What intensity/worksize/threads combo are you recommending on RX 570's 4GB for the new v8?
try 57/16/2
Topic: SRBMiner Cryptonight AMD GPU Miner V1.9.3 - native algo switching - page 144. (Read 237247 times)
October 09, 2018, 12:28:39 AM
try 57/16/2
October 08, 2018, 11:00:16 PM
Nice Dok! Can confirm ~2035 H/s for a V64 on V8. ~2135 H/s on current V7. V8 Only works using your pre-compiled kernels though. Running 18.6.1 drivers which are recommended, strange that they can't compile for V8. Just wanted to pass the info along, thx for all your wonderful work, best CN miner in my book by a long shot!
Hmm, testing with a single Vega56. Was getting ~2035 with 1.6.7, now getting ~1630 with 1.6.8. Tried both the Normal and Maxi versions, same results. Using the 18.6.1 drivers, so not sure what to make of this
October 08, 2018, 10:51:36 PM
Could anyone help me,when i start the miner with my r9 290 i get the following error:
[2018-10-09 14:48:01] Error CL_LINK_PROGRAM_FAILURE when calling clLinkProgram [ctx->Program] for DeviceID 0 (Thread 0)
[2018-10-09 14:48:01] Error CL_INVALID_PROGRAM when calling clGetProgramBuildInfo for length of build log output for DeviceID 0 (Thread 0)
[2018-10-09 14:48:01] Error CL_LINK_PROGRAM_FAILURE when calling clLinkProgram [ctx->Program] for DeviceID 0 (Thread 0)
[2018-10-09 14:48:01] Error CL_INVALID_PROGRAM when calling clGetProgramBuildInfo for length of build log output for DeviceID 0 (Thread 0)
October 08, 2018, 07:45:19 PM
Anyone still mining with the older Radeon 7970/280X or the R9 290 GPUs?
Was wondering what config you guys are using and what stability you guys get with this software, I am looking for an alternative to the Claymore XMR miner which will become obsolete when XMR forks on Oct 18th.
Was wondering what config you guys are using and what stability you guys get with this software, I am looking for an alternative to the Claymore XMR miner which will become obsolete when XMR forks on Oct 18th.
October 08, 2018, 07:41:22 PM
Nice Dok! Can confirm ~2035 H/s for a V64 on V8. ~2135 H/s on current V7. V8 Only works using your pre-compiled kernels though. Running 18.6.1 drivers which are recommended, strange that they can't compile for V8. Just wanted to pass the info along, thx for all your wonderful work, best CN miner in my book by a long shot!
October 08, 2018, 06:48:11 PM
Hey Doc, thc for update !!
Using Binary kernals on Vega 64:
v7 : HR up from 2130 to 2145
v8 : HR is 580
Is that normal, only 580 H/s? Cause i think the devs said approx 7% drop so was expecting at least 1950 h/s
I got the same HR result for v8 on both benchmark mode and mining on the test pools:
"pool": "killallasics.moneroworld.com:7777",
"difficulty": 10000,
{
"device": "GPU0",
"device_id": 0,
"model": "Radeon RX Vega",
"bus_id": 19,
"kernel_id": 1,
"hashrate": 593,
"core_clock": 1630,
"memory_clock": 1100,
"temperature": 54,
"fan_speed_rpm": 4576
},
GPU was on dual thread and auto worksize, auto intensity
Using Binary kernals on Vega 64:
v7 : HR up from 2130 to 2145
v8 : HR is 580
Is that normal, only 580 H/s? Cause i think the devs said approx 7% drop so was expecting at least 1950 h/s
I got the same HR result for v8 on both benchmark mode and mining on the test pools:
"pool": "killallasics.moneroworld.com:7777",
"difficulty": 10000,
{
"device": "GPU0",
"device_id": 0,
"model": "Radeon RX Vega",
"bus_id": 19,
"kernel_id": 1,
"hashrate": 593,
"core_clock": 1630,
"memory_clock": 1100,
"temperature": 54,
"fan_speed_rpm": 4576
},
GPU was on dual thread and auto worksize, auto intensity
how you solved this?
