My fix was -eres 0 , for gtx 1060 3gb, and this is important...do not close miner if you get errors, it will try reconnect several times, after a while miners will start w/o error.
Thanks! I started having this issue today just on one of the three miners i have with 1060 3GB cards, ODD! -eres 0 fixed it, thank you. indkt. |
|
|
|
ccminer.cpp: In function ‘void* miner_thread(void*)’:
ccminer.cpp:1852:12: error: ‘Sleep’ was not declared in this scope
Sleep(1);
^
Makefile:1836: recipe for target 'ccminer-ccminer.o' failed
linux not compile on ubuntu
Replace Sleep with usleep I made a pull request to fix this but for some reason sp didn't want to merge it. Worked!, Thanks!  |
|
|
|
|
Hello
I have been following this thread for a while
I administrate several asic farms and GPU farms, I own GPU farms.
the thing is easy ,as a general rule (it goes up and down obviously)
Asics require 2X to 2.5X more power consumption that GPU at the same profitability ratio.
Asics usually costs half of what a GPU rig costs at the same profitability ratio.
so in order, lets say, to produce 5 USD/day youll need 1 Rig of GPUs NVIDIA that will consume 650W or an asic that consumes 1600W
and yes the cost of the GPU rig is probably twice the cost of the ASIC.
but remember, power is everything, power determines how big can you go, how much can you expand
rent a place its easy, cooling it's hard but doable once you learn how the air dynamics works.
power....... it is another matter
you cannot request a: let's say 500KW in a house, the power company will refuse, it will request you to present the proper papers, as a company, since a house cannot use that level of power, as a company you need to exist as such, and pay taxes and all,and even so you still need permissions from government explaining what are you gonna do with all that power.
Anything above 300KW will probably require you to have your own high voltage transformer that will probably cost close to 100K USD plus instalation and proper certifications that takes a lot of time and money to do.
so, unless you plan to have only a very few asics in you small flat, if you want to go big, the only way is maximum power efficiency, and that it's not ASICS, it's GPU, do the math, buy both if you wan't and compare profitability vs power.
most people I know, got tired of GPU because they are difficult to configure, yes they are, but right now NVIDIA 1080Ti and AMD VEGA are the the most efficient GPUs in the market, it takes a lot of effort to make them work at maximum efficiency, but it worth it, and allows you pack a lot more of production in the same place and power capacity than will allow you with asics.
remember, power it is your ceiling , not money, power capacity is what tells you how high can you go.
cheers
indkt.
|
|
|
|
the biggest problem on EARTH, BY FAR, BY HUGE FAR IS: We have 3,8 bl people that gain less than 2$ a day, that obviously live in third world places were they can't get education, health care and all that, and they are desperately poor and want to migrate to other parts of the world were they can have an oportunity. Every year, according to OMS (WHO), more than 80 million babies are born on those third world countries. The problem is that, we face a destruction on most advanced societies. Inmigration is collapsing our way of living , our services and creating huge highways of migrations towards our countries, putting in danger our way of living as well as our possibilities to offer OUR OWN PEOPLE A RESPECTABLE WAY OF LIVING. THERE IS NO WAY migration can solve the issue. We have to help them ON THEIR OWN COUNTRIES and control very toughly those illegal inmigrations. We have to. There is no other option. before answering, watch this:_ https://www.youtube.com/watch?v=NvlSwsV7fmY +1 Agree immigration is destroying our way of living as well. this does not looking good |
|
|
|
new record on hashrate for ETH.
262350 Gh/s.
WTF is happening here??? This is destroying profit!!!!!! OMG!!!!!!!
4 hours ago we had 252.000GH/s.
As time passes by , more and more and more GH are joining the pool while THIS IS BECOMING LESS PROFITABLE. Are the people really mad?? Or is it that people have definetely abandoned many many many other coins???
In early June I calculated 10.5 mil 1060s on Ethash, 2.5 mil 1060s on Equihash. With Equihash ASIC and no zcash fork (75% of equihash nethash), 2 million + 1060s need a new home. And they will flock to Eth. +20% difficulty in Eth is expected. Even more if new GPUs with GDDR6 and better hash rate launches. ETH won't last that long. This year PoS will crash mining. Hope people still buying gpus for mining get totally rekt. Also for ASICS. They have been claiming that for about three years now and we a still mining hehe, I don't believe them LOL |
|
|
|
Don't mine with a laptop, ANY laptop, even the biggest Alienware or MSI gaming series will produce nothing and we are talking about several thousand dollar machines!. specially with HP!, HPs are poorly designed machines!, you'll endup killing it in no time. Notebooks are designed for low power, mining requires insane amount of processing power and that requires.... well.... POWER so no, build a rig specifically for mining, BIG ASS PSU, POWERFUL GPUs, if you don't have much money start small like I did, i started mining with two used AMD 5830, then sell them and buy two 7850, then 7950 eventually endup with a big farm that produces more than my work does hehe but takes time and effort and money , LOTS of money. you'll get there eventually, invest small but constant, read, investigate, and start small, but with the proper stuff, notebooks are not designed for that. cheers indkt. |
|
|
|
GPU miners are too stupid and never adapted to the changing landscape , they think their GPU are still profitable when they are not.. Asic and fpga is where the money is at.
