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

[abg00]

interesting..and on 1.6.7 you get the 1000+ speed on all cards ?

[1606541909]

1606541909

[1606541909]

1606541909

[1606541909]

1606541909

[doktor83]

interesting..and on 1.6.7 you get the 1000+ speed on all cards ?

Ok going to re-check and test.

Can you test on bittube too?

[GKumaran]

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

[doktor83]

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

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

[abg00]

Can you test on bittube too?

check rx550\4g on cn-v7
same trouble in 1.6.8
low rate
420-450 in 1.6.7
370-390 in 1.6.8


rx550\2g rate high in 1.6.8 was 460 now 490

[doktor83]

Can you test on bittube too?

check rx550\4g on cn-v7
same trouble in 1.6.8
low rate
420-450 in 1.6.7
370-390 in 1.6.8


rx550\2g rate high in 1.6.8 was 460 now 490

have you tried auto intensity for 550?

also it would be nice if you could write your settings

[abg00]

have you tried auto intensity for 550?

also it would be nice if you could write your settings

no
for 2gb 1.6.8
550
   { "id" : 6, "intensity" : 27,  "worksize" : 12, "double_threads" : true},
 GPU6 :    504 H/s [T: 49c, RPM: 0, CC: 1050 MHz, MC: 1950 MHz][BUS:4]
560
   { "id" : 7, "intensity" : 27,  "worksize" : 12, "double_threads" : true},
GPU7 :    488 H/s [T: 53c, RPM: 2635, CC: 1050 MHz, MC: 1850 MHz][BUS:2]

for 550 4gb 1.6.7
   { "id" : 0, "intensity" : 54, "double_threads" : true, "worksize": 12},
 GPU0 :    416 H/s [T: 51c, RPM: 2833, CC: 1090 MHz, MC: 2050 MHz][BUS:1]
same 550\4 in 1.6.8 only 380-390 h\s

[GKumaran]

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

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

No doc, still the same result.
My v8 config:
{
"cryptonight_type" : "normalv8",
"intensity" : 120,
"worksize" : 16,
"double_threads" : true
}

Benchmark results:
Stopping benchmark (run for 180 seconds)

Average speed: 2851.5 H/s
Shares found: 21

Abv test is for 6 Vega 64s

[doktor83]

have you tried auto intensity for 550?

also it would be nice if you could write your settings

no
for 2gb 1.6.8
550
   { "id" : 6, "intensity" : 27,  "worksize" : 12, "double_threads" : true},
 GPU6 :    504 H/s [T: 49c, RPM: 0, CC: 1050 MHz, MC: 1950 MHz][BUS:4]
560
   { "id" : 7, "intensity" : 27,  "worksize" : 12, "double_threads" : true},
GPU7 :    488 H/s [T: 53c, RPM: 2635, CC: 1050 MHz, MC: 1850 MHz][BUS:2]

for 550 4gb 1.6.7
   { "id" : 0, "intensity" : 54, "double_threads" : true, "worksize": 12},
 GPU0 :    416 H/s [T: 51c, RPM: 2833, CC: 1090 MHz, MC: 2050 MHz][BUS:1]
same 550\4 in 1.6.8 only 380-390 h\s

Intensity 54 too much with 2 threads. Why not 27 ?

[doktor83]

No doc, still the same result.
My v8 config:
{
"cryptonight_type" : "normalv8",
"intensity" : 120,
"worksize" : 16,
"double_threads" : true
}

Benchmark results:
Stopping benchmark (run for 180 seconds)

Average speed: 2851.5 H/s
Shares found: 21

Abv test is for 6 Vega 64s

Try maybe worksize 32. I don't have vega64 to test with, but i have a 56, and the best hash for it is at 120/16, it should be around the same for 64 too.

[abg00]

have you tried auto intensity for 550?

also it would be nice if you could write your settings

no
for 2gb 1.6.8
550
   { "id" : 6, "intensity" : 27,  "worksize" : 12, "double_threads" : true},
 GPU6 :    504 H/s [T: 49c, RPM: 0, CC: 1050 MHz, MC: 1950 MHz][BUS:4]
560
   { "id" : 7, "intensity" : 27,  "worksize" : 12, "double_threads" : true},
GPU7 :    488 H/s [T: 53c, RPM: 2635, CC: 1050 MHz, MC: 1850 MHz][BUS:2]

for 550 4gb 1.6.7
   { "id" : 0, "intensity" : 54, "double_threads" : true, "worksize": 12},
 GPU0 :    416 H/s [T: 51c, RPM: 2833, CC: 1090 MHz, MC: 2050 MHz][BUS:1]
same 550\4 in 1.6.8 only 380-390 h\s

Intensity 54 too much with 2 threads. Why not 27 ?