Yes something not ok, try setting by hand for example 120/16 2 threads
I really hurried to push the v8 version, so of course bugs are to be expected, and we still have time to fix them until the fork
Hey Doc,
Vega 64 - Cryptonight v8 analysis:
Using binary kernals : 583 h/s
Not Using Binary Kernals, Kernals compiled on each run and deleted
Driver: 18.5.1 GPU: 1600 Mem: 1100 Volt : 840mV
Intensity : 119 Worksize : 16 : 2019 H/s (94.338% on cn7)
Intensity : 119 Worksize : 32 : 1963 H/s (91.745% on cn7)
Intensity : 120 Worksize : 16 : 2025 H/s (94.626% on cn7)
Intensity : 120 Worksize : 32 : 1983 H/s (92.663% on cn7)
Intensity : 121 Worksize : 16 : 2032 H/s (94.953% on cn7)
Intensity : 121 Worksize : 32 : 1975 H/s (92.289% on cn7)
Intensity : 122 Worksize : 16 : 2038 H/s (95.234% on cn7)
Intensity : 122 Worksize : 32 : 1992 H/s (93.099% on cn7)
Intensity : 123 Worksize : 16 : 2043 H/s (95.467% on cn7)
Intensity : 123 Worksize : 32 : 1984 H/s (92.710% on cn7)
Intensity : 124 Worksize : 16 : 2049 H/s (95.748% on cn7)
Intensity : 124 Worksize : 32 : 2000 H/s (93.458% on cn7)
Intensity : 125 Worksize : 16 : Failed
Intensity : 125 Worksize : 32 : Failed
The above are avg HR from a run of 6 Vega 64s and the highest HR setting were tries multiple times for consistency.
Great work doc, so thats a 4.25% drop from v7 !! not bad considering devs said 7% drop
i use your int work but HR only 1100
in v8
October 08, 2018, 06:26:39 PM
What intensity/worksize/threads combo are you recommending on RX 570's 4GB for the new v8?
October 08, 2018, 05:20:45 PM
I'm using RX580 8G and version 1.6.8 is working slower with a total of 400H/s versus version 1.6.7 of heavy 
i use auto (intensity 0)
i use auto (intensity 0)
October 08, 2018, 04:59:29 PM
1.6.8 gave additional 6-8 hashes per Vega using same settings on normalv7. nice 
October 08, 2018, 02:52:56 PM
i think i found something about the 570/580 8g cards and heavy algos. Going to check out a few more things then re-upload 1.6.8 with the fixes.
check new version/ same trouble ((
also check without precompile binaries. same low rate 900-930 not 1000+ as on 1.6.7
possible need other drivers? use 18.3.4
Driver should be fine, but you could try 18.6.1.
I have a mixed rig 570/580 8G and this is the result now, everything left on auto (intensity 0)
reinstall drivers to 18.6.1 same trouble with rate ((
possible trouble because 12 card on rig?
all 12 gpu rx570 8 gb no different
swap 98 gb
but 1.6.7 1000+ rate
October 08, 2018, 02:09:09 PM
i think i found something about the 570/580 8g cards and heavy algos. Going to check out a few more things then re-upload 1.6.8 with the fixes.
check new version/ same trouble ((
also check without precompile binaries. same low rate 900-930 not 1000+ as on 1.6.7
possible need other drivers? use 18.3.4
Driver should be fine, but you could try 18.6.1.
I have a mixed rig 570/580 8G and this is the result now, everything left on auto (intensity 0)
October 08, 2018, 02:03:56 PM
V.1.6.8 is again up on Mega! Sorry for any inconvenience!
Downloaded again and tested Maxi version with driver 18.3.4 and 2 Vega 56 Reference card, one Vega 64 Asus Strix and one Rx 570 GB.
Vega problem is still here. Low performance about 4100 H/s for cards (on V7 is 7070 H/s with same settings).
That is from benchmark test 60 sec, but problem is there.
what if you forget all that strange config and comment out the gpu_conf part?
I don't understand you Doctor? That GPU_conf part is rock solid in V7 algorithm for 7 days with 7070 H/s.
Same settings in SrbMiner and overdrive settings in V8 results with 4100 H/s in benchmark mode.
Do you need some more specific test, i can test it?
yes , rock solid in v7, but the settings for v8 are not same as for v7. V8 for example is good with worksize 16 and 32, and bad with 8. Why do you guys think the same settings need to be, when this is a different algo?
vega 56 120/16/2 , vega 64 124/16/2 is optimal
October 08, 2018, 01:58:16 PM
V.1.6.8 is again up on Mega! Sorry for any inconvenience!
Downloaded again and tested Maxi version with driver 18.3.4 and 2 Vega 56 Reference card, one Vega 64 Asus Strix and one Rx 570 GB.
