What wrong with my MSI Afterbuner ?
https://drive.google.com/file/d/0B7Bc6ZeuUtjgV0V2V3ZmUml6cWM/view?usp=sharing
Topic: Claymore's CryptoNote AMD GPU Miner v11.3 - page 91. (Read 2145043 times)
Tested Claymore v10.2 on 8 rx 580 8 GB. Got the same hash rate 830-850 h/s per card as with sgminer-5.5.5. Will stay with sgminer-5.5.5 until see significant improvement in Claymore's miner
That mirrors my own observations. Thanks for sharing.
Tested Claymore v10.2 on 8 rx 580 8 GB. Got the same hash rate 830-850 h/s per card as with sgminer-5.5.5. Will stay with sgminer-5.5.5 until see significant improvement in Claymore's miner
Yes, I moved back to my xmr-stak-amd with polaris 8gb cards. Little bit more pool-side hash (but I can be wrong).Using Claymore's 10.2 with hawaii and polaris 4gb cards only.
Tested Claymore v10.2 on 8 rx 580 8 GB. Got the same hash rate 830-850 h/s per card as with sgminer-5.5.5. Will stay with sgminer-5.5.5 until see significant improvement in Claymore's miner
getting poor performance (best 3500 h/s after running for 10 mins) and sometimes GPU hangs. Config default / No overclock.
C: Drive = 20,000mb paging file / D: Drive = 100,000mb paging file
4GB Ram / Intel Celeron 2 core GPU.
9.7 I get 5600 h/s no overclock 6000 h/s with overclock
Snip from 10.2 log
22:47:47:282 1a34 driver 10.0.2442.8
22:47:47:286 1a34 AMD Cards available: 8
22:47:50:287 1a34 GPU #0 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:50:664 1a34 GPU #1 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:51:022 1a34 GPU #2 algorithm ASM, -h 900, -dmem 1 (Memory used: 3780MB)
22:47:51:379 1a34 GPU #3 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:51:741 1a34 GPU #4 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:52:106 1a34 GPU #5 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:52:504 1a34 GPU #6 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:52:888 1a34 GPU #7 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:52:892 1a34 Total cards: 8
C: Drive = 20,000mb paging file / D: Drive = 100,000mb paging file
4GB Ram / Intel Celeron 2 core GPU.
9.7 I get 5600 h/s no overclock 6000 h/s with overclock
Snip from 10.2 log
22:47:47:282 1a34 driver 10.0.2442.8
22:47:47:286 1a34 AMD Cards available: 8
22:47:50:287 1a34 GPU #0 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:50:664 1a34 GPU #1 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:51:022 1a34 GPU #2 algorithm ASM, -h 900, -dmem 1 (Memory used: 3780MB)
22:47:51:379 1a34 GPU #3 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:51:741 1a34 GPU #4 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:52:106 1a34 GPU #5 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:52:504 1a34 GPU #6 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:52:888 1a34 GPU #7 algorithm ASM, -h 1800, -dmem 1 (Memory used: 7560MB)
22:47:52:892 1a34 Total cards: 8
Hello,
This are my results whit 7 card rig ( 5xRX580+2xRX470 all bios moded )
With 9.7
Total H/S was 4690
With 10.2
Total H/S is 4790
Had to increase virtual memory to 52000MB to fix error dk2a
So, is that ok?
Also with 9.7 i had no temperatures and fan speeds with 10.2 now that is working.
This are my results whit 7 card rig ( 5xRX580+2xRX470 all bios moded )
With 9.7
Total H/S was 4690
With 10.2
Total H/S is 4790
Had to increase virtual memory to 52000MB to fix error dk2a
So, is that ok?
Also with 9.7 i had no temperatures and fan speeds with 10.2 now that is working.
this feels like pre-alpha version.
on a mixed rig of 470&370 running stable XMR for days it just refuses to work(very low hash and GPU hangs) on default miner settings, no matter the frequencies/voltages. (using 16.11.4 drivers)
not yet usable for now...
on a mixed rig of 470&370 running stable XMR for days it just refuses to work(very low hash and GPU hangs) on default miner settings, no matter the frequencies/voltages. (using 16.11.4 drivers)
not yet usable for now...
I cant get my 470 8gb running good at default h there at 310 if I set -h higher the card hashes lower the best I could get is -h 900 at 500-600 hashes which is horrible on gatelessgate I get 700-800 hashes
There is something wrong with 10.2 and Pitcairns.
