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

[fenomenyaa]

Whenever the vards initialize or not, the miner works super nice from what I have seen so far. I am busy all day optimizing my rigs for this miner. If some cards don't initialize, miner still works, just displays 0h/s for that card, and CPU usage is high, so he is probably stuck trying to initialize the cards or something.
Then, i just quit the miner and try few times and eventually all cards initialize.

Yes, me too.

For vmozara's problem, I'm 97% certain this issue would be solved with a staggered init. These Celerons are just too weak for a parallel init for > 2-3 GPUs or something. I'll add some more options now to control the behavior at startup, then send you a PM. Need to wait for todxx to provide a new build, but it would be nice to connect and do a few verification runs instead of just pushing a new version out there not knowing if it actually solves the problem.

@wyzdic, are you having the same issue as vmozara? It feels like we might have a different problem, do you get a random nr of gpus going or always the first six? Furthermore, I've heard about people that have issues in general getting > 6 Vegas going under Linux, are you on Win or some Linux version?

İ have the same problem 8card polaris systems(many) on win10 1709/Blockchain driver.G4560 -4Gb ram 32GB vm
Random number of gpus's not going,sometimes gpu0 sometimes gpu 3 or others not working, not always same gpu.

[wyzdic]

Whenever the vards initialize or not, the miner works super nice from what I have seen so far. I am busy all day optimizing my rigs for this miner. If some cards don't initialize, miner still works, just displays 0h/s for that card, and CPU usage is high, so he is probably stuck trying to initialize the cards or something.
Then, i just quit the miner and try few times and eventually all cards initialize.

Yes, me too.

For vmozara's problem, I'm 97% certain this issue would be solved with a staggered init. These Celerons are just too weak for a parallel init for > 2-3 GPUs or something. I'll add some more options now to control the behavior at startup, then send you a PM. Need to wait for todxx to provide a new build, but it would be nice to connect and do a few verification runs instead of just pushing a new version out there not knowing if it actually solves the problem.

@wyzdic, are you having the same issue as vmozara? It feels like we might have a different problem, do you get a random nr of gpus going or always the first six? Furthermore, I've heard about people that have issues in general getting > 6 Vegas going under Linux, are you on Win or some Linux version?

Win10 LTSB. 18.6.1, 8x Vega56.
It's not random, it's GPU 6 and GPU7 every time, and it doesn't show GPU6 and GPU7 at startup, it can be seen in the thread. Using -d 0,1,2,3,4,5,6,7, the problem remains

Code:

[2018-10-31 08:00:45] Successfully initialized GPU 0: Vega with 56 CU (PCIe 16:00.0) (CN 16+14)
[2018-10-31 08:00:45] Successfully initialized GPU 1: Vega with 56 CU (PCIe 10:00.0) (CN 16+14)
[2018-10-31 08:00:46] Successfully initialized GPU 2: Vega with 56 CU (PCIe 03:00.0) (CN 16+14)
[2018-10-31 08:00:47] Successfully initialized GPU 3: Vega with 56 CU (PCIe 19:00.0) (CN 16+14)
[2018-10-31 08:00:48] Successfully initialized GPU 4: Vega with 56 CU (PCIe 07:00.0) (CN 16+14)
[2018-10-31 08:00:49] Successfully initialized GPU 5: Vega with 56 CU (PCIe 22:00.0) (CN 16+14)
[2018-10-31 08:00:49] API initialized on 127.0.0.1:9755
[2018-10-31 08:00:49] Pool xxxxx login succeeded.
[2018-10-31 08:00:49] Pool xxxxx received new job. (job_id: 979590065579534250)
[2018-10-31 08:00:49] Pool xxxxx set difficulty to 600023
[2018-10-31 08:00:51] Dev pool connected and ready.
[2018-10-31 08:01:08] Pool xxxxx received new job. (job_id: 173340078480066250)
[2018-10-31 08:01:19] Stats GPU 0 - cnv8: 1.756kh/s, avg 1.499kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 1 - cnv8: 1.668kh/s, avg 1.424kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 2 - cnv8: 1.596kh/s, avg 1.405kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 3 - cnv8: 1.597kh/s, avg 1.445kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 4 - cnv8: 1.651kh/s, avg 1.543kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 5 - cnv8: 1.793kh/s, avg 1.722kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 6 - cnv8:   0.0 h/s, avg   0.0 h/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 7 - cnv8:   0.0 h/s, avg   0.0 h/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats Total - cnv8: 10.06kh/s, avg 9.037kh/s, pool   0.0 h/s

