Try CUDA 9.1, you'll slow it down with 9.2 unless you know what you're doing.
Cuda 9 fail.
Right now im working on cuda 8:
Sneak Preview
Edit***
Fail again.
The dirty results
Topic: XMR-stak-JK 2.10.7 Compiled with no devfee (Updated: 2019-8-18) - page 11. (Read 28606 times)
legendary
Activity: 1848
Merit: 1166
My AR-15 ID's itself as a toaster. Want breakfast?
Cuda 9 fail.
Right now im working on cuda 8:
Sneak Preview
Edit***
Fail again.
The dirty results
@JaredKaragen Did you ever compile XMR-STAK-2.4.8 ? Thx
2.4.5 is all that fierce-uk has released thus far..... is it another fork or something?
Update:
Im working on my 5th attempt on compiling 2.4.5. VS 2017 is uninstalling as I type this, and Ill have to rip out all IDE's and start over by installing fresh... again.
Probably going to DL and install the latest version of the cuda toolkit as well.... hopefully they play nice.
Try CUDA 9.1, you'll slow it down with 9.2 unless you know what you're doing.
legendary
Activity: 1848
Merit: 1166
My AR-15 ID's itself as a toaster. Want breakfast?
@JaredKaragen Did you ever compile XMR-STAK-2.4.8 ? Thx
2.4.5 is all that fierce-uk has released thus far..... is it another fork or something?
Update:
Im working on my 5th attempt on compiling 2.4.5. VS 2017 is uninstalling as I type this, and Ill have to rip out all IDE's and start over by installing fresh... again.
Probably going to DL and install the latest version of the cuda toolkit as well.... hopefully they play nice.
@JaredKaragen
Thanks for putting this together. I still want to compile my own, but it's been years since I've done any coding. I'm sure I'll get a wild hair on a rainy day come winter.
Thanks for putting this together. I still want to compile my own, but it's been years since I've done any coding. I'm sure I'll get a wild hair on a rainy day come winter.
I don't do any coding also, and if you follow instructions on XMR-Stak Github, you'll get there.
Worst part is installing Gb's of software - VB Studio, CUDA and AMD OpenCL libraries.
My workstation's .5tb ssd shouldn't mind, but yeah, mountain biking weather... nobody got time for that.
@JaredKaragen
Thanks for putting this together. I still want to compile my own, but it's been years since I've done any coding. I'm sure I'll get a wild hair on a rainy day come winter.
Thanks for putting this together. I still want to compile my own, but it's been years since I've done any coding. I'm sure I'll get a wild hair on a rainy day come winter.
I don't do any coding also, and if you follow instructions on XMR-Stak Github, you'll get there.
Worst part is installing Gb's of software - VB Studio, CUDA and AMD OpenCL libraries.
@JaredKaragen
Thanks for putting this together. I still want to compile my own, but it's been years since I've done any coding. I'm sure I'll get a wild hair on a rainy day come winter.
Thanks for putting this together. I still want to compile my own, but it's been years since I've done any coding. I'm sure I'll get a wild hair on a rainy day come winter.
legendary
Activity: 1848
Merit: 1166
My AR-15 ID's itself as a toaster. Want breakfast?
snip / original message deleted.
this is not a thread for 4 post users to post a binary...
or harass people for donations.
get that crap out of here.
legendary
Activity: 1848
Merit: 1166
My AR-15 ID's itself as a toaster. Want breakfast?
@JaredKaragen Did you ever compile XMR-STAK-2.4.8 ? Thx
not yet... I guess its been releasing a lot of updates.Im gonna hopefully try the VS2015 route later today... but I need sleep... havent slept yet... was a busy day yesterday.
@JaredKaragen Did you ever compile XMR-STAK-2.4.8 ? Thx
thanks for making this. I tried to follow instructions, but I had trouble with visual studio since I already have a few versions of it installed, the cuda downloader for visual studio integration failed.
