[ANN] TeamRedMiner v0.10.10 - Ironfish/Kaspa/ZIL/Kawpow/Etchash and More - page 137.

wyzdic

newbie

Activity: 46

Merit: 0

Quote from: SalvajeX on November 05, 2018, 10:09:30 AM

A great miner from my perspective, tried it on 6 x RX Vega 56 flashed as 64, all of them managed to reached 2000+ some 2180. which is considerable improvement over anything else.
Stable so far.

Can you provide the frequency and power of the card?

kerney666

member

Activity: 658

Merit: 86

Quote from: peterboy1 on November 05, 2018, 09:59:09 AM

ok, im about 2 days non stop now. so far so good. here are my statistics on one of my 13gpu rx570 4gb rig:

7331/7315/5

so i have 11 shares missing (as you discussed that the pool might not have responded)

5 rejects, 2 rejects in one card and 1 each random cards.

should i worry about the reject? on pool side, it didnt report any invalid shares. is my internet connection one of the factor? my internet is sometimes erratic.

edit:

i browsed the console, i found that all rejects happened all at once:
(gpu 7) error code: -1 invalid job id
(gpu 2) same
(gpu 9) same
(gpu 2) same
(gpu 5) same

Sounds like a network or pool hickup, and we tried to send old shares after reconnecting. Personally, I think 0.07% rejected shares is very much acceptable. So many things that can and will go wrong over a longer period of time. A pool restart there, some ddos attack there, a little internet glitch for the miner.

Do you see any other indications of pool reconnect etc before those errors?

kerney666

member

Activity: 658

Merit: 86

Quote from: fenomenyaa on November 05, 2018, 09:27:57 AM

Quote from: kerney666 on November 02, 2018, 12:11:08 PM

Quote from: fenomenyaa on November 02, 2018, 11:53:06 AM

Still same problem on win10 18.5.1 rx470-480 series G4400 4Gb ram.All rigs stable on srb 1.6.8-9

Gah this is annoying. It doesn't feel like the init problem we've been battling though. Do you have all the standard environment vars set when running the miner? Srb typically sets them with "setx" in the start bat file, so they should be persistent if you run srb as well. I mean these:
set GPU_MAX_HEAP_SIZE=100 set GPU_MAX_USE_SYNC_OBJECTS=1 set GPU_SINGLE_ALLOC_PERCENT=100 set GPU_MAX_ALLOC_PERCENT=100 set GPU_MAX_SINGLE_ALLOC_PERCENT=100
We should really include them in our example start bat file, that's a mistake. Last, if you run with something silly like 2+2 for all cards, does it work or not?

To speed up this process, would you be willing to hook up on Discord or Telegram? So much easier when you can ping-pong questions in realtime rather than this thread

.

"setx" change nothing .No problem on init but most of the cards not start .2+2 vs still same.i can just use your miner with vega rigs without problem, any of the polaris rigs not work.İ have to manage many rigs and not easy to find a spare time hook up telegram .

I finally have teamviewer access to a rig with 580s on a Onda B250 that has the 0 h/s problem, and at last I've confirmed the issue. I can't explain why it happens yet (because it shouldn't

), but at least I have a reproducible case.

kerney666

member

Activity: 658

Merit: 86

Quote from: ku4eto on November 05, 2018, 09:19:47 AM

For some reason, it is much LESS forgiving on GPUs. Running XMR-Stak, i get maybe 2-3 Rejected/Invalid shares per day on the crappy GPU.

On the TRM, i get on that in 12hours - 40-50.

But it actually BSODs less. Go figure.

It is indeed pushing the gpus harder in some parts of the algo. It's one of those cases where it's very hard to add performance unless you push the card harder. In other parts, we just execute fewer instructions than anyone else which leads to a lower power draw overall. We will get to this, it's connected to the power swings you can see as well. It takes a bigger rewrite to do what I think is the optimal design. It should both be more lean on the cards and maybe add a tad more perf.

Fwiw, I have a really crappy 580 in a ws, but I've gotten it stable, couldn't push memclk that much though. Do you see confirmed mem errors in hwinfo64 or some other tool?

SalvajeX

newbie

Activity: 20

Merit: 0

A great miner from my perspective, tried it on 6 x RX Vega 56 flashed as 64, all of them managed to reached 2000+ some 2180. which is considerable improvement over anything else.
Stable so far.

peterboy1

newbie

Activity: 168

Merit: 0

ok, im about 2 days non stop now. so far so good. here are my statistics on one of my 13gpu rx570 4gb rig:

7331/7315/5

so i have 11 shares missing (as you discussed that the pool might not have responded)

5 rejects, 2 rejects in one card and 1 each random cards.

should i worry about the reject? on pool side, it didnt report any invalid shares. is my internet connection one of the factor? my internet is sometimes erratic.

edit:

i browsed the console, i found that all rejects happened all at once:
(gpu 7) error code: -1 invalid job id
(gpu 2) same
(gpu 9) same
(gpu 2) same
(gpu 5) same

Mashy81

jr. member

Activity: 225

Merit: 1

4xvega56 + 2×rx570 4g
10.1khs
990watts at the wall Shocked

netmebtc

newbie

Activity: 72

Merit: 0

It is my pool need modify a parameter,so restart pool ,then some rigs which connecting the pool were crash and down. So need restart rig power ,then rigs can mining again.

Quote from: kerney666 on November 05, 2018, 08:17:58 AM

Quote from: netmebtc on November 05, 2018, 07:35:44 AM

Hello,dev. When the pool restart , My vega rigs runing teamredminer0.3.6 will crash and down, Waht is the problem?

Hi! What pool is this? And, why does the pool restart, is it connection down, block due to bad shares, anything else? Will try to reproduce and fix if you provide more info.

fenomenyaa

jr. member

Activity: 131

Merit: 2

Quote from: kerney666 on November 02, 2018, 12:11:08 PM

Quote from: fenomenyaa on November 02, 2018, 11:53:06 AM

Still same problem on win10 18.5.1 rx470-480 series G4400 4Gb ram.All rigs stable on srb 1.6.8-9

Gah this is annoying. It doesn't feel like the init problem we've been battling though. Do you have all the standard environment vars set when running the miner? Srb typically sets them with "setx" in the start bat file, so they should be persistent if you run srb as well. I mean these:
set GPU_MAX_HEAP_SIZE=100 set GPU_MAX_USE_SYNC_OBJECTS=1 set GPU_SINGLE_ALLOC_PERCENT=100 set GPU_MAX_ALLOC_PERCENT=100 set GPU_MAX_SINGLE_ALLOC_PERCENT=100
We should really include them in our example start bat file, that's a mistake. Last, if you run with something silly like 2+2 for all cards, does it work or not?

To speed up this process, would you be willing to hook up on Discord or Telegram? So much easier when you can ping-pong questions in realtime rather than this thread

.

"setx" change nothing .No problem on init but most of the cards not start .2+2 vs still same.i can just use your miner with vega rigs without problem, any of the polaris rigs not work.İ have to manage many rigs and not easy to find a spare time hook up telegram .

ku4eto

jr. member

Activity: 194

Merit: 4

For some reason, it is much LESS forgiving on GPUs. Running XMR-Stak, i get maybe 2-3 Rejected/Invalid shares per day on the crappy GPU.

On the TRM, i get on that in 12hours - 40-50.

But it actually BSODs less. Go figure.

ArtSmart

newbie

Activity: 12

Merit: 0

kerney666 any solution to my issue?

Mashy81

jr. member

Activity: 225

Merit: 1

Damn WTF!!!

How you do dis.
I am getting better than v7 hashrates and low power.
Super easy to use to. Thanks guys

kerney666

member

Activity: 658

Merit: 86

Quote from: netmebtc on November 05, 2018, 07:35:44 AM

Hello,dev. When the pool restart , My vega rigs runing teamredminer0.3.6 will crash and down, Waht is the problem?

Hi! What pool is this? And, why does the pool restart, is it connection down, block due to bad shares, anything else? Will try to reproduce and fix if you provide more info.

netmebtc

newbie

Activity: 72

Merit: 0

Hello,dev. When the pool restart , My vega rigs runing teamredminer0.3.6 will crash and down, Waht is the problem?

ArtSmart

newbie

Activity: 12

Merit: 0

Tahnk You. It was running perfectly on 18.8.1 drivers with the same clocks/voltages. But only 6 GPUs were seen. So I don't think that it is clocks issue.

Quote from: todxx on November 05, 2018, 05:38:06 AM

The miner is successfully finding and running all 8 GPUs in that screen shot. As for the error, the 'Failed to enqueue kernel' error means that something has gone very, very wrong with the driver, possibly due to cards crashing.
What clocks/voltages are you running on those cards?

netmebtc

newbie

Activity: 72

Merit: 0

Solve it, 6xVega64 get 12.5Kh @1280W

Quote from: vmozara on November 05, 2018, 02:44:45 AM

netmebtc,
since you mentioned power target -20%, i assume you have wattman installed.
You need to reinstall the driver without wattman as very often it will conflict with miner, drop the hashrate or cause instability. Use DDU to clean up your driver, then install adrenalin driver ONLY, no amd settings, no amd problem reporting, no nothing. (just power tables because they get deleted when driver is reinstalled)
Finally, use video card restart sequence to restart your cards, apply the ODN settings, and start the miner. This will have your vegas hashing at top speed. If you don't have this let us know

todxx

member

Activity: 176

Merit: 76

Quote from: ArtSmart on November 05, 2018, 02:57:57 AM

Hi, thanks for good product. Had newer drivers, but in this setup only 6GPU from 8GPUS were recognised. Installed 18.6.1 drivers - now get this error. Please help

The miner is successfully finding and running all 8 GPUs in that screen shot. As for the error, the 'Failed to enqueue kernel' error means that something has gone very, very wrong with the driver, possibly due to cards crashing.
What clocks/voltages are you running on those cards?

keksik

jr. member

Activity: 169

Merit: 1

Quote from: heavyarms1912 on November 04, 2018, 08:58:13 PM

Micron 8gbs can do 1k+.
Elipdas 4gb can do 1k+
Hynix 4gb AJR can do 1k+

All of these at 1250 core or lower.

everyone saying as it is possible, no one share the straps. we understand that it is possible for every card to do 1k+, but im not able to reach it under my current condition. i have all set and tuned, im mising good strap for my cards. are u able to share it with me please?

netmebtc

newbie

Activity: 72

Merit: 0

@vmozara @todxx,thanks you two guys, I'll try do as you say. Then return the result.

ArtSmart

newbie

Activity: 12

Merit: 0

Hi, thanks for good product. Had newer drivers, but in this setup only 6GPU from 8GPUS were recognised. Installed 18.6.1 drivers - now get this error. Please help

https://image.ibb.co/fSQoQ0/Screen-Shot-2018-11-05-at-09-39-52.png

Topic: [ANN] TeamRedMiner v0.10.10 - Ironfish/Kaspa/ZIL/Kawpow/Etchash and More - page 137. (Read 211877 times)