On my 270X 4Gb with -dmem 1 and stock frequences 1070/1400 speed 515 h/s and about 50-75% incorrect shares. No rejects. Without -dmem 1 option card even with overclocking gives only 485 h/s maximum. I didn't check for incorrect shares.
On 9.7 with overclock i recieve 595 h/s and no rejects. Returning to 9.7 for now. Only 1 plus of 10.2 - it more power effective and card cooler and quieter.
Hope that Claymore fix this problem.
On my 270X 4Gb with -dmem 1 and stock frequences 1070/1400 speed 515 h/s and about 50-75% incorrect shares. No rejects. Without -dmem 1 option card even with overclocking gives only 485 h/s maximum. I didn't check for incorrect shares.
On 9.7 with overclock i recieve 595 h/s and no rejects. Returning to 9.7 for now. Only 1 plus of 10.2 - it more power effective and card cooler and quieter.
Hope that Claymore fix this problem.
Does Miners Works with Nvidia's Card ?
When miner runs long-term, hash reduce...Restart hash increase again.
I've noticed this as well. And it's not simply restarting the miner; I run mine with -r 240 so it should be restarting every 4 hours or GPU fail (I've not confirmed if this is happening in the logs yet FWIW). I got up this morning and hash rate had dropped and didn't come back until I rebooted. Bounced my machine and blamo... right back up to the 5130-5140 range.
Wolf0 congrats, people says your miner is faster than Claymore!
Meh, it is in some cases, but only a bit. If I wanted to have a contest, I wouldn't hand the competitor my entry, duh.
My public one is on Github! Or in SGMiner-GM.
Hey wolf, thank you for all your work. Have you done any testing with the vega frontier? Any clue of optimal settings? 4 threads possible/better?
Since xmr-stak-amd is just a new miner shell plus my GPU code ripped off, I guess the stats for it are identical.
Your work is very good. It gets hightest hash.
When miner runs long-term, hash reduce...Restart hash increase again.
Some things to check.
- One of my rigs started using 3,5 GB of ram. Closed the miner and it dropped to the usual 420 Mb.
- On another rig one GPU dropped to 0 and another GPU dropped to 500 (RX 470 usually working at 750-770). Restarted the miner also.
- One of my rigs started using 3,5 GB of ram. Closed the miner and it dropped to the usual 420 Mb.
- On another rig one GPU dropped to 0 and another GPU dropped to 500 (RX 470 usually working at 750-770). Restarted the miner also.
Sapphire Nitro+ RX 580's - 6
Current Cal Hashrate (Nanopool): 6,300.0 H/s (or 1050 per card)
Claymore output: 5140 H/s (avg 856 per card)
Coreclock: 1380 (if I try for 1390 it hangs)
Memclock: 2250
Power: ~115w per card
-h: 1800 (default)
Temp: 58-60c
Drivers: Blockchain beta
Changing the -h flag in any way (other than the default of -h 1800) provides worse results or restarts. Attempting to set voltage control hangs/bootloops. I had to set my core clock lower with v10.x.
For comparison, when running 9.7 I would get would get average of 5009 H/s, ~129w /card, temp 62c with
Coreclock: 1425
Memclock: 2250
-h: 2030
Mem and core volt: 875
Power: ~129w/card
Temp: 62c
Current Cal Hashrate (Nanopool): 6,300.0 H/s (or 1050 per card)
Claymore output: 5140 H/s (avg 856 per card)
Coreclock: 1380 (if I try for 1390 it hangs)
Memclock: 2250
Power: ~115w per card
-h: 1800 (default)
Temp: 58-60c
Drivers: Blockchain beta
Changing the -h flag in any way (other than the default of -h 1800) provides worse results or restarts. Attempting to set voltage control hangs/bootloops. I had to set my core clock lower with v10.x.
For comparison, when running 9.7 I would get would get average of 5009 H/s, ~129w /card, temp 62c with
Coreclock: 1425
Memclock: 2250
-h: 2030
Mem and core volt: 875
Power: ~129w/card
Temp: 62c
New 10.2 is not good for all cards, it seems...