[kerney666]

Win10 LTSB. 18.6.1, 8x Vega56.
It's not random, it's GPU 6 and GPU7 every time, and it doesn't show GPU6 and GPU7 at startup, it can be seen in the thread. Using -d 0,1,2,3,4,5,6,7, the problem remains

If you start with -d 2,3,4,5,6,7, then -d 2,3,4,5,6,7,0,1, what happens? I'm curious if it's always the last two cards in the provided device order that are failing or if it's always your GPU6/7 in the OpenCL enumeration...

[wyzdic]

Win10 LTSB. 18.6.1, 8x Vega56.
It's not random, it's GPU 6 and GPU7 every time, and it doesn't show GPU6 and GPU7 at startup, it can be seen in the thread. Using -d 0,1,2,3,4,5,6,7, the problem remains

If you start with -d 2,3,4,5,6,7, then -d 2,3,4,5,6,7,0,1, what happens? I'm curious if it's always the last two cards in the provided device order that are failing or if it's always your GPU6/7 in the OpenCL enumeration...

I restarted the miner 20+ times. Now GPU6 starts working, but GPU7 still doesn't work. I think this is because of the CPU problem?

ps:GPU6 doesn't work when I turn off the miner to reopen

ps2:I tested using -d 0,1,2,3,4, gpu3 and gpu4 didn't work.

[kerney666]

Win10 LTSB. 18.6.1, 8x Vega56.
It's not random, it's GPU 6 and GPU7 every time, and it doesn't show GPU6 and GPU7 at startup, it can be seen in the thread. Using -d 0,1,2,3,4,5,6,7, the problem remains

If you start with -d 2,3,4,5,6,7, then -d 2,3,4,5,6,7,0,1, what happens? I'm curious if it's always the last two cards in the provided device order that are failing or if it's always your GPU6/7 in the OpenCL enumeration...

I restarted the miner 20+ times. Now GPU6 starts working, but GPU7 still doesn't work. I think this is because of the CPU problem?

ps:GPU6 doesn't work when I turn off the miner to reopen

ps2:I tested using -d 0,1,2,3,4, gpu3 and gpu4 didn't work.

Wow, that is some weird stuff. If you run "-d 0,1", do you get any mining at all? . Maybe we should wait until we have the staggered and delayed init in place, then just test again? That should be ready tonight. If that doesn't resolve your problem, I'd love to dig deeper into your specific issue though.

[VictorGT]

I am getting "pool login failed" messages. Does the miner support address.paymentID? Using  supportXMR.com

Same settings working OK with other miner  .

[kerney666]

I am getting "pool login failed" messages. Does the miner support address.paymentID? Using  supportXMR.com

Same settings working OK with other miner  .

Nope, we're missing that feature in our stratum implementation, will be added shortly, stay tuned!

[geck]

hi,

im not able to get it to work with my vega64 or 8x vega56.
on both rigs i get the same error message “Failed to initialize device number 0.”

ive tried using -d 1, but then the error becomes device 1 is out of range.

any ideas on what to try next?

edit forgot to mention im on win 10 and 18.6.1 drivers

Have you tried this ?

--platform=1

hi,
yes, i have tried that. Did not work either.

turn off hbcc... it's probably recognizing it has gfx901.

Just checked that, hbcc is turned off. SRBminer detects it as gfx900.

[Bepson]

On win7x64 rx550-2 could not start. Always crash driver. Does the miner work in win 7?

[FablomsFDP]

My indicators (Vega 64 Sapphire Nitro +) 16 + 14
Stats GPU 0 - cnv8: 2.338kh / s, avg 2.335kh / s, pool 2.640kh / s
There is not enough card temperature indicator, otherwise everything is super !!! Left for a day to test.

[Grumo]

hi,

im not able to get it to work with my vega64 or 8x vega56.
on both rigs i get the same error message “Failed to initialize device number 0.”

ive tried using -d 1, but then the error becomes device 1 is out of range.

any ideas on what to try next?

edit forgot to mention im on win 10 and 18.6.1 drivers

Have you tried this ?

--platform=1

hi,
yes, i have tried that. Did not work either.

turn off hbcc... it's probably recognizing it has gfx901.

Just checked that, hbcc is turned off. SRBminer detects it as gfx900.

same problem here

[kerney666]

hi,

im not able to get it to work with my vega64 or 8x vega56.
on both rigs i get the same error message “Failed to initialize device number 0.”

ive tried using -d 1, but then the error becomes device 1 is out of range.

any ideas on what to try next?

edit forgot to mention im on win 10 and 18.6.1 drivers

Have you tried this ?

--platform=1

hi,
yes, i have tried that. Did not work either.

turn off hbcc... it's probably recognizing it has gfx901.

Just checked that, hbcc is turned off. SRBminer detects it as gfx900.

same problem here

Can you post the output from the command “clinfo”? You can run it from the command prompt, it’ll be a long text dump. Also, include your command line and the output from the miner when you run it. Think we’ll be able to nail down this one. Everything points to a platform with a single device being tried, doesn’t match your rig description.

And, we desperately need to add automatic platform id detection...

[5BTC]

I can’t get more than 5 of my VEGA 56 to hash at the same time, currently running rigs with 8 x Vega 56

Celeron processor
Ubuntu mate 18.04
Amdgpupro 18.30

The miner initializes all 8 cards, all 8 cards show full activity on LEDs, but always 0 hashrate reported for 3 of them, only 5 ever have a hash rate above 0

Any ideas?

[kerney666]

I can’t get more than 5 of my VEGA 56 to hash at the same time, currently running rigs with 8 x Vega 56

Celeron processor
Ubuntu mate 18.04
Amdgpupro 18.30

The miner initializes all 8 cards, all 8 cards show full activity on LEDs, but always 0 hashrate reported for 3 of them, only 5 ever have a hash rate above 0

Any ideas?

Well, there is some issue here that multiple users have been seeing. Not at all certain it's the reason, but Celeron CPUs and many gpus seem to be the common data point so far. On startup, can you confirm you see the output line "... Successfully initialized GPU X" for all gpus, or is it only for five of them, i.e. the ones that then hash successfully?

[heavyarms1912]

I can’t get more than 5 of my VEGA 56 to hash at the same time, currently running rigs with 8 x Vega 56

Celeron processor
Ubuntu mate 18.04
Amdgpupro 18.30

The miner initializes all 8 cards, all 8 cards show full activity on LEDs, but always 0 hashrate reported for 3 of them, only 5 ever have a hash rate above 0

Any ideas?

Well, there is some issue here that multiple users have been seeing. Not at all certain it's the reason, but Celeron CPUs and many gpus seem to be the common data point so far. On startup, can you confirm you see the output line "... Successfully initialized GPU X" for all gpus, or is it only for five of them, i.e. the ones that then hash successfully?

celeron processor.  9 gpu rig works fine.  3 x vega 56 + 6 x RX570s.  Win 10. amd 18.6.1 drivers

[alucard20724]

@kerney666 very good job  
Just tested under Hive OS (drivers 18.10) on rx 480 4gb (elpida) 1250/950 1950/950 hashrate is 990 - 1000 depend on the card.
Do you plan to add other CN variants like arto, stellite and others? If yes could you add algo switching like SRBminer?

This was a big investment time-wise, we had to build all the support for CN stratum mining etc, so it’d be very unwise for us to not add more variants . We might also add an algo switching scheme, but haven’t decided yet.

my recommendation is for cn heavy.   there are plenty of algo switching software out there already, so that functionality doesn't really add much value for the miner since the important thing is hashrate. imho 

[wyzdic]

I can’t get more than 5 of my VEGA 56 to hash at the same time, currently running rigs with 8 x Vega 56

Celeron processor
Ubuntu mate 18.04
Amdgpupro 18.30

The miner initializes all 8 cards, all 8 cards show full activity on LEDs, but always 0 hashrate reported for 3 of them, only 5 ever have a hash rate above 0

Any ideas?

Well, there is some issue here that multiple users have been seeing. Not at all certain it's the reason, but Celeron CPUs and many gpus seem to be the common data point so far. On startup, can you confirm you see the output line "... Successfully initialized GPU X" for all gpus, or is it only for five of them, i.e. the ones that then hash successfully?

Only 6 for "... Successfully initialized GPU X".
Like this.