Vega problem is still here. Low performance about 4100 H/s for cards (on V7 is 7070 H/s with same settings).
That is from benchmark test 60 sec, but problem is there.
what if you forget all that strange config and comment out the gpu_conf part?
I don't understand you Doctor? That GPU_conf part is rock solid in V7 algorithm for 7 days with 7070 H/s.
Same settings in SrbMiner and overdrive settings in V8 results with 4100 H/s in benchmark mode.
Do you need some more specific test, i can test it?
And for commenting gpu_conf:
[
{ "id" : 0, "intensity" : 120, "worksize" : 16, "threads" : 1}, - Vega 56 Reference 1
{ "id" : 0, "intensity" : 120, "worksize" : 16, "threads" : 1}, - Vega 56 Reference 1
{ "id" : 1, "intensity" : 56, "worksize" : 8, "threads" : 1}, - RX 570 4GB 1
{ "id" : 1, "intensity" : 56, "worksize" : 8, "threads" : 1}, - RX 570 4GB 1
{ "id" : 2, "intensity" : 120, "worksize" : 8, "threads" : 1}, - Vega 64 Asus Strix 1 - better work with worksize 8, at least in V7
{ "id" : 2, "intensity" : 122, "worksize" : 8, "threads" : 1}, - Vega 64 Asus Strix 1
{ "id" : 3, "intensity" : 120, "worksize" : 16, "threads" : 1}, - Vega 56 Reference 2
{ "id" : 3, "intensity" : 120, "worksize" : 16, "threads" : 1} - Vega 56 Reference 2
]
October 08, 2018, 01:29:47 PM
V.1.6.8 is again up on Mega! Sorry for any inconvenience!
Downloaded again and tested Maxi version with driver 18.3.4 and 2 Vega 56 Reference card, one Vega 64 Asus Strix and one Rx 570 GB.
Vega problem is still here. Low performance about 4100 H/s for cards (on V7 is 7070 H/s with same settings).
{
"cryptonight_type" : "normalv8",
"intensity" : 0,
"double_threads" : true,
"min_rig_speed" : 6500,
"gpu_conf" :
[
{ "id" : 0, "intensity" : 120, "worksize" : 16, "threads" : 1},
{ "id" : 0, "intensity" : 120, "worksize" : 16, "threads" : 1},
{ "id" : 1, "intensity" : 56, "worksize" : 8, "threads" : 1},
{ "id" : 1, "intensity" : 56, "worksize" : 8, "threads" : 1},
{ "id" : 2, "intensity" : 120, "worksize" : 8, "threads" : 1},
{ "id" : 2, "intensity" : 122, "worksize" : 8, "threads" : 1},
{ "id" : 3, "intensity" : 120, "worksize" : 16, "threads" : 1},
{ "id" : 3, "intensity" : 120, "worksize" : 16, "threads" : 1}
]
}
That is from benchmark test 60 sec, but problem is there.
what if you forget all that strange config and comment out the gpu_conf part?
October 08, 2018, 01:22:56 PM
V.1.6.8 is again up on Mega! Sorry for any inconvenience!
Downloaded again and tested Maxi version with driver 18.3.4 and 2 Vega 56 Reference card, one Vega 64 Asus Strix and one Rx 570 GB.
Vega problem is still here. Low performance about 4100 H/s for cards (on V7 is 7070 H/s with same settings).
{
"cryptonight_type" : "normalv8",
"intensity" : 0,
"double_threads" : true,
"min_rig_speed" : 6500,
"gpu_conf" :
[
{ "id" : 0, "intensity" : 120, "worksize" : 16, "threads" : 1},
{ "id" : 0, "intensity" : 120, "worksize" : 16, "threads" : 1},
{ "id" : 1, "intensity" : 56, "worksize" : 8, "threads" : 1},
{ "id" : 1, "intensity" : 56, "worksize" : 8, "threads" : 1},
{ "id" : 2, "intensity" : 120, "worksize" : 8, "threads" : 1},
{ "id" : 2, "intensity" : 122, "worksize" : 8, "threads" : 1},
{ "id" : 3, "intensity" : 120, "worksize" : 16, "threads" : 1},
{ "id" : 3, "intensity" : 120, "worksize" : 16, "threads" : 1}
]
}
That is from benchmark test 60 sec, but problem is there.
