[ANN] TeamRedMiner - Old Thread

[pbfarmer]

Hello, does anyone know if this miner on Win10 requires a large page file like some other miners with other algos?

My page file is set on 64GB for my 4 Vega 64 rig. I have absolutely no problem with this or any other miner.

Total GPU mem size + 4GB should be more than enough for any algorithm - so 36GB should be enough for your 4 vega rig.  Really, 4GB * num GPUs + 4GB (system) should be plenty in most cases.

Vega sets core clock based on the core voltage. This is a major principle in all known reference Vegas...

...But when you set core voltage at 1.000V and clock at 1600Mhz will result only about 1300Mhz on the core because this is the clock for P3 state as default.

Core clock is adjusted based on voltage + load.  At idle, (assuming lower p states disabled,) you will see core clock at/near what you set.  While mining, it will depend on the miner load signature - some algos will tax the core more than others, causing more of a clock throttle if you're significantly undervolted.  This is all independent of other state settings (i.e. it is not related to P3.)  Also, this only happens when ACG is engaged, which *should* only be for P5,6,7.  In linux, you can step down to P4 to avoid this throttling, while in Windows, I have been unsuccessful in such attempts.

All my 4 Vegas consumes 986W of power at the wall and that include PCI-E slot power and PCI-E 2x8pin socket power. I use 2 1200W HSTNS-PD19 Platinum PSU because at that power level it is only worm and  very quiet. So 214W from HWinfo is not that inaccurate since it is not at the wall power.

I can't speak definitively to lyra2rev3 performance, as i have yet to benchmark, but with every other algo i've run, i've found Vegas to become increasingly inefficient above 1400Mhz (actual) / ~175w (at the wall.)  If you're just out for raw production, and don't care about heat or the lifespan of the GPUs, then mine away at 225w or wherever you're at.  Otherwise, you might want to check your hash/watt on some lower settings and compare (i usually find an 'actual' core speed of ~1300-1400Mhz to be a good spot for other algos.) 

[jimmyD30]

I know with Ethash on Claymore it ‘requires’ a page file at least equal to total GPU memory, so for 6 8GB Vega 64s that’s 48GB, maybe due to DAG file size.

But my question here is/was does this miner “require’ some amount of page file? Because I’m sizing my drives based on this.

I think I answered it though, basically, no. I switched over to this miner (Lyra2rev3) keeping my old ethash/claymore page file size (48GB), then switched it to system managed size (which ended up being much smaller) and have seen no issues or reduction in HR.

[pbfarmer]

I know with Ethash on Claymore it ‘requires’ a page file at least equal to total GPU memory, so for 6 8GB Vega 64s that’s 48GB, maybe due to DAG file size.

But my question here is/was does this miner “require’ some amount of page file? Because I’m sizing my drives based on this.

I think I answered it though, basically, no. I switched over to this miner (Lyra2rev3) keeping my old ethash/claymore page file size (48GB), then switched it to system managed size (which ended up being much smaller) and have seen no issues or reduction in HR.

If a miner truly requires full GPU mem page file, you would want what I wrote as the first option: total GPU mem size (all GPUs mem combined) + 4GB (for system).  Are you sure Claymore really required that tho - I'm betting that was the requirement when GPU mem was limited to 2 or 3GB - when you're getting up to 8GB, I'm not sure it's necessary anymore.   If it really is due to DAG size, then 3GB per GPU should be enough through at least mid-2019.

[jimmyD30]

Does anyone know if there is a miner watchdog script available here or are people basically writing there own?

[ku4eto]

Does anyone know if there is a miner watchdog script available here or are people basically writing there own?

People are writing their own.

[wacholek]

Core clock is adjusted based on voltage + load.  At idle, (assuming lower p states disabled,) you will see core clock at/near what you set.  While mining, it will depend on the miner load signature - some algos will tax the core more than others, causing more of a clock throttle if you're significantly undervolted.  This is all independent of other state settings (i.e. it is not related to P3.)  Also, this only happens when ACG is engaged, which *should* only be for P5,6,7.  In linux, you can step down to P4 to avoid this throttling, while in Windows, I have been unsuccessful in such attempts.

Generally I agree. However core freq throttling is also dependent what core voltage you set even if the card is in idle state.
When I was mining CNv8 ( which is not so core intensive algo) I set 1.100V because I could not get 1660Mhz on the core. Than my P7 state was set @ 1750Mhz (defult for LC). On this algo I was at 180W. On idle (no miner load) clock was as I set 1750Mhz.
On Lyra2rev3 core power is the main issue so on my settings I have 0.950V on load @ 1500Mhz and 1.000V @1600Mhz on idle. And that is with P7 set @ 1700Mmhz.
So Vega is throttling even on idle if you set very low core voltage.
I always use my Vegas with only P7 enable. I found that more stable when some network problems or pool problems occurs.

All above is more like a guide how to set your Vega, not a ready recipe.

I can't speak definitively to lyra2rev3 performance, as i have yet to benchmark, but with every other algo i've run, i've found Vegas to become increasingly inefficient above 1400Mhz (actual) / ~175w (at the wall.)  If you're just out for raw production, and don't care about heat or the lifespan of the GPUs, then mine away at 225w or wherever you're at.  Otherwise, you might want to check your hash/watt on some lower settings and compare (i usually find an 'actual' core speed of ~1300-1400Mhz to be a good spot for other algos.) 

Since the biggest degradation of silicon IC's, especially for HBM, is temperature I use water to cool down that power hungry card. I have 38*C on the core and HBM under full 218W load.
Try that on air.
Using liquid cooling I also have higher efficiency but in my case this is completely irrelevant. 

[jimmyD30]

awesome! can you disable or change the watchdog?

What are you looking to do? The watchdog checks for a number of things: shuts on/off a gpu if the temp go out of range, executes your watchdog script if a gpu dies, then some internal checks as well.

Hello, miner doesn’t seem to be executing watchdog.bat when using —watchdog_test, just notifies, is this normal behavior?

UPDATE: My watchdog.bat script works when executed by itself, but not via the miner watchdog_test.

[iRonNuke]

Hi, any plans for adding Equihash 144.5 (Zhash) to your miner? All the other miners just sucks..

Btw, anyone knows why most AMD miners just plain sucks to compete with Nvidia miners? I mean in games, AMD cards are quite good. Like a RX 580 is faster in most games than a GTX 1060. But at mining it can't compete at all. Is it much harder to code Open Cl?

[ku4eto]

awesome! can you disable or change the watchdog?

What are you looking to do? The watchdog checks for a number of things: shuts on/off a gpu if the temp go out of range, executes your watchdog script if a gpu dies, then some internal checks as well.

Hello, miner doesn’t seem to be executing watchdog.bat when using —watchdog_test, just notifies, is this normal behavior?

UPDATE: My watchdog.bat script works when executed by itself, but not via the miner watchdog_test.

Use absolute paths instead of relative. Windows may not like relative.

[jimmyD30]

awesome! can you disable or change the watchdog?

What are you looking to do? The watchdog checks for a number of things: shuts on/off a gpu if the temp go out of range, executes your watchdog script if a gpu dies, then some internal checks as well.

Hello, miner doesn’t seem to be executing watchdog.bat when using —watchdog_test, just notifies, is this normal behavior?

UPDATE: My watchdog.bat script works when executed by itself, but not via the miner watchdog_test.

Use absolute paths instead of relative. Windows may not like relative.

Hi, in the watchdog and miner startup batch files I am using absolute paths. Both scripts work fine when executed by themselves (by double clicking them), it’s just I can’t get the miner itself to execute watchdog.bat during testing and I’m assuming it won’t work while running without watchdog_test as well.

[jimmyD30]

awesome! can you disable or change the watchdog?

What are you looking to do? The watchdog checks for a number of things: shuts on/off a gpu if the temp go out of range, executes your watchdog script if a gpu dies, then some internal checks as well.

Hello, miner doesn’t seem to be executing watchdog.bat when using —watchdog_test, just notifies, is this normal behavior?

UPDATE: My watchdog.bat script works when executed by itself, but not via the miner watchdog_test.

Use absolute paths instead of relative. Windows may not like relative.

Ok, I got it working. I thought you only needed —watchdog_script parameter if using other than default script name/location, you need to add the parameter regardless.

Thank you for your suggestions tho :-)

[kerney666]

Hi, any plans for adding Equihash 144.5 (Zhash) to your miner? All the other miners just sucks..

Btw, anyone knows why most AMD miners just plain sucks to compete with Nvidia miners? I mean in games, AMD cards are quite good. Like a RX 580 is faster in most games than a GTX 1060. But at mining it can't compete at all. Is it much harder to code Open Cl?

Ha, this is a long discussion... Generalizing a little, but the Nvidia software stack (CUDA) is just a simpler and more competent tool to work with when trying to coerce the compiler to utilize the hardware the way you want. AMD's OpenCL stack(s) isn't that difficult to work with, but in many cases it's near impossible to express your intent properly, at least when we're talking about non-trivial use cases. And, the f-ing regression issues they keep having... Stuff that compiled just fine with the Blockchain driver way back when and produced really tight and nice code is completely broken in later amdcl2 drivers. The ROCm stack adds a few very useful features, but at the end of the day we revert to low level tools for most algos in this miner to produce code mapped to the hardware in a better way.

From a hardware perspective, there are scenarios where AMD GCN can truly shine, others where Nvidia wins the day. The latter case tends to be more common . The way efficiency is becoming the key metric rather than raw hashrate also plays into Nvidia's hands, my experience is that AMD tends to be a bit more power hungry regardless of use case. This is an aspect you don't care that much about in gaming.

To answer your first question: no, we're not doing any Equihash work right now. It could change in the future, but first we need to focus on the upcoming XMR fork.

[kerney666]

awesome! can you disable or change the watchdog?

What are you looking to do? The watchdog checks for a number of things: shuts on/off a gpu if the temp go out of range, executes your watchdog script if a gpu dies, then some internal checks as well.

Hello, miner doesn’t seem to be executing watchdog.bat when using —watchdog_test, just notifies, is this normal behavior?

UPDATE: My watchdog.bat script works when executed by itself, but not via the miner watchdog_test.

Use absolute paths instead of relative. Windows may not like relative.

Ok, I got it working. I thought you only needed —watchdog_script parameter if using other than default script name/location, you need to add the parameter regardless.

Thank you for your suggestions tho :-)

Glad you got it working! And yes, you always need the --watchdog_script parameter if you want a script to be executed. The --watchdog_test only triggers a fake alarm after 20-30s of mining.

[BeeDeeEff]

How goes progress on the new XMR algo for the upcoming fork?

[fluxy12]

How goes progress on the new XMR algo for the upcoming fork?

Yes any news ?

[kerney666]

How goes progress on the new XMR algo for the upcoming fork?

Yes any news ?

We're working on it, looking to be ready for the fork. The release will probably be a day before the fork or so, don't think we're ready with everything before then.

[Spoonbox]

I got some old GPUs on my rig so the miner shuts down on startup. However I would like to use this miner for Vegas on that rig so any workaround?

[UnclWish]

I got some old GPUs on my rig so the miner shuts down on startup. However I would like to use this miner for Vegas on that rig so any workaround?

Your bat-file?

[cregsanjay]

How goes progress on the new XMR algo for the upcoming fork?

Yes any news ?

We're working on it, looking to be ready for the fork. The release will probably be a day before the fork or so, don't think we're ready with everything before then.

Will the Radeon VII be supported in the new release?

[kerney666]

How goes progress on the new XMR algo for the upcoming fork?

Yes any news ?

We're working on it, looking to be ready for the fork. The release will probably be a day before the fork or so, don't think we're ready with everything before then.

Will the Radeon VII be supported in the new release?

Supporting it is trivial, I have two VIIs myself, it’s just about adding another known device name in the code. We’ll get it in there for the compute algos. For CN, we need to tune it properly first, or at least add a big fat disclaimer that the miner isn’t optimized for the gpu (yet). We’ll see where we end up.