Yours worked instantly! 2 ez
you are welcome.Yours worked instantly! 2 ez
I have expeirenced a very strange connect error to nanopool on only one of my machines.... Anyone else getting this issue?
I'm not sure what the hell happens (because so far I have been asleep each time) but I have some issue which seems to crash three instances of the miner (2.4.4) running on two machines. Last night it was after running great for 16 hours, mining Loki (which has a very low difficulty right now which is good, but has taken a price hit - but I am HODLing anyway).
Did you manage to compile 2.4.5 yet? I was going to try to teach myself how sometime but it's like going back to f'ing school.....
no luck on trying with 2.4.5 yet. Probably sometime tomorrow since Ill be at a location where I can uninstall and reinstall everything for the third time to try and compile it.
The issue was a fluke; seems right before the completion of the new windows 18xx update. I tried running 2.4.2, and it gave the same no connect error.
When I came home later that day my machine had rebooted itself automatically, and 2.4.4 launched no problems again. It was an odd one.
Go into the CPU config, and remove the CPU threads.... leave it blank brackets {}. See if stability comes back. I have a LGA775 machine right now that gets TLS socket errors and reconnects every so often. The auto emails from nanopool from that machine are quite annoying.
I was also sick of win 10 updates and all that "telemetry" running. I now use win 8.1 on all my rigs. xmr-stak runs OK for Nvidia GPU mining - I get up to 740h/s per gtx1070 (30 threads, 60 blocks, core 0, memory +750, powerlimit 75%).
@JaredKaragen - did you ever manage to compile 2.4.5? I really need to teach myself all that is necessary to compile for Windows or even maybe run linux.
thanks for making this. I tried to follow instructions, but I had trouble with visual studio since I already have a few versions of it installed, the cuda downloader for visual studio integration failed.
Yours worked instantly! 2 ez
you are welcome.Yours worked instantly! 2 ez
I have expeirenced a very strange connect error to nanopool on only one of my machines.... Anyone else getting this issue?
I'm not sure what the hell happens (because so far I have been asleep each time) but I have some issue which seems to crash three instances of the miner (2.4.4) running on two machines. Last night it was after running great for 16 hours, mining Loki (which has a very low difficulty right now which is good, but has taken a price hit - but I am HODLing anyway).
Did you manage to compile 2.4.5 yet? I was going to try to teach myself how sometime but it's like going back to f'ing school.....
no luck on trying with 2.4.5 yet. Probably sometime tomorrow since Ill be at a location where I can uninstall and reinstall everything for the third time to try and compile it.
The issue was a fluke; seems right before the completion of the new windows 18xx update. I tried running 2.4.2, and it gave the same no connect error.
When I came home later that day my machine had rebooted itself automatically, and 2.4.4 launched no problems again. It was an odd one.
Go into the CPU config, and remove the CPU threads.... leave it blank brackets {}. See if stability comes back. I have a LGA775 machine right now that gets TLS socket errors and reconnects every so often. The auto emails from nanopool from that machine are quite annoying.
nice thread here, looking forward 2.4.5 version
I have complied for amd and running with my vegas on my farm. I will share with you folks tomorrow when I make sure the stability.
nice thread here, looking forward 2.4.5 version
I really appreciate if you can compile the latest version 2.4.5. Thanks
Read post #92, ya lazy newb.
lol, yeah, ive been under the weather the past week and my life has definitely not been normal. I am hoping to get into work today and work late to try and make up for lost time.
As soon as I am able to sit down at a high speed internet location, ill attempt another recompile.
I noticed with 2.4.4 my cuda libraries would not be able to use the old gtx 5xx series and older.... no bueno. I may have to remove cuda 9, and reinstall cuda 7.5 and cuda 8, and see if I can get support back in the 4.4.5 version. I attempted swapping for my old cuda_backend.dll file that worked; and it crashed the app, so I am assuming that happens because the new algos and functions added to cuda_backend.dll were moved around and changed somehow making the old library foreign language to the new app build.
Ill be walking into uncharted territory, because I believe to compile with cuda 7.5 and 8, Ill need to install VS2015.... so yeah. fingers crossed I guess.