Just tested on my old Pitcairn R7 370 2GB
On v9.7 - 490-495 h/s
With new 10.2:
- if I start directly default -dmem tries 2.5+ GB of memory - obviously it doesn't work and miner gets stuck
- if I disable -dmem (-dmem 0), the best I can do is -h 832, but in this cases the h/s are only 475-480 (btw, Claymore - GPU0 must stay at -h 480, otherwise it crashes; but still gets these 480 h/s).
So it's not an improvement, but the opposite...
Unfortunately, good features and info of v10.2 cannot not be used... I'm sticking with 9.7...
And 10.2 became very sensitive to overclocking. On my 270X 4Gb on 9.7 i have 490-495 h/s, on 10.2 with the same frequinces and -dmem 1 I have 530 h/s and many of incorrect shares. Try to reduce OC and have only 515 h/s almost without incorrect shares.Just tested on my old Pitcairn R7 370 2GB
On v9.7 - 490-495 h/s
With new 10.2:
- if I start directly default -dmem tries 2.5+ GB of memory - obviously it doesn't work and miner gets stuck
- if I disable -dmem (-dmem 0), the best I can do is -h 832, but in this cases the h/s are only 475-480 (btw, Claymore - GPU0 must stay at -h 480, otherwise it crashes; but still gets these 480 h/s).
So it's not an improvement, but the opposite...
Unfortunately, good features and info of v10.2 cannot not be used... I'm sticking with 9.7...
If I disable -dmem 1 setting it to 0 then I recieve only 480 h/s with the same -h 512 and OC parameters that on 9.7 gives 490.
New 10.2 is not good for all cards, it seems...
Just tested on my old Pitcairn R7 370 2GB
On v9.7 - 490-495 h/s
With new 10.2:
- if I start directly default -dmem tries 2.5+ GB of memory - obviously it doesn't work and miner gets stuck
- if I disable -dmem (-dmem 0), the best I can do is -h 832, but in this cases the h/s are only 475-480 (btw, Claymore - GPU0 must stay at -h 480, otherwise it crashes; but still gets these 480 h/s).
So it's not an improvement, but the opposite...
Unfortunately, good features and info of v10.2 cannot not be used... I'm sticking with 9.7...
Just tested on my old Pitcairn R7 370 2GB
On v9.7 - 490-495 h/s
With new 10.2:
- if I start directly default -dmem tries 2.5+ GB of memory - obviously it doesn't work and miner gets stuck
- if I disable -dmem (-dmem 0), the best I can do is -h 832, but in this cases the h/s are only 475-480 (btw, Claymore - GPU0 must stay at -h 480, otherwise it crashes; but still gets these 480 h/s).
So it's not an improvement, but the opposite...
Unfortunately, good features and info of v10.2 cannot not be used... I'm sticking with 9.7...
Hey folks
here are my results
Nitro+ rx 470 4gb:
before 750 H/s ; after 790 H/s = +40/50 H/s
power draw before 650w ; after 450w (500w) = -150w
-h 900 -dmem 1
rx 470 8gb mining edition (samsung)
hashrate 845 H/s
power draw 75w
-h 1800 -dmem 1
if you need more details check here http://1stminingrig.com/claymores-cryptonote-amd-gpu-miner-v10-2-update-review/
Nice results. Does overclocking parameters work or I should use afterburner? here are my results
Nitro+ rx 470 4gb:
before 750 H/s ; after 790 H/s = +40/50 H/s
power draw before 650w ; after 450w (500w) = -150w
-h 900 -dmem 1
rx 470 8gb mining edition (samsung)
hashrate 845 H/s
power draw 75w
-h 1800 -dmem 1
if you need more details check here http://1stminingrig.com/claymores-cryptonote-amd-gpu-miner-v10-2-update-review/
if you need more details check here http://1stminingrig.com/claymores-cryptonote-amd-gpu-miner-v10-2-update-review/
You are killing noobs with your wrong environment vars setting in bat file )))
Quote
GPU_FORCE_64BIT_PTR 1
GPU_MAX_HEAP_SIZE 100
GPU_USE_SYNC_OBJECTS 1
GPU_MAX_ALLOC_PERCENT 100
GPU_SINGLE_ALLOC_PERCENT 100
GPU_MAX_HEAP_SIZE 100
GPU_USE_SYNC_OBJECTS 1
GPU_MAX_ALLOC_PERCENT 100
GPU_SINGLE_ALLOC_PERCENT 100
No linux + vega support?
Jump to: