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

[doktor83]

--gpureorder and --listdevicesreordered does not work as expected. More precisely, the first parameter works, but I get the result with the command "SRBMiner-CN.exe --listdevicesreordered", I see that I need to set "worksize = 8" for BUSID 12 and 13, which corresponds to DeviceID [7] and DeviceID [8] When setting up ' gpu_conf for IDs 7 and 8, but as a result, the values for cards with BUSID 2 and 4 change

ver 1.5.1
ver 1.5.2

Right, i will look into it. Also im thinking about dropping GPU id thing, and just use BUSid instead, it would be much cleaner, as bus id is always the same, but gpu index changes on reordering, and everything has to be reordered in back, but looks like i left out something somewhere.

[heavyarms1912]

--gpureorder and --listdevicesreordered does not work as expected. More precisely, the first parameter works, but I get the result with the command "SRBMiner-CN.exe --listdevicesreordered", I see that I need to set "worksize = 8" for BUSID 12 and 13, which corresponds to DeviceID [7] and DeviceID [8] When setting up ' gpu_conf for IDs 7 and 8, but as a result, the values for cards with BUSID 2 and 4 change

ver 1.5.1
ver 1.5.2

Right, i will look into it. Also im thinking about dropping GPU id thing, and just use BUSid instead, it would be much cleaner, as bus id is always the same, but gpu index changes on reordering, and everything has to be reordered in back, but looks like i left out something somewhere.

Yup that looks to be much cleaner approach since all other 3rd party tools also use similar approach.  it's easier to debug a faulty gpu too.

[ipe4enko]

Disable GPU by key pressed not working correctly.
When i try to disable gpu 10, 11, 12 with Shift key nothing happen, but if then i try to disable gpu 1 when all 4 card will disable

[2018-05-10 01:36:43] Stopping GPU1, please wait...
[2018-05-10 01:36:43] Stopping GPU10, please wait...
[2018-05-10 01:36:43] Stopping GPU11, please wait...
[2018-05-10 01:36:43] Stopping GPU12, please wait...

[heavyarms1912]

for some reason cryptonight heavy hashrates seems to be inconsistent on similar settings.  It goes 1530 drops to 1380 if monitor is turned off or sometimes just starts at 1380.

If we play around with intensities like increase to 60 the hashrate would be 13xx/14xx but then again when you drop it back to 56 and restart the hashrates are 1560 on Vega 56.  So weird)



Turn off display and the hashrate goes away.  back to 14xx.

[mutual.consent]

@Doktor83 - does SRB cache the GPU compile data anywhere on the hard drive to speed up the process in future? I've noticed that one GPU (Vega 64) will not perform better on Heavy no matter what I do but when testing Alloy it outperforms the other GPUs (Vega 56). My current suspicions are that there's either a bug in your implementation for Vega 64 or there's some caching happening and I need to delete these files.

Thanks.

[rusman2]

hey guys, what sort of tool do you use otherthan overdriven? all my GPUs (580,570,560) are having MUCH higher voltages than the 875-900 mv in the bios.... this is new though with 18.3.4 and latest windows 1709 (DON'T do april, as usual, microsoft is sooo stupid destroying everything).
Also, does anyone have Good Elpida (BABG) and Hynix (AJR) timings (4gb Ellsemere GPUs) other than the one click?
I have some 560s which have no more gains with clocking memory over 2000 (eplida) hash rate is maxed at 548 changing core from 1050-1200+++ doesnt do anything.

I wouldn't call them 'good' as all i did was copy 1500 strap to the higher versions. I have 2 Hynix 570 Gaming X 4GB cards that are hashing right at 700h/s (CN-heavy) using autointensity (50). 1350/1000mv and mem is 2000/850. I see little to no benefit in memory clock and only see a difference in hashrate on Core clock.
My one other card is the same but has Samsung. I am having issues with it hanging/crashing the miner itself or windows. I believe it doesn't like the memory timing so I need to change it I just haven't taken the time to do that yet and currently deal with restarting the miner every ~24 hours or so.

[rusman2]

@Doktor83 - is there any way to increase logging output? I am having a daily crash and the log is not catching anything. The log is continuing normal with no errors and the next timestamp is when I restart the miner.

[rusman2]

@Doktor83 - does SRB cache the GPU compile data anywhere on the hard drive to speed up the process in future? I've noticed that one GPU (Vega 64) will not perform better on Heavy no matter what I do but when testing Alloy it outperforms the other GPUs (Vega 56). My current suspicions are that there's either a bug in your implementation for Vega 64 or there's some caching happening and I need to delete these files.

Thanks.

I thought I saw this in the log or CLI but I don't see it on my system now. There is a C:\Users\<username>\AppData\Local\SRBMiner directory path but there was nothing, not even hidden files in there on my system.

[doktor83]

@Doktor83 - does SRB cache the GPU compile data anywhere on the hard drive to speed up the process in future? I've noticed that one GPU (Vega 64) will not perform better on Heavy no matter what I do but when testing Alloy it outperforms the other GPUs (Vega 56). My current suspicions are that there's either a bug in your implementation for Vega 64 or there's some caching happening and I need to delete these files.

Thanks.

in miner dir, delete .srb files

[doktor83]

@Doktor83 - is there any way to increase logging output? I am having a daily crash and the log is not catching anything. The log is continuing normal with no errors and the next timestamp is when I restart the miner.

if the miner crashes, then it can't write anything to the log
Which version are you using?

[mutual.consent]

@Doktor83 - does SRB cache the GPU compile data anywhere on the hard drive to speed up the process in future? I've noticed that one GPU (Vega 64) will not perform better on Heavy no matter what I do but when testing Alloy it outperforms the other GPUs (Vega 56). My current suspicions are that there's either a bug in your implementation for Vega 64 or there's some caching happening and I need to delete these files.

Thanks.

in miner dir, delete .srb files

There aren't any .srb files in the miner folder and I've searched the drive for .srb files but nothing was found. I'm using 1.5.1 and running miner from .bat files located in another directory to the miner. Anywhere else the cache files could be?

[doktor83]

@Doktor83 - does SRB cache the GPU compile data anywhere on the hard drive to speed up the process in future? I've noticed that one GPU (Vega 64) will not perform better on Heavy no matter what I do but when testing Alloy it outperforms the other GPUs (Vega 56). My current suspicions are that there's either a bug in your implementation for Vega 64 or there's some caching happening and I need to delete these files.

Thanks.

in miner dir, delete .srb files

There aren't any .srb files in the miner folder and I've searched the drive for .srb files but nothing was found. I'm using 1.5.1 and running miner from .bat files located in another directory to the miner. Anywhere else the cache files could be?

No, it creates the files in the miner dir, filenames are like:
RadeonRX580Series12256368CNH.srb

[rusman2]

if the miner crashes, then it can't write anything to the log
Which version are you using?

True! I was hoping it would log something before crash that might give me a clue as to what it was.
I'm using 1.5.1, but I see 1.5.2 was released on the 8th - is there any logging enhancement or stability enhancement worth checking out? AMD driver 18.2.1.

[doktor83]

Disable GPU by key pressed not working correctly.
When i try to disable gpu 10, 11, 12 with Shift key nothing happen, but if then i try to disable gpu 1 when all 4 card will disable

[2018-05-10 01:36:43] Stopping GPU1, please wait...
[2018-05-10 01:36:43] Stopping GPU10, please wait...
[2018-05-10 01:36:43] Stopping GPU11, please wait...
[2018-05-10 01:36:43] Stopping GPU12, please wait...

it's keyboard language and layout dependant, so 10 is really ')' , 11 is '!' etc, its for English (United States) - US keyboard so shift+1 = ! , etc..

[doktor83]

if the miner crashes, then it can't write anything to the log
Which version are you using?

True! I was hoping it would log something before crash that might give me a clue as to what it was.
I'm using 1.5.1, but I see 1.5.2 was released on the 8th - is there any logging enhancement or stability enhancement worth checking out? AMD driver 18.2.1.

How does it crash, a window opens saying the app has crashed, or ?

[rusman2]

How does it crash, a window opens saying the app has crashed, or ?

It crashes with a blue screen!
That was a joke (a bad one)

Usually happens overnight. I'll wake up to find one of two things:
1) System has restarted on its own.
2) System is hung. I can't even get the monitor to wake up (using on board video, do not have remote capability yet). However, the cards are all behaving like they are hashing in the sense that they fans are running around~2000 RPM but the pool web page says no hashes from my rig. My only option at this point is to force restart the system.

It's pretty consistent that it will do this in ~24 hour timeframe.

[doktor83]

How does it crash, a window opens saying the app has crashed, or ?

It crashes with a blue screen!
That was a joke (a bad one)

Usually happens overnight. I'll wake up to find one of two things:
1) System has restarted on its own.
2) System is hung. I can't even get the monitor to wake up (using on board video, do not have remote capability yet). However, the cards are all behaving like they are hashing in the sense that they fans are running around~2000 RPM but the pool web page says no hashes from my rig. My only option at this point is to force restart the system.

It's pretty consistent that it will do this in ~24 hour timeframe.

Have you tried other drivers , or lowering the intensity ?
There was a bug that made random app crashes, but that was fixed, that's why i asked if a crash window appears.

[hesido]

Doktor, what can you say about the random hash rate?

SRB does get the best hashrate, and I even managed 1000 mh/s for all my cards, in one run, but what happens is this:
e.g. Sometimes, GPU0 gives 920h/s, sometimes the same gpu, just by restarting miner, gives 1000 mh/s, but then some other card, e.g. GPU5 which gave 1000mh/s, may (or may not) do 920-940 mh/s.. The total may vary +/- 800 mh/s, and it doesn't change much during the run (it gets stabilized in a few minutes, stays that way)

Sometimes a gpu gets 0 hashrate, I restart, it fails on compilation phase. I use XMR-Stak to get it working back, switch back to SRB and it begins working (SRB cannot get it unstuck however many restarts!)

[livada]

i have 5 pool and when i pres p mine rgo on next pool. ok. and i must wait for pres p again and again.
any option go direct to pool 5? ctr-shift 5 etc? or pres p for elect  poll in list?

[doktor83]

Doktor, what can you say about the random hash rate?

SRB does get the best hashrate, and I even managed 1000 mh/s for all my cards, in one run, but what happens is this:
e.g. Sometimes, GPU0 gives 920h/s, sometimes the same gpu, just by restarting miner, gives 1000 mh/s, but then some other card, e.g. GPU5 which gave 1000mh/s, may (or may not) do 920-940 mh/s.. The total may vary +/- 800 mh/s, and it doesn't change much during the run (it gets stabilized in a few minutes, stays that way)

Sometimes a gpu gets 0 hashrate, I restart, it fails on compilation phase. I use XMR-Stak to get it working back, switch back to SRB and it begins working (SRB cannot get it unstuck however many restarts!)

this is true only for heavy algo, and i haven't look into it yet. What error do you get when the kernel compile fails? And what do you mean by using xmr stak to get it back working?