Topic: SRBMiner Cryptonight AMD GPU Miner V1.9.3 - native algo switching - page 294. (Read 237247 times)

Hello, Arto sucessfully doing a HF to change the PoW to custom PoW (CN Lite Special Variant but with 2MB pad so it should similar to Monero V7). Our community naming this PoW "SoHeavy" even actually the PoW is "NotThatHeavy". We need more hashing power so if one of you have an idle rig you are invited to try our new PoW algo 

On SRBMiner our new algo is artocash. The official pool url: http://pool.arto.cash

Thank you,

I belive the program in Crpto Heavy is unstable, my RX 580 8gb they do 850 h/s and sometimes they do 950 h/s, always restarting program and changing intensity till gives me the 950 h/s, same options sometimes give 850 h/s, other times 950h/s. Would be nice a fix in this.

I have found that if I lower my intensity, I get higher results by the pool ? I used to be at 44i with 4.2-3kh/s and with 40i I am constantly hashing with 4.6++ and periodically going over 6kh/s. What may be causing this, doesn't sound logic?

I tried set intensity 104 or 112 but the has hrate not stable, so I go back with intensity 72. just curious with that intensity your hash rate stable?

Yes, you are right. With 122 I'm having nan H/s.
I think something is not OK with V7 using cards with more than 4 GB.
There is almost no difference of H/s setting intensity higher than 61 (The most for my RX 580 4GB where H/s is very stable).
Testing I 70-100 the H/s varies up & down, is very difficult to get the best value even with the 30 minutes average, looks like the memory is not used over 4 GB.
In Heavy there is no such problem, I can set to the max. without nan, also H/s is much higher with the 8 GB card than with the 4 one.

I have 2 rig, rig1 6k h/s and rig2 11k h/s.
I also have this kind of issue with fairpool using SRBMiner only with rig2. with cast xmr I never got this issue maybe because they have fastjobswitch.
maybe the other reason fairpool have starting difficulty very low which is 2000 because I tried other pool hashvault with starting difficulty 25000 using SRBMiner never have this kind of issue.

if you want to replicate this issue you can try fairpool with high hashrate (more than 10k maybe) in one rig, the issue should happen quite often.

thats not just on vegas, to be honest i don't know the reason 

Have you tried 1.4.9 ? If you have pool switching issues (from user to dev pool), this version fixes that.

possibly yes, i think they order by busid, i will check that, just to finish up a few things first.

Hi Dok, indeed! I am using 1.4.6 at the moment as it is the most stable version for my rigs.

yes with heavy algo

first launch with intensity 56
[2018-04-28 23:01:07] GPU0:    1359.0 H/S [T: 47c, RPM: 3333, CC: 1430 MHz, MC: 930 MHz]
[2018-04-28 23:01:07] GPU1:    1375.0 H/S [T: 43c, RPM: 3356, CC: 1440 MHz, MC: 958 MHz]
[2018-04-28 23:01:07] GPU2:    1348.0 H/S [T: 45c, RPM: 3326, CC: 1420 MHz, MC: 920 MHz]
[2018-04-28 23:01:07] GPU3:    1399.0 H/S [T: 45c, RPM: 3287, CC: 1465 MHz, MC: 970 MHz]
[2018-04-28 23:01:07] Total:   5481.0 H/S

then i relaunch the miner with intensity 60 ( a bit too high for my vegas)
[2018-04-28 23:02:55] GPU0:    1334.0 H/S [T: 49c, RPM: 3660, CC: 1430 MHz, MC: 930 MHz]
[2018-04-28 23:02:55] GPU1:    1353.0 H/S [T: 48c, RPM: 3673, CC: 1440 MHz, MC: 958 MHz]
[2018-04-28 23:02:55] GPU2:    1322.0 H/S [T: 48c, RPM: 3721, CC: 1420 MHz, MC: 920 MHz]
[2018-04-28 23:02:55] GPU3:    1372.0 H/S [T: 49c, RPM: 3646, CC: 1465 MHz, MC: 970 MHz]
[2018-04-28 23:02:55] Total:   5381.0 H/S

hashrates are lower

and then i relaunch the miner with intensity 56 and then booom to 1500
[2018-04-28 23:04:15] GPU0:    1512.0 H/S [T: 49c, RPM: 3393, CC: 1430 MHz, MC: 930 MHz]
[2018-04-28 23:04:15] GPU1:    1529.0 H/S [T: 45c, RPM: 3376, CC: 1440 MHz, MC: 958 MHz]
[2018-04-28 23:04:15] GPU2:    1497.0 H/S [T: 48c, RPM: 3390, CC: 1420 MHz, MC: 920 MHz]
[2018-04-28 23:04:15] GPU3:    1558.0 H/S [T: 49c, RPM: 3382, CC: 1465 MHz, MC: 970 MHz]
[2018-04-28 23:04:15] Total:   6096.0 H/S

doktor83 is there a way to preserve the openCL device order detection according to the HWinfo/atiflash/gpu-z order?

Hi doktor,
Your miner is great!
Higher hashrate and lower power consumption compared to cast and stak 
4 Vega 56 : 7840h/s (v7 algo), 690W at the wall , 100% accepted shares
Very stable so far (24h)
Thank you very much!

i have identical problem with vega. Yes this is only in heavy algo other algo work fine..
HBM memory not work and i must restart  push up voltage on gpu and mem  disable-enable card and work fine (1500hr)after this.

if i go minign other algo and back to haevy problem again . hr 1300hr per card. and again reset - push up voltage disablče.enable card and work fine 1500hr.

Accepted shares 99-100%...

is that on heavy algo?

hi

i have a question
when i launch srbminer after a reboot i get 1350 hs per card
and when i kill the process and relaunch it, i get 1500 - 1550 h per card
do u know why ?
the intensity is always 56
and i have reference vega 56 with stock bios

All
Did not try to calculate how much idle your devices are, trying to cope with the difficulties ....
And how many blocks do you miss ....

Well did you try other software ?
There is nothing special here, miner receives target from pool and does its hashing... it looks like their diff algo is too sophisticated
The miner itself does not 'raise difficulty', it's something that the pool dictates.

I have all my rigs working on fairpool, all working fine. What version of SRB are you using?