I just wish this crap would continue working after the first time you reboot the machine.
Not helping that im doing it on a laptop, but I ritually hibernate it now. just cause.
Crossed my fingers... I wish I would help you...
hope you will get it fixed... legendary
Activity: 1848
Merit: 1166
My AR-15 ID's itself as a toaster. Want breakfast?
I really appreciate if you can compile the latest version 2.4.5. Thanks
Read post #92, ya lazy newb.
lol, yeah, ive been under the weather the past week and my life has definitely not been normal. I am hoping to get into work today and work late to try and make up for lost time.
As soon as I am able to sit down at a high speed internet location, ill attempt another recompile.
I noticed with 2.4.4 my cuda libraries would not be able to use the old gtx 5xx series and older.... no bueno. I may have to remove cuda 9, and reinstall cuda 7.5 and cuda 8, and see if I can get support back in the 4.4.5 version. I attempted swapping for my old cuda_backend.dll file that worked; and it crashed the app, so I am assuming that happens because the new algos and functions added to cuda_backend.dll were moved around and changed somehow making the old library foreign language to the new app build.
Ill be walking into uncharted territory, because I believe to compile with cuda 7.5 and 8, Ill need to install VS2015.... so yeah. fingers crossed I guess.
I just wish this crap would continue working after the first time you reboot the machine.
Not helping that im doing it on a laptop, but I ritually hibernate it now. just cause.
I really appreciate if you can compile the latest version 2.4.5. Thanks
Read post #92, ya lazy newb.
thanks for making this. I tried to follow instructions, but I had trouble with visual studio since I already have a few versions of it installed, the cuda downloader for visual studio integration failed.
Yours worked instantly! 2 ez
you are welcome.Yours worked instantly! 2 ez
I have expeirenced a very strange connect error to nanopool on only one of my machines.... Anyone else getting this issue?
I'm not sure what the hell happens (because so far I have been asleep each time) but I have some issue which seems to crash three instances of the miner (2.4.4) running on two machines. Last night it was after running great for 16 hours, mining Loki (which has a very low difficulty right now which is good, but has taken a price hit - but I am HODLing anyway).
Did you manage to compile 2.4.5 yet? I was going to try to teach myself how sometime but it's like going back to f'ing school.....
no luck on trying with 2.4.5 yet. Probably sometime tomorrow since Ill be at a location where I can uninstall and reinstall everything for the third time to try and compile it.
The issue was a fluke; seems right before the completion of the new windows 18xx update. I tried running 2.4.2, and it gave the same no connect error.
When I came home later that day my machine had rebooted itself automatically, and 2.4.4 launched no problems again. It was an odd one.
Go into the CPU config, and remove the CPU threads.... leave it blank brackets {}. See if stability comes back. I have a LGA775 machine right now that gets TLS socket errors and reconnects every so often. The auto emails from nanopool from that machine are quite annoying.
I really appreciate if you can compile the latest version 2.4.5. Thanks
Eventually I solved this issue by running separate instances of XMR-stak. 1st one running the 1080 and the 1st 2 1070s plus cpu core 0, second one running the remaining two 1070s and cpu core 2. With this setup, now finally all five cards are at 650-700h/s. It took quite a bit of messing with the configs of the second instance to get it to even run, and I have to set it to verbosity 4 and 30secs because pressing h doesn't respond and give me hash-rate.
If anyone else is running a 4 or more GPU rig (especially if you use 1070s/1080s) and you aren't getting full hash rate on all GPUs I suggest trying what I did - it works great.
Like I said, getting the second instance of xmr-stak running on the same machine means changing quite a few things in the configuration.txt so if anyone is interested I will try to remember what I did.
Great tip, it should be on a top 10 list. I 'accidentally' found this too with my rig with mis-matched cards, that running different instances of the miner for each card(s) helps a lot.
so are you running an individual instance for every card? does that give you more stability? I have divided the work up into 1 CPU core & 3 GPUs on one instance and 1 CPU core 2 and 2 GPUs on a second instance, but like I said in my previous post - it all still seems to crash eventually. Sometimes it manages to live for a long time and other times only a few hours, but this is also happening on a second machine with only 3 GPUs (one miner instance). I was beginning to suspect that it's OhGodAnEthLargementPill I am running for better 1080 performance rather than xmr-stak that is the issue, but I don't know. I guess I need to make it keep a log and dig into that.
It seems that the majority of cryptonight miners are more AMD focussed so I guess there is a lack of stability, but unfortunately the only ccminer that does cryptonight hasn't updated to the newer ASIC restistant variations of the algorithm like cryptonight-heavy.
Well, no, not for every card, only if one card looked like it needed special treatment.
I also ran different instances of a miner if I wanted to spread the risk and mine different coins on the same rig.
legendary
Activity: 1848
Merit: 1166
My AR-15 ID's itself as a toaster. Want breakfast?
thanks for making this. I tried to follow instructions, but I had trouble with visual studio since I already have a few versions of it installed, the cuda downloader for visual studio integration failed.
Yours worked instantly! 2 ez
you are welcome.Yours worked instantly! 2 ez
I have expeirenced a very strange connect error to nanopool on only one of my machines.... Anyone else getting this issue?
I'm not sure what the hell happens (because so far I have been asleep each time) but I have some issue which seems to crash three instances of the miner (2.4.4) running on two machines. Last night it was after running great for 16 hours, mining Loki (which has a very low difficulty right now which is good, but has taken a price hit - but I am HODLing anyway).
Did you manage to compile 2.4.5 yet? I was going to try to teach myself how sometime but it's like going back to f'ing school.....
no luck on trying with 2.4.5 yet. Probably sometime tomorrow since Ill be at a location where I can uninstall and reinstall everything for the third time to try and compile it.
The issue was a fluke; seems right before the completion of the new windows 18xx update. I tried running 2.4.2, and it gave the same no connect error.
When I came home later that day my machine had rebooted itself automatically, and 2.4.4 launched no problems again. It was an odd one.
Go into the CPU config, and remove the CPU threads.... leave it blank brackets {}. See if stability comes back. I have a LGA775 machine right now that gets TLS socket errors and reconnects every so often. The auto emails from nanopool from that machine are quite annoying.
Eventually I solved this issue by running separate instances of XMR-stak. 1st one running the 1080 and the 1st 2 1070s plus cpu core 0, second one running the remaining two 1070s and cpu core 2. With this setup, now finally all five cards are at 650-700h/s. It took quite a bit of messing with the configs of the second instance to get it to even run, and I have to set it to verbosity 4 and 30secs because pressing h doesn't respond and give me hash-rate.
If anyone else is running a 4 or more GPU rig (especially if you use 1070s/1080s) and you aren't getting full hash rate on all GPUs I suggest trying what I did - it works great.
Like I said, getting the second instance of xmr-stak running on the same machine means changing quite a few things in the configuration.txt so if anyone is interested I will try to remember what I did.
Great tip, it should be on a top 10 list. I 'accidentally' found this too with my rig with mis-matched cards, that running different instances of the miner for each card(s) helps a lot.
so are you running an individual instance for every card? does that give you more stability? I have divided the work up into 1 CPU core & 3 GPUs on one instance and 1 CPU core 2 and 2 GPUs on a second instance, but like I said in my previous post - it all still seems to crash eventually. Sometimes it manages to live for a long time and other times only a few hours, but this is also happening on a second machine with only 3 GPUs (one miner instance). I was beginning to suspect that it's OhGodAnEthLargementPill I am running for better 1080 performance rather than xmr-stak that is the issue, but I don't know. I guess I need to make it keep a log and dig into that.
It seems that the majority of cryptonight miners are more AMD focussed so I guess there is a lack of stability, but unfortunately the only ccminer that does cryptonight hasn't updated to the newer ASIC restistant variations of the algorithm like cryptonight-heavy.
Jump to: