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

[nordmann666]

today...without changing ynthing i got message EXCEPTION 90012. FAULTY HDMI DEVICE INT3H.

what is it?

switched to JCE immediately to check if its HW related...but the miner works :/

It is part of the anti debugger stuff, you installed/executed an app that the miner does not like to be run with at the same time...

i copied sysinternals on the device for getting remote access to other device...after reboot it´s still problem...i dont start anything else...is there a way to find the process?!

ok..installed new awesome miner agent...maybe thats the problem..but on other rig i also installed the updated agent and i can still use srb

It is not the awesome miner, it is because of sysinternals. There are some intersting tools in there..
I will make changes in next version so it won't interfere.

oh...ok...so i have to deinstall that sxxt, cause my rig now is haning every 10-20 minutes and i cant get on the rig with TV or RDP and also with psexec or just remote powershell - i can ping the rig...i can connect to explorer stuff and awesome miner also see the rig as stopped...but the rig did not recognized any remote command...i have to ask the technician to hard reset the rig and than i can use it again...for 10 minutes (was good for weeks now)

[NCarter84]

Seems like 8gb cards always fall short of 4gb cards except for Heavy type algos... Roughly 300hs short on some algos...

I've cranked the intensity up until it crashes... What am I missing?

[nordmann666]

today...without changing ynthing i got message EXCEPTION 90012. FAULTY HDMI DEVICE INT3H.

what is it?

switched to JCE immediately to check if its HW related...but the miner works :/

It is part of the anti debugger stuff, you installed/executed an app that the miner does not like to be run with at the same time...

i copied sysinternals on the device for getting remote access to other device...after reboot it´s still problem...i dont start anything else...is there a way to find the process?!

ok..installed new awesome miner agent...maybe thats the problem..but on other rig i also installed the updated agent and i can still use srb

It is not the awesome miner, it is because of sysinternals. There are some intersting tools in there..
I will make changes in next version so it won't interfere.

oh...ok...so i have to deinstall that sxxt, cause my rig now is haning every 10-20 minutes and i cant get on the rig with TV or RDP and also with psexec or just remote powershell - i can ping the rig...i can connect to explorer stuff and awesome miner also see the rig as stopped...but the rig did not recognized any remote command...i have to ask the technician to hard reset the rig and than i can use it again...for 10 minutes (was good for weeks now)

after i kicked out sysinternals...i spend last hours to "debug" - but every time...srbminer-cn.exe is the only task i cant kill remote...the whole system works...but i cant connect to it - srbminer-cn.exe is running cause i cant rename the file with using explorer and administrative permissions

[zombezi]

Hi doctor,
Please add new algo by forked XCash new POW!
Here is the github page:

https://github.com/X-CASH-official/X-CASH/releases/tag/1.5.0
https://github.com/X-CASH-official/XCASH_ALL_Miner

I will look into it, priority now is wownero/monero
[/quote]

Please, add the "effort"

[thenextdinosaur]

I have a MSI Vega 56 (Hynix) and this card is crashing the whole 6x Vega rig (Windows 10) when mining CryptoNight-GPU. It seems like this card is getting strangely hotter than all others when mining this algorithm as well. Usually it's the colder card of the rig.

P7 is at 1408;950
Mem3 is at 890;945

[livada]

I have a MSI Vega 56 (Hynix) and this card is crashing the whole 6x Vega rig (Windows 10) when mining CryptoNight-GPU. It seems like this card is getting strangely hotter than all others when mining this algorithm as well. Usually it's the colder card of the rig.

P7 is at 1408;950
Mem3 is at 890;945

what suipšly you have?

for GPU algo u must have 1800+W power
try with this card and other 4 card- disabe 1 card and try again

i use only 880/900mv for gpu and mem

[rdome]

This cards have all bad coolers. Look
https://www.reddit.com/r/Amd/comments/9eqhnc/my_vega_64_le_was_having_horrific_throttling/

On 4 my vegas i changed cooler to watercooler because of this, 3*56 and one limited edition 64.
Solder was all over the place when i opened them up. All survived, mounted cpu water coolers on them and on one mounted accelero 3 air cooler. All coolers performed much better then original vapor chamber.
The card was set to max 65c and 4900rpm and original coolers still has melted. So original coolers are bad.
Even the 120mm radiator water cpu cooler is beter then vapor chamber (much lower temps). For my 4th card i intend to buy 240mm aio cpu cooler to test it out

And all my cards are set to.
P7 is at 1408:900
Mem3 is at 1100:880
Samsung memory

[thenextdinosaur]

what suipšly you have?

for GPU algo u must have 1800+W power
try with this card and other 4 card- disabe 1 card and try again

i use only 880/900mv for gpu and mem

I have 2x 1000W so 2000W. I don't think that power is a problem.

