 Show Posts
|
|
Pages: [1] 2 3 4 5 6 »
|
I have a bunch of 4gb cards mining using smos and phoenix miner. I try changing the riser.
switching to phoenix miner seems to have solved the issue. Thanks! |
|
|
|
did you try using a different pool for mining eth?
also can you flush the oc settings and see if that is causing any issues?
did you try reverting back to old claymore version and check if it crashes anymore?
I did not try a different pool yet, I have wiped the overclock settings as that was my first thought on what would be causing it. It did help a little (in terms of frequency of the rigs going down) however they still go down. Did not go back to an older version of claymores as I believe the newest version of claymores made all previous versions stop working. Unless that has been fixed or changed since I was forced to move to version 15 by smos. |
|
|
|
|
As the Title says is anyone experiencing issues mining ethereum on simplemining with claymores 15? I have started to see just today a lot of miners with 4gb rx 570's crash about every hour or so. Same thing happens on ethminer to. I tried to do some research about it and all i can see is that 4gb should still work technically but there may be issues and everyone is saying to upgrade to 8gb cards anyways. Next time one crashes ill post the crash log.
GPU #4 recognized as Radeon RX 470/570
GPU #5: Ellesmere (Radeon RX 570 Series), 522 MB available, 32 compute units (pci bus 8:0:0)
GPU #5 recognized as Radeon RX 470/570
GPU #6: Ellesmere (Radeon RX 570 Series), 522 MB available, 32 compute units (pci bus 9:0:0)
GPU #6 recognized as Radeon RX 470/570
POOL/SOLO version
AMD ADL library not found.
GPU #0: algorithm ASM 1
GPU #1: algorithm ASM 1
GPU #2: algorithm ASM 1
GPU #3: algorithm ASM 1
GPU #4: algorithm ASM 1
GPU #5: algorithm ASM 1
GPU #6: algorithm ASM 1
cudaGetDeviceCount returned error 38
-_ no CUDA-capable device is detected
No NVIDIA CUDA GPUs detected.
Total cards: 7
AMD Cards available: 7
GPU #0: Ellesmere (Radeon RX 570 Series), 503 MB available, 32 compute units (pci bus 1:0:0)
GPU #0 recognized as Radeon RX 470/570
GPU #1: Ellesmere (Radeon RX 570 Series), 503 MB available, 32 compute units (pci bus 2:0:0)
GPU #1 recognized as Radeon RX 470/570
GPU #2: Ellesmere (Radeon RX 570 Series), 503 MB available, 32 compute units (pci bus 5:0:0)
GPU #2 recognized as Radeon RX 470/570
GPU #3: Ellesmere (Radeon RX 570 Series), 503 MB available, 32 compute units (pci bus 6:0:0)
GPU #3 recognized as Radeon RX 470/570
GPU #4: Ellesmere (Radeon RX 570 Series), 503 MB available, 32 compute units (pci bus 7:0:0)
GPU #4 recognized as Radeon RX 470/570
GPU #5: Ellesmere (Radeon RX 570 Series), 503 MB available, 32 compute units (pci bus 8:0:0)
GPU #5 recognized as Radeon RX 470/570
GPU #6: Ellesmere (Radeon RX 570 Series), 503 MB available, 32 compute units (pci bus 9:0:0)
GPU #6 recognized as Radeon RX 470/570
POOL/SOLO version
AMD ADL library not found.
GPU #0: algorithm ASM 1
The miners will stop mining, reload and when they do reload they come up and get stuck looking like this.
|
|
|
|
|
how much are you looking to pay?
|
|
|
|
The S9i runs fine on one hash board, I tested both PSU's. When I power two, this time instead of getting the red fault light and fans slow down. I get one of the PSU breakout boards blinking, i.e., the digital voltage readout blinks, also the PSU green LED blinks.
Are those breakout boards linked together? Not sure about parallels break out boards but other setups that are similar to this such as a 750w + 1200w psu can be used to power an S9. No need to link the breakout boards. I use the 750w for the control board and one hashboard and the 1200w for the other 2 hashboards. This sounds weird but the correct way to fire it up is to turn the 750w on first and then quickly following turn the 1200w on next. Either way If both these psu's are 750w and you are running on 110v I would say thats not enough, your stressing the psus to much. They might be able to do it on 200+volt but even then I would recommended (2) 800watts or like I said a 750 and 1200. Hopefully you get it sorted out and no damage to the machine occurs. |
|
|
|
|
get a multimeter and test the voltage coming out of your wall socket, look for drops or spikes
|
|
|
|
|
Awesome Dude! Love those Octo Miner Cases! So Many GPU's LOL!!
|
|
|
|
I wouldn't be able to run more than the eight with my power as is. Plus this is at home and I'll hope to venture out to cheaper power and bigger space if/as needed. So 3k would be plenty and the 1.6k cfm would be about the minimum then?
are they all s9's? I wouldn't want my farm to have "The minimum amount of cfm exhaust" I would rather have more than enough to accommodate for the few days of the year where it might get extremely hot outside. Calculating the CFM on these really isn't like hard numbers, by that I mean the fans on the machines are variable speeds. At full speed the s9's move almost 300cfm, If you take that into consideration 8*300=2400CFM. Thats worst case scenario (I would build it to those numbers), the colder the intake temp on the miners the slower the fans spin meaning the less cfm they actually move. |
|
|
|
Priced out the build, can defiantly save on the fan and hopefully other items. Thoughts? Only concern was ensuring intake, any others? or other suggestions? I just want to do one duct register in the ceiling, so would 3k CFM removal be overkill for a small farm of 8 asics on a 100F day? Bout as hot as it would typically get through a few summer days here.
[...]
Yes 3000CFM will be fine for 8 units, it allows you to have some wiggle room and if you wanted to throw in one or 2 more machines you could if you wanted. |
|
|
|
|
Raritan makes 415Volt pdu's that step it down and output at 240volt which is perfect for miners. I think the make close to 60KVA units
|
|
|
|
|
Your not being limited by the transformer here, your limit is your 400amp 240volt panel. at 80% (I wouldn't recommend going much above 80% though) you can have roughly 46 s9's
fun fact those transformers are under rated and over built. Ask the utility company, on a 2500ka unit they told us they were ok with loading it to 120%
*EDIT* Didn't see this was 3 phase! You are correct on the 95 machines you should be ok hooking those up!
|
|
|
|
|
Use SMOS (Linux) This mobo can run 7 cards if you update the bios to the mining version and use the m.2 slot for one of the cards.
Have 7 rigs running 7 cards with this mobo right now.
|
|
|
|
|
do you still have the 1070 zotac?
|
|
|
|
I own a big farm, and I'm getting annoyed by W10.
Can't run Linux because the person who maintains the farm isn't familiar with it, so I decided to give smOS a try.
I'll be installing the OS tomorrow and running it for a day, or two, on a 13 GPU rig and see how it goes.
Awesome project, kudos.
Everything fine so far. Took me 5-10 minutes to understand everything. The only thing I can't understand is where I can give names to rig cards individually? It's possible to name groups, but not GPUs, right? You can name individual rigs just click on the name for example if it says "noname" click it and it will bring up a box where you can name it whatever you want.  |
|
|
|
|
You need to have an intake opening or openings for the same cfm that you are exhausting. If your one intake is only 12000cfm you need 3 more to match your exhaust.
|
|
|
|
Most of the time, reflashing usb stick or changing with new one solves the problem  After changing to cheap 32GB SSD's stability over time improved a lot. No need to use USB sticks. You just need to choose advanced mode in etcher to burn to SSD/HDs. Use Rufus to burn to SSD's and HDD's Average burn time for me is about a minute etcher takes like 5-8 minutes for some reason |
|
|
|
I don't know how useful it is to compare the temps like that. The chip temps are reported in degrees Celsius. Even with 100F ambient air typically the S9 will be in the 80-85c range from my experience.
Correct, with proper exhausting though! OP you should post some pictures of your setup! |
|
|
|
|
More info would be helpful for troubleshooting.
What pool are you trying to point it to?
Is the machine Cold when you start it up? the auto tune machines need to be somewhat warm during startup, if its in a cold room sitting off for a while when you go to turn it back on it may not boot up properly, has to do with opening the circuits on the boards I think.
Whats your network configuration like? dhcp, static addresses?
Are you sure you are not missing anything from the pool config tab? If you are not using the failover pool slots its best to wipe all of the info out of those lines. I suggest setting up accounts on other pools and putting them into the slots in case your main pool does go down.
|
|
|
|
|