The longer you hold gpus the more rekt you will get
That is true for the short term if you get a first batch one. Otherwise once ASICS are out the difficulty will kill your profits and you have a loud doorstopper. FPGA will be a different story as long as the Devs keep up with the algos If you bought the S9 in the first batch, you can still make some profits now. Really?, at 4.5 USD/day @ 1600W power consumption at the wall the profit is marginal at best, GPU still get you twice the profit at the current power consumption if you aim at maximum power efficiency |
|
|
|
GPU miners are too stupid and never adapted to the changing landscape , they think their GPU are still profitable when they are not.. Asic and fpga is where the money is at.
The longer you hold gpus the more rekt you will get
If you have free power yes, if not ASIC, it is not profitable. FPGA only if you design your solution in-house, you'll have an edge, otherwise will not, either. |
|
|
|
|
target: 55
limit: 70 (shutdown temp)
My rigs always stays between 50 and 60, in summer (now we are in winter here, so rigs at 45C not more) it could get very rough, some gpus reach 62/63 but not more.
anything above 70C will deteriorate cards seriously and i always babysit my cards a lot, they are too expensive to risk damaging them.
|
|
|
|
|
BITMAIN happens, another algo destroyed by this company.
|
|
|
|
If your network is ASIC resistant you need to worry about being prone to GPU miners that have huge hash rates. If you are GPU mined like the majority of at least the minor alts you have a much larger pool of potential adversaries. That's the point I'm trying to make.
It is also worth noting that one tends to underestimate the scale of GPU mining operations. Those are commercial operations buying GPUs wholesale. Mining has become an uphill battle for the hobbyist ever since it turned into an industry, no matter if we're talking about GPU or ASIC mining.
That being said, I'm fully with you that the large upside of GPUs is that they are easier acquired by regular people. I honestly wish that ASICs would also be as accessible.
Hello I don't think most people understand how SERIOUS this is. This is about ONE company, BITMAIN, yeah there is innosilicon right, which is owned by bitmain i'm pretty sure. they have a kill switch embedded inside the miners, the miner calls home every time and if you block this, the miner will not start, i know, i've tested it. one company owns ALL of the hashrate for every algorithm. really you don't see it as something dangerous? GPUs are general purpose computation devices, it can programmed by anyone, it can be installed by anyone. ASICS are not, only bitmain can design, build and program, why? because even if you design the solution in a FPGA you cannot create 100 asics for you, or 1000 no one will produce it, you need to order 100.000 or MORE, specially if you want to use 14nm or better technology, you need to order more than a Million chips minimum. the problem is not WHO mine with it the problem is that BITMAINS owns the status quo for the ENTIRE MINING OPERATION IN THE WORLD that's WHY asic are so dangerous. and ZCASH community doesn't see it, I can't believe it, unless, someone pays enough ppl to look the other way , which is possible, but i have no proof of that, just a thought. cheers indkt. |
|
|
|
Hello, I just like to report that, newer versions (11,2 forward, crashes with the OhGodACompany's Enlargement pill mempatch with GTX 1080s) Nvidia driver crash, doesn't happen with 9.8 Version, which works just fine.
without the mempatch works fine but the hash rate on the 1080 is a joke due to the GDDR5X memory.
Lubuntu 14.04 , driver 378.09
actually i noticed that 11.6 does that (low hash rate with ethelargement pill), but 11.4 runs fine at 50+ mhs. driver doesnt crash on 11.6, just low hash rate. havent tried 11.7 yet. thought it just operator error on my part.have 2 1080tis on ethlargment. Thanks!, interesting!, I will download 11.4, i don't have it, i tested 11.2, 11.6 and 11.7 also 9.8 , 9.7 and 9.4. there is a new 11.8 will test it now. Could you possibly show the power the cards are using?
me?, TPL power used? 1080 top efficiency is at 100W TPL 1080 Ti top efficiency is at 140W TPL  |
|
|
|
|
Hello, I just like to report that, newer versions (11,2 forward, crashes with the OhGodACompany's Enlargement pill mempatch with GTX 1080s) Nvidia driver crash, doesn't happen with 9.8 Version, which works just fine.
without the mempatch works fine but the hash rate on the 1080 is a joke due to the GDDR5X memory.
Lubuntu 14.04 , driver 378.09
|
|
|
|
|
Hello
I got segfault almost all the time with driver crashes, some times, when i don't get segfault, it works perfectly for a long time, it always segfault at the beginning, not during mining.
Linux 14.04 LTS (64) 8GB RAM, 6x 1080 Strix from flash drive
Run enlargement pill with & and then the miner.
claymore 11.7
any ideas what could cause it? the strange thing is that, sometimes i manage to make it work and it never crashes until the next reboot.
I only have segfault if i run the enlargement pill and then claymore , if i run claymore alone works fine, and other miners also works ok.
i am not overclocking or anything, just set fan speeds.
thanks
indkt.
[EDIT]
It seems to work fine with claymore 9.8, will try other versions and post, i would like to hear from someone else with the same problem (1080 only, 1080Ti is a completely different beast)
|
|
|
|
|
Hello
Someone bring me this machine to check it, and the controller is not responding
i can PING to the default address and i get a response but when i telnet to the unit port 8100 i get connection refused
also connection refused when i try to access the webpage
i programmed a router in the same subnet to test it.
the chains have a steady red light on.
the unit is using 160W at the wall, and the fans are at near full speed (PWM)
at the board there is only 1 green led on.
i se link and activity on the eth port but green and red light are off.
any ideas what could be wrong?
what can i test?
thank you!
indkt.
|
|
|
|
|
Hello
has anyone been able to set two threads to the new xmr-stak multi GPU miner?
I was able to compile it , runs very well on CPU/NVIDIA but for AMD i cannot set more than two threads per GPU which is a requirement for RX VEGA to hash faster.
I have three RX VEGA, this line works perfectly on the OLD xmr-stak-amd (amd only) miner, but it is ignored on the new miner.
"gpu_threads_conf" :
[
{ "index" : 0, "intensity" : 2016, "worksize" : 8, "affine_to_cpu" : false },
{ "index" : 0, "intensity" : 1736, "worksize" : 8, "affine_to_cpu" : false },
{ "index" : 1, "intensity" : 2016, "worksize" : 8, "affine_to_cpu" : false },
{ "index" : 1, "intensity" : 1736, "worksize" : 8, "affine_to_cpu" : false },
{ "index" : 2, "intensity" : 2016, "worksize" : 8, "affine_to_cpu" : false },
{ "index" : 2, "intensity" : 1736, "worksize" : 8, "affine_to_cpu" : false },
],
I'm on windows 10 with the classic blockchain driver, it is not a machine issue, it is the miner, compiled it again but still refuses to create two threads per GPU.
anyone fixed it? any ideas?
thanks!
indkt.
|
|
|
|
|
Hello
I'm trying to make claymore work with VEGA
I get:
AMD Cards available: 2
GPU #0: gfx900, 8176 MB available, 56 compute units
GPU #0 recognized as vega
GPU #1: gfx900, 8176 MB available, 56 compute units
GPU #1 recognized as vega
Cannot build OpenCL program for GPU 0
Cannot build OpenCL program for GPU 1
later it says this:
GPU0, OpenCL error -48 (0) - cannot create DAG on GPU
GPU1, OpenCL error -48 (0) - cannot create DAG on GPU
I'm using lubuntu 1604, with latest rocm opensource drive (which is the only one that works in OpenCL so far with VEGA)
I can mine with the XMR miner perfectly so OpenCL is working!, but for some reason Claymore doesn't see it.
The machine is i3, with 16GB of RAM, SSD, and 2xRX VEGA 56.
any ideas why it doesn't work?
thank you
Indkt.
|
|
|
|
I like this miner.
Cards: 4x EVGA 1080 Ti FTW3
Efficiency: ~4.6 Sols/W
If you run this at 2.8 sols/W you must have very cheap electricity costs.
How did you achieve 4.6 Sols/W  , i can't get anything above 3,85 even with heavy OC and very low TPL. |
|
|
|
after long tests and usage my opinion about dstm for GTX1080ti is here - it is the same as ewbf
why? here is why :
hashrate for one gtx 1080ti aprox 770sol, so 4 x 760 = 3040 sol/s
while with same settings it is 2960 sol on ewbf.
so you it is 2.7% win for dstm. but, there is a dev fee 2% and ewbf we can use without devfee - that means that actual difference is less then 1%
and taking into account, that ewbf is more stable for my rig, the choice is obvious
however! for 2 x gtx1070 on dstm i'm getting 960 sol/s and 920 sol/s with ewbf
so it is 4.3% difference, minus 2% dev fee and we comes to 2.3% benefit.
I can accept that bonus for gtx1070
I second that, I see exactly the same production difference. Only my numbers are different because you are pushing the cards hard in power limit, and I am aiming at top efficiency, but regardless the result is the same, my numbers says 2.2 % difference bettween EWBF and DSTM miner, so taking into the account the 2% fee the production is the same. at least on the 1080Ti's i am testing on the 1080s (non-Tis) I'll post back tomorrow with results, but it seems that production is the same, i'm also don't notice much CPU usage right now, it looks lower than previous versions. BTW: zm miner tends to spike power usage (it's not constant, it goes up and down) the average endup being the same as EWBF but EWBF miner uses constant power, no spikes of any kind (i have an AC current meter bar graph (vu-meter like) to measure this behavior) so in the end no much difference, but as someone said here dstm is active developing it, so I'm more inclined towards using ZM miner for the time being. cheers Indkt. |
|
|
|
Hi dstm! Testing your version since yesterday today i wasn't @ home, ISP went down for about an hour, when I get home (now) miner is working, so it's all good!, the only thing I don't know if it was consuming power or not, it should stop mining if there is no connection , dunno if that happened, i have an AC power meter connected but it does not have log capability (something i gotta do ) will keep testing but so far so good!, rock solid. cheers Indkt. |
|
|
|
|
Show Posts
