|
I picked up some Octominer X8ULTRA PLUS chassis on the cheap locally from someone that's bailing.
Can AM interact with the on-board smart hardware watchdog & fan controller on these?
I know Hive OS can, and I have most of my rigs running the Hive OS image but having AM manage everything.
What about if I run windows on the rigs?
Thanks
|
|
|
|
The miner profitability issue seems to have sorted itself, but I have been seeing $0.00 for ETH along the bottom bar for several days now:  This is what I have for statistical providers:  Any idea what might be going on? Buddy of mine has the same settings as me and is seeing the same issue. |
|
|
|
|
Anyone noticed how the value of ETH along the bar at the bottom of the AM window is wrong? Mine is currently showing $1,865.95, but ETH sits at $1,428.94 according to Coinbase and coinmarketcap.
Also, I have all my rigs on Ergo right now and AM is showing that my 5.92 GH/s is earning $29.45/day. That's way off as the 2miners pool I'm on is showing $7.47/day, which is much more realistic and in-line with the Ergo coins I have been paid out so far.
Will this discrepancy sort itself out once things "settle down"?
|
|
|
|
|
On my all LHR rig, I can lock the core on each GPU by passing the following parameters:
--cclock @1350,@1200,@1200,@1400,@1200,@1350,@1200
It has the following GPUs:
0 3070Ti
1 3080Ti
2 3080Ti
3 3060Ti
4 3080
5 3070Ti
6 3080Ti
However, I also have a rig with a mix of 1000, 2000 and 3000 cards as follows:
0 1070Ti
1 3060
2 1070Ti
3 2080Ti
4 3080Ti
I want to lock the core on GPU 4 @1200. How do I do that since some of the other GPUs don't support core lock?
EDIT:
Like this:
--cclock ,,,,@1200
|
|
|
|
I manage my rigs with Awesome Miner, so I don't set that stuff manually from the command line. But here's what the rig with the arguments below looks like:  And here's the command line parameters that AM is sending: Command line parameters: --cclock @1450,@1200,@1200,@1400,@1200,@1450,@1200 --no-watchdog -a ethash -o stratum+tcp://us1.ethermine.org:4444 -u 0x304AcBc3bfa713AC08bD2F6d5a59AaE2df047F72.miner006 --api 0.0.0.0:4028 |
|
|
|
i startet a test yesterday and NBminer works very well
any tips for optimizing the hashrate or power drow?
Use the absolute core lock argument. i.e.: --cclock @1450,@1200,@1200,@1400,@1200,@1450,@1200 |
|
|
|
|
Works for me with Awesome Miner and windows clients for 3060Ti, 3070Ti, 3080 and 3080Ti (all LHR of course).
|
|
|
|
Did some initial tuning and this rig is performing well now:  "pool side":  I was at 525 MH/s before using T-Rex, so this is an impressive improvement of 140 MH/s. Like getting a 3090 for free! My 3080Tis are now more efficient than my 3090s, which are all doing around 122 MH/s @ 300 Watts. |
|
|
|
Looks good. LHR rig before:  Same rig after (before any tuning):  |
|
|
|
How do I assign custom name to each GPU in the rig? I did that a while ago with one of my rigs but totally forgot how to do it, search didn't return any answers. I want to be able to later assign clocking profiles to them based on names (which include manufacturer/model/cooling type)
There are unfortunately no great feature for this today. For a selected miner, on the GPU tab, you do have a button for "Set name", but it's a bit limited at the moment. The name will only be used on the GPU tab itself. A future improvement would be to display it everywhere if defined. Interesting, I think I saw it somewhere in the manual, OC settings rule based on GPU name? You can define Clocking Profile Groups (see the Options dialog, GPU Clocking Profiles, where you find the groups as well) looking at the reported GPU Name. This is however the true name of the GPU like "nVidia GeForce RTX 3090", not any custom name you defined yourself. With the Clocking Profile Groups you can for example define that all GPU names containing "3090" should have one profile or all GPU brands matching "ASUS" should have another profile and so on. Well, that's the issue I have. Even though all my cards are 3080s I have to use slightly different settings for them being from various manufacturers and cooled differently. As a result at the moment my clocking profiles and profile groups look like a huge mess hard even for myself to sort out what's what. I understand your point here. Are you also mining different algorithms so you define clocking settings not only per GPU but also per algorithm? Or are you only mining one single coin/algorithm all the time? I have this same issue. My OC profiles are a mess too:  Different profiles for different models and different algos. |
|
|
|
Here's one of my 3090's in a test rig with just the default Xintensity:  Probably got more in her (EVGA KingPin 3090 on water), but figured I'd do a baseline'ish run first. This is after about one hour. |
|
|
|
I have a RX 480 8GB Reference with a Swiftech Komodo full block that also blew a power choke a few days ago.  Has Samsung memory too:  Nice card until it blew up:  Guess I'll put it up on eBay as an auction to see what it will fetch, with the water block. |
|
|
|
God damn, what kind of solar setup is this? You said you don't need batteries for backup? How is that even possible? What happens once the sun goes down in the evening and night? Wowu
My solar farm is tied to the grid. So anything I produce in excess of what I consume, flows back into the grid. At night and on cloudy/rainy days, I draw power from the grid. With the Net Metering arrangement I have, I'm credited for every kWh I put back into the grid at the same rate they charge me, i.e. ~ $0.12 per kWh. |
|
|
|
It takes a pretty serious investment to produce meaningful power from solar. I have a 50 kWh farm in Virginia and I try to match my mining rig consumption with what I produce. I have Net Metering with the power company, so I don't need to worry about messing with batteries or anything like that. Here's my production vs. consumption for 2020:  Its a bit misleading since I move my rigs to the house in the winter months to supply all the heating I need. So the above graph only captures the consumption of the mining rigs when they are located in my shop building. 2021 so far:  I sold all my 1080Ti, 1070 and 1070Ti rigs in November and December of 2020 and have been slowly building out 3000 series rigs this year as I've been able to get GPUs. That is why you see a gradual climb in consumption since April (which is typically when I move the rigs from the house to the shop). Looking at last month specifically:  So on nice sunny days, I generate about 300 kWh/day, and my rigs consume about 175 kWh/day. I do run a few rigs at the house all year, so the actual daily consumption is more like 200 kWh, or about 8,000 watts at the wall as seen here:  Areal of solar farm:  As mentioned above, I have Net Metering with the POCO. So any excess energy I produce, goes into a "bank" of kWh on my account to be consumed during the winter months.  They do make me pay a basic meter charge of $30.75 plus $3 in County taxes, but I can live with that.  Btw, the meter is also supplying power to my main house, so it basically gets a "free ride" and "free" heat in the winter. |
|
|
|
@crazydane You have auto-exchange on to BTC? Looks like it would take forever to get payment. Maybe change it to LTC? Just a thought  Just got my first payment:  So 15 days for a little over $90 on a 5800 and 5950X. Not bad at all! |
|
|
|
|
Any chance of adding CoinGecko as a Statistics Provider so that we can pull in coins like Raptoreum that is not listed on any of the existing providers from what I can tell.
You already have the Ghostrider algo, just not the coin, so I don't seem to be able to CPU mine Raptoreum properly at the moment using AM.
|
|
|
|
I guess the EVGA queue gravy train has come to a screeching halt for those of us that were using it: --- cut --- Update 10/4/2021 Hello, In an effort to expedite those in queue who are still waiting for a EVGA GeForce RTX 30 Series Card, and to be fair to all EVGA customers, we are making a running change: Accounts that have NOT purchased an EVGA 30 series card yet as part of the queue will be prioritized. Accounts that have already purchased an EVGA 30 series card in the queue system, will be behind those who have not. Note: B-Stock (RX) or Step-Up 30 Series products will not affect queue positions. We hope that this update will allow more gamers to get their hands on an EVGA GeForce RTX 30 series card. If there are any questions please contact our support at sales@evga.com. Thanks EVGA --- cut --- I got about a dozen cards via the queue, will another 5 within minutes of popping. |
|
|
|
|
TeamBlackMiner 1.10 was released yesterday that introduced API.
|
|
|
|
|
Show Posts