27 is ok for 1.6.8 rx550\4gb rate 440 h\s
30 is not ok. bad rate

[doktor83]

have you tried auto intensity for 550?

also it would be nice if you could write your settings

no
for 2gb 1.6.8
550
   { "id" : 6, "intensity" : 27,  "worksize" : 12, "double_threads" : true},
 GPU6 :    504 H/s [T: 49c, RPM: 0, CC: 1050 MHz, MC: 1950 MHz][BUS:4]
560
   { "id" : 7, "intensity" : 27,  "worksize" : 12, "double_threads" : true},
GPU7 :    488 H/s [T: 53c, RPM: 2635, CC: 1050 MHz, MC: 1850 MHz][BUS:2]

for 550 4gb 1.6.7
   { "id" : 0, "intensity" : 54, "double_threads" : true, "worksize": 12},
 GPU0 :    416 H/s [T: 51c, RPM: 2833, CC: 1090 MHz, MC: 2050 MHz][BUS:1]
same 550\4 in 1.6.8 only 380-390 h\s

Intensity 54 too much with 2 threads. Why not 27 ?

27 is ok for 1.6.8 rx550\4gb rate 440 h\s
30 is not ok. bad rate

so what is the conclusion at the end?
1.6.8 hashing better on rx550 and rx560 cards ?

also where did you get intensity 30 from?
You are really confusing me here i don;t know even what algo are we talking about now.

550 4g should be ok on V7 with : 27/12/2
560 4g should be ok on V7 with : 27/8/2

550 4g should be ok on V8 with : 28/16/2
560 4g should be ok on V8 with : 29/32/2

[abg00]

so what is the conclusion at the end?
1.6.8 hashing better on rx550 and rx560 cards ?

also where did you get intensity 30 from?
You are really confusing me here i don;t know even what algo are we talking about now.

550 4g should be ok on V7 with : 27/12/2
560 4g should be ok on V7 with : 27/8/2

550 4g should be ok on V8 with : 28/16/2
560 4g should be ok on V8 with : 29/32/2

true.
but have trouble on cn-heavy with rx470\480\570\580 8gb card. plz check my config
   { "id" : 1, "intensity" : 58, "double_threads" : true},
on 1.6.7 ok and rate 1060
 GPU1 :    1027 H/s [T: 67c, RPM: 1465, CC: 1150 MHz, MC: 2030 MHz][BUS:2]
on 1.6.8 low rate
GPU1 :    875 H/s [T: 63c, RPM: 945, CC: 1150 MHz, MC: 2030 MHz][BUS:2]

[GKumaran]

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  

[doktor83]

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

[doktor83]

so what is the conclusion at the end?
1.6.8 hashing better on rx550 and rx560 cards ?

also where did you get intensity 30 from?
You are really confusing me here i don;t know even what algo are we talking about now.

550 4g should be ok on V7 with : 27/12/2
560 4g should be ok on V7 with : 27/8/2

550 4g should be ok on V8 with : 28/16/2
560 4g should be ok on V8 with : 29/32/2

true.
but have trouble on cn-heavy with rx470\480\570\580 8gb card. plz check my config
   { "id" : 1, "intensity" : 58, "double_threads" : true},
on 1.6.7 ok and rate 1060
 GPU1 :    1027 H/s [T: 67c, RPM: 1465, CC: 1150 MHz, MC: 2030 MHz][BUS:2]
on 1.6.8 low rate
GPU1 :    875 H/s [T: 63c, RPM: 945, CC: 1150 MHz, MC: 2030 MHz][BUS:2]

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.

[GKumaran]

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 !!

[livada]

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

[doktor83]

nice test. try Worksize 8
my vega64LE have better HR with 8 worksize

V8 is not good with worksize 8 on any card.

[doktor83]

V.1.6.8 is again up on Mega! Sorry for any inconvenience!