Hi there, the thread is long and i couldnt find an answer, anyway from time to time i got 0 sols on all cards and i need to close the miner and open it again. IS there any automatic solution like script or something which can do this automatically? If cards start making 0 sols, end the process and start it again?
|
|
|
Don't really think most of us will mind this small increase when the difficulty is doubling every 30 days and in 45 days the block times are going be double. So losing 1MH/s won't really be that much when the difficulty quadruples anyways.
So now you believe me 29-22 drop is not really a 1mh but anyway, where can i read about blocktime increase?
|
|
|
RX series and 380 and 290/390 don't lose speed with DAG increases. Its only Pitcairn and Tahiti that do that.
Make me a favour and run a claymore on RX with -benchmark 135, -benchamrk 155 and -benchmark 199 if you dont have any RX gpus, i will be happy to show you screenshots RX and Tonga are the new tahiti i dont have hawaii to check. About the diff rise rate, there is just not enough gpu-s on the market to keep the diff rate. And even if there were enought if gpu diffrate increase with the same speed, once the profit becomes lower the new miners will be lower too. If this day a GPU gets ROI in 2 months, and diff goes x4 (which is impossible simple because there is no enough gpus in the world) and price stays the same , which means the ROI going to be 8 month, do you think the people will keep investing in rigs ? No of course. You should be really stupid to think that diff increase rate is constant.
|
|
|
For the ice age I have this quote from Buterin on the next fork "In general, if we get it out before the end of June it’s even better, and if we get it out before the end of August that’s still, like, less bad. My recommendation is probably still end of June as a … normal case target and end of August as a worst case deadline" For difficulty it only took about 5 weeks to double and the price is starting to get stable around the high $300 close to $400 range So a example estimate Today 164 mh = $1000 month ~ July 24 = $500 month ~ September 1 = $250 month ~ October 7 = $125 month This is assuming a static price with a static climb in difficulty, the climb will begin to slow as profit margins become thin for people and the gold rush ends. I think realistically you can expect profits to get pretty thin around November or December at best, unless price goes up. Vega will be a non issue, even if it did amazballs with hashing the limited units would be a drop in the bucket. This would be true if there were unlimited gpu quantity, but most is sold out.Peaople start mining with 1050ti and 560 but there being sold out fast too. Also RX is starting to loose hash over time like it was with tahiti 18 months ago. On epoch 155 the RX would be around 22 mh which results in diff decrease.
|
|
|
What speeds do you get on dagger? No matter what settings i try i always get around 2mh/s slower then claymore..
|
|
|
I suppose I wasn't clear enough. Modded BIOS'es for Claymore's may not work well with excavator, and you need to flash back the BIOS to the stock one first. You can mod the BIOS for excavator, but please keep in mind that excavator may require different settings.
Ok, but whats the point of stock if i still dont have hardware acceleration?
|
|
|
Should it be runned with something like this: [ {"time":0,"commands":[ {"id":1,"method":"algorithm.add","params":["daggerhashimoto","daggerhashimoto.eu.nicehash.com:3353","34HKWdzLxWBduUfJE9JxaFhoXnfC6gmePG.testrig"]} ]}, {"time":3,"commands":[ {"id":1,"method":"worker.add","params":["0","0"]} ]}, {"time":10,"loop":10,"commands":[ {"id":1,"method":"worker.print.speed","params":["0"]}, {"id":1,"method":"algorithm.print.speeds","params":["0"]} ]} ]
With in 10 mins my speed slowly increased and settled on 27.3mh/s on RX 480 8gb with modded timigns. With claymor i got 29.2-29.4 with same clocks. Is it normal to get so much time to get to optimal speed and ist it suppose to be that much slower than claymores? p.s. I didnt notice that i need to run hardware acceleration, but the test mode on windows didnt help with the driver signature enforcement and all my rigs are remotely located so i guess i have to wait for the next version. p.s.2. Ok, i found the bat file for disabling the sign req and it worked but i still got: [07:12:18][0x00000b24][info] opencl | Loaded phymem for direct PCIe access. [07:12:19][0x00000b24][info] core | Found OpenCL device: AMD Radeon RX 470 [07:12:19][0x00000b24][info] opencl | PCI Vendor/Device ID: 0x1002, 0x67df [07:12:19][0x00000b24][info] opencl | PCI Subsystem Vendor/Device ID: 0x174b, 0xe347 [07:12:19][0x00000b24][info] opencl | Memory Type: GDDR5 [07:12:19][0x00000b24][info] opencl | Memory Size: 8192MB [07:12:19][0x00000b24][error] opencl | Failed to gather information for hardware acceleration. [07:12:19][0x00000b24][error] opencl | Hardware acceleration will be disabled. p.s.3. Same thing On RX 480 8gb, 480 4Gb. 470 8gb
|
|
|
I noticed that Nitro 4xx got Vddc offset on the fast bios and not on the silent one. But for some reason VDDCI mod is not working on the fast bios where the offset is available. Is it possible to add a static voltage and offset(where it is available) together?
|
|
|
How does -mpsw works? I thought it should be -mpsw password and then set it in the remote manager , but its not like that..
Exactly, set "-mpsw xxxxx" in miner, then set same password in EthMan in the miner properties, that is all. Yep the problem was i didnt set it in ethman. Another question, I setn gmail smtp and i get "Could not load SSL library", any fix?
|
|
|
Can you point me how is the best way to configure dual mining: Ether on pool and only change pascal/decred/sia on nicehash for best profit?
|
|
|
Wattman is showing 0 when you set it that way, but it's really not 0. (can't be zero ) I think it's a bug in wattman. Also you need to have vddc offset added in your bios if you want it to work the way you say. There are only a few cards that have vddc offset added in stock bios. I dont know but at stock bios when i start the miner i see the gpu going through the states and when the actual mining begins and gpu get steady 100% load it locks on P7, but when i set them as i showed you and start the miner i just see how from 300 goes to 1100 and no fluctuation... So in your exp which is better alk states the same and static voltage per states or global offset?
|
|
|
Yes it's better, but voltage setting by states is not working on every card.
I think there is a workaround without needing to set voltage on each state. When the core clock is set to lets say 1100 you can actually do this: Set all states to 1100 and increse the P0 state with lets say 100 just to be safe if somebody needs offset 200(because default 800-200 offset is kind of low 600mv) Then when the card boots you got this: All states but the P1 automatically eliminated because they have same core clock and auto voltage. From there you just set the offset(in the bios of course) and your card cant crash because P4 state with -150 offset is not stable. How does it sounds?
|
|
|
Ok, then i didnt get you the first time. My point is with global offset which is for every state, the card may be stable on the last P7 with -150 offset but not stable on the P5 for example with -150. So in that case if in any moment the OS reduces the GPU load during mining the GPU state will fluctuate to a lower one and can become unstable. So my question is isnt it better to set all P states to the desired core clock which can prevent the card become unstable with the choosed offset on some of the lower states(as it is on my first screenshot).
|
|
|
Hi, can you add more steps to the offset. My cards are on core 1100 and auto voltage -150 offset is about 925mv but the can handle less then 900. Also isnt it better the clock to be set on all 7 states, because i think the offset is for all states and if the card is stable on the last state with -150 it can crash lets say on state 4 if for some reason the load of gpu drops for a second during mining?
when setting vddc offset that means every state's voltage is changed accordingly to the offset you chose. I dont know i just thought its like that... is it only for the last state? Ok, jsut checked and looks like i'm right. Here is a shot of a card with my settings, you can see in idle (state 0) vddc is 800mv as it is suppose to be: Here is a card with global offset -125 done with the editor, and in idle vddc is 675mv: Managed to get a shot when in P6, its suppose to be 1150mv but its 1025(1150-125):
|
|
|
Hi, can you add more steps to the offset. My cards are on core 1100 and auto voltage -150 offset is about 925mv but the can handle less then 900. Also isnt it better the clock to be set on all 7 states, because i think the offset is for all states and if the card is stable on the last state with -150 it can crash lets say on state 4 if for some reason the load of gpu drops for a second during mining?
when setting vddc offset that means every state's voltage is changed accordingly to the offset you chose. I dont know i just thought its like that... is it only for the last state?
|
|
|
Hi, can you add more steps to the offset. My cards are on core 1100 and auto voltage -150 offset is about 925mv but the can handle less then 900. Also isnt it better the clock to be set on all 7 states, because i think the offset is for all states and if the card is stable on the last state with -150 it can crash lets say on state 4 if for some reason the load of gpu drops for a second during mining?
|
|
|
On my only win 7 machine i got strange problem since 2.9. Since time to time the node is stopping, the miner keep running.
On the win 10 machines, no problems.
have you an error message or just quit? No this is how i find it. Btw maybe is good idea to add dcri parameter for a eth only, since ASM option dcri is affecting eth only mode too. the default is 30 but my best value for my rigs is 10. If i change it from -dcri (DCRI) to -dcri 10 in the config file i cant go dual mode when i decide with just enabling it on the website because i need to edit the config too back to -dcri (DCRI)
|
|
|
On my only win 7 machine i got strange problem since 2.9. Since time to time the node is stopping, the miner keep running. On the win 10 machines, no problems.
|
|
|
Hi there, nice update. But in the current options i dont think i can mine ETH on pool and DCR on nicehash on ALGO? LEts say pascal is getting more profitable than DCR on nicehash to switch?
|
|
|
|