October 08, 2018, 01:16:55 PM
i think i found something about the 570/580 8g cards and heavy algos. Going to check out a few more things then re-upload 1.6.8 with the fixes.
check new version/ same trouble ((
also check without precompile binaries. same low rate 900-930 not 1000+ as on 1.6.7
possible need other drivers? use 18.3.4
October 08, 2018, 12:35:00 PM
V.1.6.8 is again up on Mega! Sorry for any inconvenience!
October 08, 2018, 12:34:23 PM
nice test. try Worksize 8
my vega64LE have better HR with 8 worksize
V8 is not good with worksize 8 on any card.
October 08, 2018, 12:33:06 PM
Yes something not ok, try setting by hand for example 120/16 2 threads
I really hurried to push the v8 version, so of course bugs are to be expected, and we still have time to fix them until the fork
Hey Doc,
Vega 64 - Cryptonight v8 analysis:
Using binary kernals : 583 h/s
Not Using Binary Kernals, Kernals compiled on each run and deleted
Driver: 18.5.1 GPU: 1600 Mem: 1100 Volt : 840mV
Intensity : 119 Worksize : 16 : 2019 H/s (94.338% on cn7)
Intensity : 119 Worksize : 32 : 1963 H/s (91.745% on cn7)
Intensity : 120 Worksize : 16 : 2025 H/s (94.626% on cn7)
Intensity : 120 Worksize : 32 : 1983 H/s (92.663% on cn7)
Intensity : 121 Worksize : 16 : 2032 H/s (94.953% on cn7)
Intensity : 121 Worksize : 32 : 1975 H/s (92.289% on cn7)
Intensity : 122 Worksize : 16 : 2038 H/s (95.234% on cn7)
Intensity : 122 Worksize : 32 : 1992 H/s (93.099% on cn7)
Intensity : 123 Worksize : 16 : 2043 H/s (95.467% on cn7)
Intensity : 123 Worksize : 32 : 1984 H/s (92.710% on cn7)
Intensity : 124 Worksize : 16 : 2049 H/s (95.748% on cn7)
Intensity : 124 Worksize : 32 : 2000 H/s (93.458% on cn7)
Intensity : 125 Worksize : 16 : Failed
Intensity : 125 Worksize : 32 : Failed
The above are avg HR from a run of 6 Vega 64s and the highest HR setting were tries multiple times for consistency.
Great work doc, so thats a 4.25% drop from v7 !! not bad considering devs said 7% drop
nice test. try Worksize 8
my vega64LE have better HR with 8 worksize
October 08, 2018, 11:27:43 AM
Yes something not ok, try setting by hand for example 120/16 2 threads
I really hurried to push the v8 version, so of course bugs are to be expected, and we still have time to fix them until the fork
Hey Doc,
Vega 64 - Cryptonight v8 analysis:
Using binary kernals : 583 h/s
Not Using Binary Kernals, Kernals compiled on each run and deleted
Driver: 18.5.1 GPU: 1600 Mem: 1100 Volt : 840mV
Intensity : 119 Worksize : 16 : 2019 H/s (94.338% on cn7)
Intensity : 119 Worksize : 32 : 1963 H/s (91.745% on cn7)
Intensity : 120 Worksize : 16 : 2025 H/s (94.626% on cn7)
Intensity : 120 Worksize : 32 : 1983 H/s (92.663% on cn7)
Intensity : 121 Worksize : 16 : 2032 H/s (94.953% on cn7)
Intensity : 121 Worksize : 32 : 1975 H/s (92.289% on cn7)
Intensity : 122 Worksize : 16 : 2038 H/s (95.234% on cn7)
Intensity : 122 Worksize : 32 : 1992 H/s (93.099% on cn7)
Intensity : 123 Worksize : 16 : 2043 H/s (95.467% on cn7)
Intensity : 123 Worksize : 32 : 1984 H/s (92.710% on cn7)
Intensity : 124 Worksize : 16 : 2049 H/s (95.748% on cn7)
Intensity : 124 Worksize : 32 : 2000 H/s (93.458% on cn7)
Intensity : 125 Worksize : 16 : Failed
Intensity : 125 Worksize : 32 : Failed
The above are avg HR from a run of 6 Vega 64s and the highest HR setting were tries multiple times for consistency.
Great work doc, so thats a 4.25% drop from v7 !! not bad considering devs said 7% drop
That's my man, i really love it when someone takes time and tries himself something
So it looks like the binary kernel is faulty ? I will check that out.
Thank you for your efforts
edit:
And YES, the binary IS faulty. Great that i pushed the 1.6.8 version 10 days before the fork so we can smooth these kind of things out
:)no prob mate !! Jump to: