Number 9! Ninth altcoin thread. Back to the moon Baby!

[bmoscato]

Sometimes they get really stable.

My 2 card hp omen  rigs are very good.  1080ti's about 2 2/3 years old. they run perfectly 50 days in a row no issues

the amd are flakey

A) my service may have an issue
B) smos servers may have an issue
C) nicehash may have an issue

But they are easy peasy to run and I like the direct conversions of eth to btc.

Do your AMD cards get a lot of restarts in SMOS against Nice Hash? I'm at 17 restarts today using SMOS with Phoenix, I can't get TeamRed Miner to work against NiceHash (using the proper BTC address), but it works great against Ethereum with ETH address.

[philipma1957]

Do you guys get a lot of rejects on NiceHash?  Are your rigs as stable as mining straight eth?

Yes, I only get errors on Nice Hash, I can run against Ethermine for days with no issues.

your gear—— nicehash—— a renter——— a pool


your gear—- a pool

much less connections to hand off.

also the renter can point to a shit coin and switch all over.

so it could be six connections or more vs two connections.

[alucard20724]

Do your AMD cards get a lot of restarts in SMOS against Nice Hash? I'm at 17 restarts today using SMOS with Phoenix, I can't get TeamRed Miner to work against NiceHash (using the proper BTC address), but it works great against Ethereum with ETH address.

Here's what I use for nicehash with my BTC address with TRM version 0.7.15 with SMOS.
-a ethash -o stratum+tcp://daggerhashimoto.usa.nicehash.com:3353 -u btcaddresshere.$rigName -p x --api_listen --watchdog_script.

I have three systems using trm on smos. these three systems work better with trm than they do with phoenix.  I have a fourth system that works better with phoenix than it does with trm.

[bmoscato]

Here's what I use for nicehash with my BTC address with TRM version 0.7.15 with SMOS.
-a ethash -o stratum+tcp://daggerhashimoto.usa.nicehash.com:3353 -u btcaddresshere.$rigName -p x --api_listen --watchdog_script.

I have three systems using trm on smos. these three systems work better with trm than they do with phoenix.  I have a fourth system that works better with phoenix than it does with trm.

That definitely caused some issues after running about a minute, I guess NiceHash isn't in my future until they come out with the newer NHOS.

[2020-11-03 16:16:00]
[2020-11-03 16:16:00] Stats Uptime: 0 days, 00:01:32
[2020-11-03 16:16:00] ----------------------------------------- GPU Status -------------------------------------------
[2020-11-03 16:16:00] GPU 0 [50C, fan 24%] ethash: 14.26Mh/s, avg 30.84Mh/s, pool 58.19Mh/s a:2 r:0 hw:0
[2020-11-03 16:16:00] GPU 1 [36C, fan 64%] ethash: 36.02kh/s, avg 25.74Mh/s, pool 29.16Mh/s a:1 r:0 hw:0
[2020-11-03 16:16:00] GPU 2 [34C, fan 29%] ethash: 36.21kh/s, avg 25.87Mh/s, pool 29.23Mh/s a:1 r:0 hw:0
[2020-11-03 16:16:00] Total ethash: 14.33Mh/s, avg 82.45Mh/s, pool 116.6Mh/s a:4 r:0 hw:0
[2020-11-03 16:16:00] ----------------------------------------- Pool Status ------------------------------------------
[2020-11-03 16:16:00] daggerhashimoto.usa.nicehash.com ethash: 14.32Mh/s, avg 80.64Mh/s, pool 116.4Mh/s a:4 r:0
[2020-11-03 16:16:00] ------------------------------------------------------------------------------------------------
[2020-11-03 16:16:10] GPU 0 Ramp-up completed.
[2020-11-03 16:16:13] Pool daggerhashimoto.usa.nicehash.com share accepted. (GPU0) (a:5 r:0) (104 ms) (diff 5.61 GH)
[2020-11-03 16:16:17] Pool daggerhashimoto.usa.nicehash.com received new job. (job_id: 00000000b9a70a04)
[2020-11-03 16:16:20] GPU 1: detected DEAD (07:00.0), will execute restart script watchdog.sh
[2020-11-03 16:16:20] GPU 2: detected DEAD (0a:00.0), will execute restart script watchdog.sh
[2020-11-03 16:16:20] Watchdog script executor thread executing script 'watchdog.sh'
Rebooting
(by mining software). Check crash logs.

[BogdanCo]

Here's what I use for nicehash with my BTC address with TRM version 0.7.15 with SMOS.
-a ethash -o stratum+tcp://daggerhashimoto.usa.nicehash.com:3353 -u btcaddresshere.$rigName -p x --api_listen --watchdog_script.

I have three systems using trm on smos. these three systems work better with trm than they do with phoenix.  I have a fourth system that works better with phoenix than it does with trm.

That definitely caused some issues after running about a minute, I guess NiceHash isn't in my future until they come out with the newer NHOS.

[2020-11-03 16:16:00]
[2020-11-03 16:16:00] Stats Uptime: 0 days, 00:01:32
[2020-11-03 16:16:00] ----------------------------------------- GPU Status -------------------------------------------
[2020-11-03 16:16:00] GPU 0 [50C, fan 24%] ethash: 14.26Mh/s, avg 30.84Mh/s, pool 58.19Mh/s a:2 r:0 hw:0
[2020-11-03 16:16:00] GPU 1 [36C, fan 64%] ethash: 36.02kh/s, avg 25.74Mh/s, pool 29.16Mh/s a:1 r:0 hw:0
[2020-11-03 16:16:00] GPU 2 [34C, fan 29%] ethash: 36.21kh/s, avg 25.87Mh/s, pool 29.23Mh/s a:1 r:0 hw:0
[2020-11-03 16:16:00] Total ethash: 14.33Mh/s, avg 82.45Mh/s, pool 116.6Mh/s a:4 r:0 hw:0
[2020-11-03 16:16:00] ----------------------------------------- Pool Status ------------------------------------------
[2020-11-03 16:16:00] daggerhashimoto.usa.nicehash.com ethash: 14.32Mh/s, avg 80.64Mh/s, pool 116.4Mh/s a:4 r:0
[2020-11-03 16:16:00] ------------------------------------------------------------------------------------------------
[2020-11-03 16:16:10] GPU 0 Ramp-up completed.
[2020-11-03 16:16:13] Pool daggerhashimoto.usa.nicehash.com share accepted. (GPU0) (a:5 r:0) (104 ms) (diff 5.61 GH)
[2020-11-03 16:16:17] Pool daggerhashimoto.usa.nicehash.com received new job. (job_id: 00000000b9a70a04)
[2020-11-03 16:16:20] GPU 1: detected DEAD (07:00.0), will execute restart script watchdog.sh
[2020-11-03 16:16:20] GPU 2: detected DEAD (0a:00.0), will execute restart script watchdog.sh
[2020-11-03 16:16:20] Watchdog script executor thread executing script 'watchdog.sh'
Rebooting
(by mining software). Check crash logs.

I think your issues are OC settings not Nicehash, can we see you OC settings and what cards are you using?

[bmoscato]

I think your issues are OC settings not Nicehash, can we see you OC settings and what cards are you using?

Keep in mind that these settings are stable on a straight Ethereum pool.

3 PowerColor RX 5700 XT running at:
Core MHz: 1300
Mem MHz: 915
Core Undervolt: 800
Power Stage: 4

I'm going to try dropping the Mem MHz to 900...

***UPDATE***

I've dropped my settings to the following and have crashed 3 times in 30 minutes:

Core MHz: 1300
Mem MHz: 900
Core Undervolt: 800
Power Stage: 4

[BogdanCo]

I think your issues are OC settings not Nicehash, can we see you OC settings and what cards are you using?

Keep in mind that these settings are stable on a straight Ethereum pool.

3 PowerColor RX 5700 XT running at:
Core MHz: 1300
Mem MHz: 915
Core Undervolt: 800
Power Stage: 4

I'm going to try dropping the Mem MHz to 900...

***UPDATE***

I've dropped my settings to the following and have crashed 3 times in 30 minutes:

Core MHz: 1300
Mem MHz: 900
Core Undervolt: 800
Power Stage: 4

Looks good, can you try also with Power Stage 7 instead of 4?

[rdluffy]

I've dropped my settings to the following and have crashed 3 times in 30 minutes:

I'll tell you something funny that I discover with my card (5700 XT Gigabyte oc Gaming)
There's a "gap" on memory frequency, that not works 100%, but this gap is not the limit

I use my memory at 1900mhz, but if I try something like 1885 or 1890 the memory became unstable, but if put 1900mhz, it works 100% without any issue, running here for 24/7
I think this is not an exact science, it's all test and try  

[bmoscato]

Looks good, can you try also with Power Stage 7 instead of 4?

Thanks, so far I've been stable for almost 4 hours without a restart using the last settings that I posted and pushing the Power Stage to 7.

***UPDATE***

Stable at 25 hours.

[philipma1957]

Looks good, can you try also with Power Stage 7 instead of 4?

Thanks, so far I've been stable for almost 4 hours without a restart using the last settings that I posted and pushing the Power Stage to 7.

***UPDATE***

Stable at 25 hours.

Nice numbers. Good work.

[bmoscato]

What's weird is that I'm running 3 PowerColor and 1 Sapphire RX5700XT and noticed that the power draw on the sapphire is a lot less than the PowerColor.

1. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 110W
2. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 109W (fan continually ramps up 100% - temp only @ 49C)
3. PowerColor Red Devil Radeon™ RX 5700 XT 8GB GDDR6 - 112W
4. SAPPHIRE NITRO+ Radeon RX 5700 XT 8GB GDDR6 - 95W

It seems like buying Sapphire cards would theoretically be more power efficient based on my small data sampling.

[badbart]

What's weird is that I'm running 3 PowerColor and 1 Sapphire RX5700XT and noticed that the power draw on the sapphire is a lot less than the PowerColor.

1. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 110W
2. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 109W (fan continually ramps up 100% - temp only @ 49C)
3. PowerColor Red Devil Radeon™ RX 5700 XT 8GB GDDR6 - 112W
4. SAPPHIRE NITRO+ Radeon RX 5700 XT 8GB GDDR6 - 95W

It seems like buying Sapphire cards would theoretically be more power efficient based on my small data sampling.

Yes, same experience here.  Next to blower 5700 xts Sapphire cards are the best.

[favebook]

What's weird is that I'm running 3 PowerColor and 1 Sapphire RX5700XT and noticed that the power draw on the sapphire is a lot less than the PowerColor.

1. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 110W
2. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 109W (fan continually ramps up 100% - temp only @ 49C)
3. PowerColor Red Devil Radeon™ RX 5700 XT 8GB GDDR6 - 112W
4. SAPPHIRE NITRO+ Radeon RX 5700 XT 8GB GDDR6 - 95W

It seems like buying Sapphire cards would theoretically be more power efficient based on my small data sampling.

Could it be that power draw difference is due to fans?

[bmoscato]

What's weird is that I'm running 3 PowerColor and 1 Sapphire RX5700XT and noticed that the power draw on the sapphire is a lot less than the PowerColor.

1. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 110W
2. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 109W (fan continually ramps up 100% - temp only @ 49C)
3. PowerColor Red Devil Radeon™ RX 5700 XT 8GB GDDR6 - 112W
4. SAPPHIRE NITRO+ Radeon RX 5700 XT 8GB GDDR6 - 95W

It seems like buying Sapphire cards would theoretically be more power efficient based on my small data sampling.

Could it be that power draw difference is due to fans?

The first two are the 2 fan model using the 8 and 6 pin power connectors, but both the Red Devil and Nitro+ are 3 fan models using 2x 8 pin connectors.

So, theoretically the Red Devil and the Nitro+ should be pulling close to the same power, but instead, the Red Devil is closer to the 2 fan model in power draw.

[BogdanCo]

What's weird is that I'm running 3 PowerColor and 1 Sapphire RX5700XT and noticed that the power draw on the sapphire is a lot less than the PowerColor.

1. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 110W
2. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 109W (fan continually ramps up 100% - temp only @ 49C)
3. PowerColor Red Devil Radeon™ RX 5700 XT 8GB GDDR6 - 112W
4. SAPPHIRE NITRO+ Radeon RX 5700 XT 8GB GDDR6 - 95W

It seems like buying Sapphire cards would theoretically be more power efficient based on my small data sampling.

Measured at the wall or into the SW (miner)?

[bmoscato]

What's weird is that I'm running 3 PowerColor and 1 Sapphire RX5700XT and noticed that the power draw on the sapphire is a lot less than the PowerColor.

1. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 110W
2. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 109W (fan continually ramps up 100% - temp only @ 49C)
3. PowerColor Red Devil Radeon™ RX 5700 XT 8GB GDDR6 - 112W
4. SAPPHIRE NITRO+ Radeon RX 5700 XT 8GB GDDR6 - 95W

It seems like buying Sapphire cards would theoretically be more power efficient based on my small data sampling.

Measured at the wall or into the SW (miner)?

In SMOS. The rig is closer to 570 at the wall, SMOS shows the rig around 545. I'm not sure how to measure each card independently @ the wall without running each card solo to get a sampling over a set time period.

[adaseb]

What's weird is that I'm running 3 PowerColor and 1 Sapphire RX5700XT and noticed that the power draw on the sapphire is a lot less than the PowerColor.

1. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 110W
2. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 109W (fan continually ramps up 100% - temp only @ 49C)
3. PowerColor Red Devil Radeon™ RX 5700 XT 8GB GDDR6 - 112W
4. SAPPHIRE NITRO+ Radeon RX 5700 XT 8GB GDDR6 - 95W

It seems like buying Sapphire cards would theoretically be more power efficient based on my small data sampling.

Measured at the wall or into the SW (miner)?

In SMOS. The rig is closer to 570 at the wall, SMOS shows the rig around 545. I'm not sure how to measure each card independently @ the wall without running each card solo to get a sampling over a set time period.

Basically if you want an exact number you run the rig with all GPUs. And then just disconnect 1 GPU at a time and measure the chance in wattage. This is a pain to do because there is no easy way to disconnect the GPUs, you have to disconnect the riser, and disconnect the PCIe power cables. If you have the riser plugged in without PCIe cables it will draw power and vice versa. Hence it won't be accurate.

You also have to do this with the GPUs at a hot temperature because the rig uses less power when colder than hotter. One reason is that the fans spin faster and I think GPUs in general use more wattage when running hotter, no idea why. However the SMOS reading is pretty accurate I would say at 545 from being an estimate somewhat.

[favebook]

You also have to do this with the GPUs at a hot temperature because the rig uses less power when colder than hotter. One reason is that the fans spin faster and I think GPUs in general use more wattage when running hotter, no idea why. However the SMOS reading is pretty accurate I would say at 545 from being an estimate somewhat.

More heat = more power needed to work as intended as a lot of power is going into heat. Same is with ASICs, during winter some use almost 100W less than in summer.

Basically if you want an exact number you run the rig with all GPUs. And then just disconnect 1 GPU at a time and measure the chance in wattage. This is a pain to do because there is no easy way to disconnect the GPUs, you have to disconnect the riser, and disconnect the PCIe power cables. If you have the riser plugged in without PCIe cables it will draw power and vice versa. Hence it won't be accurate.

Or instead, just run system with 0 cards, measure, connect 1GPU, measure and substract previous measure, connect 2 GPUs, measure and substract previous measure...
You will know exactly how much your idle system uses, you will know how much each GPU uses and you can do it while making your rig or after cleaning when you are rebuilding it.

[DrG]

What's weird is that I'm running 3 PowerColor and 1 Sapphire RX5700XT and noticed that the power draw on the sapphire is a lot less than the PowerColor.

1. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 110W
2. PowerColor Radeon™ RX 5700 XT 8GB GDDR6 (base model) - 109W (fan continually ramps up 100% - temp only @ 49C)
3. PowerColor Red Devil Radeon™ RX 5700 XT 8GB GDDR6 - 112W
4. SAPPHIRE NITRO+ Radeon RX 5700 XT 8GB GDDR6 - 95W

It seems like buying Sapphire cards would theoretically be more power efficient based on my small data sampling.

Measured at the wall or into the SW (miner)?

In SMOS. The rig is closer to 570 at the wall, SMOS shows the rig around 545. I'm not sure how to measure each card independently @ the wall without running each card solo to get a sampling over a set time period.

Basically if you want an exact number you run the rig with all GPUs. And then just disconnect 1 GPU at a time and measure the chance in wattage. This is a pain to do because there is no easy way to disconnect the GPUs, you have to disconnect the riser, and disconnect the PCIe power cables. If you have the riser plugged in without PCIe cables it will draw power and vice versa. Hence it won't be accurate.

You also have to do this with the GPUs at a hot temperature because the rig uses less power when colder than hotter. One reason is that the fans spin faster and I think GPUs in general use more wattage when running hotter, no idea why. However the SMOS reading is pretty accurate I would say at 545 from being an estimate somewhat.

While chemical batteries poop out in cold weather (due to their decreased ability to have the electrons transfer between cathode and anode due to slower chemical reactions with decreased ambient energy), electrical circuits work better when there is less heat. Hence the holy grail of non sub-zero superconductors would be ground breaking but we're not there yet. This is one of the reasons why high voltage systems are used since they pass a lower amperage and tend to generate less heat (like transmission lines and battery electric vehicles). And of course as mentioned the fans have to spin up higher per the fan curve and that's going to eat 5-10watts as well.

[bmoscato]

Basically if you want an exact number you run the rig with all GPUs. And then just disconnect 1 GPU at a time and measure the chance in wattage. This is a pain to do because there is no easy way to disconnect the GPUs, you have to disconnect the riser, and disconnect the PCIe power cables. If you have the riser plugged in without PCIe cables it will draw power and vice versa. Hence it won't be accurate.

Or instead, just run system with 0 cards, measure, connect 1GPU, measure and substract previous measure, connect 2 GPUs, measure and substract previous measure...
You will know exactly how much your idle system uses, you will know how much each GPU uses and you can do it while making your rig or after cleaning when you are rebuilding it.

Honestly, I'm good with the ballpark numbers. I have no desire to do this. lol