Code:

[2018-10-31 08:00:45] Successfully initialized GPU 0: Vega with 56 CU (PCIe 16:00.0) (CN 16+14)
[2018-10-31 08:00:45] Successfully initialized GPU 1: Vega with 56 CU (PCIe 10:00.0) (CN 16+14)
[2018-10-31 08:00:46] Successfully initialized GPU 2: Vega with 56 CU (PCIe 03:00.0) (CN 16+14)
[2018-10-31 08:00:47] Successfully initialized GPU 3: Vega with 56 CU (PCIe 19:00.0) (CN 16+14)
[2018-10-31 08:00:48] Successfully initialized GPU 4: Vega with 56 CU (PCIe 07:00.0) (CN 16+14)
[2018-10-31 08:00:49] Successfully initialized GPU 5: Vega with 56 CU (PCIe 22:00.0) (CN 16+14)
[2018-10-31 08:00:49] API initialized on 127.0.0.1:9755
[2018-10-31 08:00:49] Pool xxxxx login succeeded.
[2018-10-31 08:00:49] Pool xxxxx received new job. (job_id: 979590065579534250)
[2018-10-31 08:00:49] Pool xxxxx set difficulty to 600023
[2018-10-31 08:00:51] Dev pool connected and ready.
[2018-10-31 08:01:08] Pool xxxxx received new job. (job_id: 173340078480066250)
[2018-10-31 08:01:19] Stats GPU 0 - cnv8: 1.756kh/s, avg 1.499kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 1 - cnv8: 1.668kh/s, avg 1.424kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 2 - cnv8: 1.596kh/s, avg 1.405kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 3 - cnv8: 1.597kh/s, avg 1.445kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 4 - cnv8: 1.651kh/s, avg 1.543kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 5 - cnv8: 1.793kh/s, avg 1.722kh/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 6 - cnv8:   0.0 h/s, avg   0.0 h/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats GPU 7 - cnv8:   0.0 h/s, avg   0.0 h/s, pool   0.0 h/s 0/0
[2018-10-31 08:01:19] Stats Total - cnv8: 10.06kh/s, avg 9.037kh/s, pool   0.0 h/s

[kerney666]

On win7x64 rx550-2 could not start. Always crash driver. Does the miner work in win 7?

Sorry Bepson, missed this earlier today. We have not tested the miner on Win 7 unfortunately. I have an Rx550-2GB waiting for me, I’ll start playing around with it tomorrow.

It might be that the miner tries to push the intensities too hard. Can you try using —cn_config=2+2 as a very conservative test setting?

[kerney666]

Only 6 for "... Successfully initialized GPU X".

That’s actually a good thing from my perspective, it points to the init phase not working properly, not the core mining process.

I have a build available very shortly. Are you available for a quick test? I’d love to see if this solves your issue before we push it as an official release.

Anyone else available right now that has had the some-subset-of-my-gpus-don’t-work issue and are willing to test a patch, reach out to me over PM or find me on Discord (Zcoin/fpga/mineority/lux/cryptobridge are some servers I’m available on).

[dickmo]

          Team Red Miner version 0.3.4
[2018-10-30 15:25:55] Pool gin.f2pool.com successfully subscribed.
[2018-10-30 15:25:55] Pool gin.f2pool.com authorization succeeded.
[2018-10-30 15:25:55] Pool gin.f2pool.com set difficulty to 15.999756
[2018-10-30 15:25:55] Pool gin.f2pool.com received new job. (job_id: aKXuutTIJS)
[2018-10-30 15:25:55] Successfully initialized GPU 0: Vega with 56 CU (PCIe 10:00.0)
[2018-10-30 15:25:55] Successfully initialized GPU 1: Vega with 56 CU (PCIe 0d:00.0)
[2018-10-30 15:25:55] Dev pool connected and ready.
[2018-10-30 15:25:55] Successfully initialized GPU 2: Vega with 56 CU (PCIe 09:00.0)
Segmentation fault


i just upgrade amd-gpupro driver from 18.10 to 18.30 on hiveos(ubuntu 16.04.5),then start like above

Well, that is annoying, hmm. Would it be possible for you to check if this is a regression bug in this new version, i.e. since you run lyra2z, if you download and run an older version of the miner on the 18.30 driver, do you still get the same seg fault?

run 0.33 and get the same fault，then i tried cast1.5，it works normal