[heavyarms1912]

I have a MSI Vega 56 (Hynix) and this card is crashing the whole 6x Vega rig (Windows 10) when mining CryptoNight-GPU. It seems like this card is getting strangely hotter than all others when mining this algorithm as well. Usually it's the colder card of the rig.

P7 is at 1408;950
Mem3 is at 890;945

cn-gpu is core heavy.  maybe you were mining coins that weren't core heavy before?
Check the fan control on this particular one, check positioning of this card.

[rdome]

what suipšly you have?

for GPU algo u must have 1800+W power
try with this card and other 4 card- disabe 1 card and try again

i use only 880/900mv for gpu and mem

I have 2x 1000W so 2000W. I don't think that power is a problem.

Just use the flashlight, and point it to hdmi conectors and look if the cooler has gone bad. Just like i sad i have sapphire, xfx, and powercolor cards an on 4 out of 12 coolers are melted. When this happens somethimes it hashes less, somethimes it crashes rig. If you put your hand over the card it looks cooler then the rest because ic cooles with half or less cooler that is still soldered to copper base. So air seems cooler but when you look in software card hashes less and trotles.
Good luck

[morinfen]

what hashrate is given by card 552 on an algorithm cn-gpu?

my hashrate - 182, it seems as there has to be more

[sharix80]

what hashrate is given by card 552 on an algorithm cn-gpu?

my hashrate - 182, it seems as there has to be more

Sorry, but RX500 very weakly for cn-gpu. This algo need more powerful gpu chip, but no need memory. Memory can be reduced to 300 megahertz without loss of performance.

[thenextdinosaur]

cn-gpu is core heavy.  maybe you were mining coins that weren't core heavy before?
Check the fan control on this particular one, check positioning of this card.

I'm mining on MoneroOcean, so I was mining whatever they had before there. Never had an issue. HBM temperature never goes above 75°C (normal-v8). You can check this image to see the current temperatures (and maximums). The maximum fan speed is usually when the miner restarts due to algorithm change. I don't know why the fans for all cards go crazy when it does the Vega resetting, but they do.

[abg00]

hi dok!
heavyx ready? if bata availabre ready for beta test ))

[heavyarms1912]

cn-gpu is core heavy.  maybe you were mining coins that weren't core heavy before?
Check the fan control on this particular one, check positioning of this card.

I'm mining on MoneroOcean, so I was mining whatever they had before there. Never had an issue. HBM temperature never goes above 75°C (normal-v8). You can check this image to see the current temperatures (and maximums). The maximum fan speed is usually when the miner restarts due to algorithm change. I don't know why the fans for all cards go crazy when it does the Vega resetting, but they do.

cn-gpu is different and doesn't care about the memory speed.  in fact you can drop your hbm to 800 Mhz and still get the same speed.  When i say temp it's basically core temp.

[rednoW]

forget about cn-gpu crap. It is much better on nvidia. And will be easy for fpga/asic if have any volume

[thenextdinosaur]

forget about cn-gpu crap. It is much better on nvidia. And will be easy for fpga/asic if have any volume

I agree and tend to believe in the same thing. I removed CN-GPU from my list.
About the temperatures, what other AMD Vega miners are experiencing? Sometimes I think HBM at 75°C might be too hot, although it doesn't throttle at all.

[heavyarms1912]

forget about cn-gpu crap. It is much better on nvidia. And will be easy for fpga/asic if have any volume

Yup. it's much better on Nvidia but floating point math is not efficient on FPGAs and ASICs.

[sgnl.dsrptr]

When is cn/wow implementation expected to be done?

[maria3131]

hi dok!
heavyx ready?