, but is there still profit in Cryptonight mining?
I used to do Cryptonight mining with my powerful xeon processors in January - February but stopped in March as profits < electricity (I'm paying EUR 0.13 kWh).
AMD R9 Fury also is not covering electricity costs with Crypto on Nicehash.
So am I missing something? Is crypto mining still viable?

[doktor83]

Same error but it says Thread 0 now which indicates it is actually the HD 6950.

Oh well I guess that card is going to have to be on neoscrypt now.

xmr-stak suffers the same problems and all he offered was some patches you could add an compile yourself but no binaries.

https://github.com/fireice-uk/xmr-stak-amd/issues/22

i will look into it, to integrate that 'patch'

[UnclWish]

Situation with 270X stays the same.
If I try to launch 270X alone, miner gives error CL_INVALID_DEVICE.
If I launch it with 2 RX580 launcher works.
Temperature and fan control for 270X didn't work.

Working program says the version 1.0 still...

[doktor83]

I guess it only works with GCN or newer cards like almost every other miner created

[2018-03-25 05:08:15] Error CL_INVALID_WORK_GROUP_SIZE when calling clEnqueueNDRangeKernel for kernel 4.

scrolling...

Try V1.1.1, i applied a fix that addresses this issue.

[doktor83]

Situation with 270X stays the same.
If I try to launch 270X alone, miner gives error CL_INVALID_DEVICE.
If I launch it with 2 RX580 launcher works.
Temperature and fan control for 270X didn't work.

Working program says the version 1.0 still...

Fixed the version thing
Also could you re-download V1.1.1 from mega and try 270x alone, it should give a longer description on error like :

X threads found, here's a list of deviceIdx : ...

Could you paste that here please

[UnclWish]

Situation with 270X stays the same.
If I try to launch 270X alone, miner gives error CL_INVALID_DEVICE.
If I launch it with 2 RX580 launcher works.
Temperature and fan control for 270X didn't work.

Working program says the version 1.0 still...

Fixed the version thing
Also could you re-download V1.1.1 from mega and try 270x alone, it should give a longer description on error like :

X threads found, here's a list of deviceIdx : ...

Could you paste that here please

Here log:

[2018-03-26 00:53:40] Heating up system, please wait...
[2018-03-26 00:53:40] Error CL_INVALID_DEVICE when calling clCreateContext
[2018-03-26 00:53:40] 1 threads found, here's a list of deviceIdx:
[2018-03-26 00:53:40] ID: 2 [ThreadID: 0]

[Mind163]

/* This is just an example, edit it and remove comment lines (the slash and star) !! */
"gpu_conf" :
[
{ "id" : 0, "intensity" : 13, "double_threads" : true},
{ "id" : 1, "intensity" : 13, "double_threads" : true},
{ "id" : 2, "intensity" : 13, "double_threads" : true},
{ "id" : 3, "intensity" : 13, "double_threads" : true},
{ "id" : 4, "intensity" : 13, "double_threads" : true},
{ "id" : 5, "intensity" : 12, "double_threads" : true},
]

/* Make sure not to delete this last blank line */

this is how the end of the config looks like

Try download this example config and run it :

https://mega.nz/#F!yJxTTZpK!ZA6WVs3kIfcw_slX3dgw1Q

In the new version, everything turned out. I would very much like that you could make additional pools in the config. It's real?

I also tried to do this, but nothing happened
"gpu_conf" :
[
{ "id" : 0, "intensity" : 13, "double_threads" : true},
{ "id" : 1, "intensity" : 13, "double_threads" : true},
{ "id" : 2, "intensity" : 13, "double_threads" : false},
{ "id" : 2, "intensity" : 12, "double_threads" : false},
{ "id" : 3, "intensity" : 13, "double_threads" : true},
{ "id" : 4, "intensity" : 12, "double_threads" : true},
]

[UnclWish]

/* This is just an example, edit it and remove comment lines (the slash and star) !! */
"gpu_conf" :
[
{ "id" : 0, "intensity" : 13, "double_threads" : true},
{ "id" : 1, "intensity" : 13, "double_threads" : true},
{ "id" : 2, "intensity" : 13, "double_threads" : true},
{ "id" : 3, "intensity" : 13, "double_threads" : true},
{ "id" : 4, "intensity" : 13, "double_threads" : true},
{ "id" : 5, "intensity" : 12, "double_threads" : true},
]

/* Make sure not to delete this last blank line */

this is how the end of the config looks like

Try download this example config and run it :

https://mega.nz/#F!yJxTTZpK!ZA6WVs3kIfcw_slX3dgw1Q

In the new version, everything turned out. I would very much like that you could make additional pools in the config. It's real?

I also tried to do this, but nothing happened
"gpu_conf" :
[
{ "id" : 0, "intensity" : 13, "double_threads" : true},
{ "id" : 1, "intensity" : 13, "double_threads" : true},
{ "id" : 2, "intensity" : 13, "double_threads" : false},
{ "id" : 2, "intensity" : 12, "double_threads" : false},
{ "id" : 3, "intensity" : 13, "double_threads" : true},
{ "id" : 4, "intensity" : 12, "double_threads" : true},
]

Why you two times point the same gpu?

[Bulitt]

I guess it only works with GCN or newer cards like almost every other miner created

[2018-03-25 05:08:15] Error CL_INVALID_WORK_GROUP_SIZE when calling clEnqueueNDRangeKernel for kernel 4.

scrolling...

Try V1.1.1, i applied a fix that addresses this issue.

You did it! Thank you so much my HD 6950 thanks you, it even appears to work in AEON mode without "AMD Invalid Result" like XMR-stak does

edit: I spoke too soon, getting rejected shares compute error now At least it works with monero and thats an improvement on whats out there right now.

[Mind163]

Is it possible to make the intensity more differentiated. For example 12.5. For example, RX550 2gb-12 is not enough for them, 13 is the limit. They are already hard, periodically disconnected.