i have problems with teamviewer :)) …celeron 1840 …8 gb ram
Confirmed that the 0.3.3 version is stable on my ubuntu 16.04 server with 6 RX 480 and the hash rate looks good.
yeah, optiminer is in the same league and i don’t want to babysit my rigs.
In any case, eth is almost exactly the same $$ right now.
same $ but not same power draw …
Same, I have been running 5 RX 480 stock since its release and no crashes. Good to know b/c I am going to be gone for a week!
Any new miner from Optiminer to beat Claymore hasrate?
me 2 , iver hade a couple system crashes, and had to reinstall windows once already due to a horrid crash
Claymoreing for 2 days straight. (ok almost, had to restart my rigs today due to Windows forced update. )
No problems.
Temps got a little bit high though.
The RigPC does not pick up the temps of the 2nd card in the rig!?
I guess it is about 70C also.
Version 0.3.4 released!
Fixes a deadlock in the stratum client that many people have reported.
Also adds a GPU watchdog that allows to execute a command when any GPU gets stuck. The only solution to reliably reset a GPU seems to be to reboot the system. See ‘mine.sh’ and ‘watchdog-cmd.sh’ for examples.
1 Like
Performance improvement?
watchdog-cmd.sh example missing in tar.gz file
Nice miner you built !
Thanks for reporting. ‘watchdog-cmd.sh’ added to the archive.
No speed improvements for this release.
Optimizer can you improve the hash-rate soon? Because Claymore “eat us all the pie”. Thank’s for your work.
1 Like
finally got optiminer working in the simpleminer.com OS.
1 rig with 2 x 290, 2 x 290X, 1 x 390 860 sol
1 rig with 3 x 290, 2 x 380 620 sol
BS. Big farms also care about monitoring tools, ease of use, reliability, and OCing. There is no upside to using Nix right now besides occasionally devs limiting builds to nix for shits and giggles (then concurrently running into a bunch of broken dependencies and things not compiling correctly).
Hmm, all of those except maybe OCing which I agree is easier at the moment on Windows, I would count towards Linux. It’s far easier to launch (and keep launched), maintain consistency, is more reliable, has way better monitoring tools… and ease of use if you know what you are doing. If you have problems with dependencies and compiling you need to switch distros, not to Windows. Windows 10 has horrible nagging updates that kill uptime (hence $$) and are harder and harder to disable.
1 Like
upside of windows, Claymore.
True, but it’s been leapfrog between optiminer and claymore. Optiminer was beating Claymore for awhile as well.
I would agree that long term management of Nix is easier. In a enterprise setting where you just want to turn something on and leave it running for a year or years on end without ever having to touch it, it’s definitely a great solution. Cryptos are constantly changing and evolving. There are a ton of people that want monitoring, OCing, and watchdog controls in nix so they ask someone like Claymore to add it. That’s why his miner is so robust, because it’s a absolute pain to do that stuff in nix.
Cryptos change though and there are a lot of them, Claymore isn’t everywhere, he can’t go to every algo and every piece of hardware. Windows is just way more robust for mining in addition to allowing someone of pretty much any skill level to fix it when it’s broken. Crypto mining requires constant attention and constantly changes. How many different releases have went through Equihash in the last week? The same OCs you use for instance in Ethereum no longer work in Equihash due to the algos stressing different parts of your GPU, everything needs to be changed.
That’s not something you can go and do in nix easily. People are just used to the idea of nix being superb for anything ‘big’ as far as computing goes. Windows isn’t bad especially when you configure it correctly (like people complaining about restarts because of updates don’t know how to manage policies or services, something a advanced windows user easily knows how to do), it just has a huge stigma associated with it.
I knew you would mention group policy but you do know Microsoft makes that a moving target with 10. My last Pro copy I run at home did have ways to turn off updates now it’s degraded down to only being able to postpone… no telling what will be next. You need Enterprise versions now to control it (or hacks). I agree different user demographics require different platforms that those users are already use to/running. The original comment was about big farms though and with Windows licensing costs, update upkeep, non stable configuration unless you block said updates, I believe we need more focus on just making better linux gpu oc tools. AMD will be launching their power management and other tools for Linux (according to them) so that might be another way to bring it up to par on that side.
I don’t know why anyone would want to pay for an OS to run something that makes just as much money on something free especially if it can be all made automated. Not to mention supporting opensource but that wasn’t brought up so I’ll leave it alone.
Anyway, sorry for rambling. I have one helpful link and one tip for people wanting to do power / clock management in linux: GitHub - DominiLux/amdgpu-pro-fans: An alpha version written in bash script for the development of overall concepts for a c++ project I will be releasing soon as open source. - It helps you script fan control and also if you want to change clocks you need to first echo “manual” into /sys/class/drm/cardX/device/power_drm_force_performance_level then changing the gpu clock is done in the file pp_drm_sclk in that same device directory. cat it first to see what speeds are available and then echo the number (ex: 5) into it to change it. I’ve noticed that optiminer will adjust this after you start it so you need to set it again after running optiminer.