OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.0 - page 371.

belcat

newbie

Activity: 13

Merit: 0

Quote from: cimpex on April 16, 2013, 03:42:31 PM

i have a problem with my gigabyte 7950 rig. clocked at 1150/1400 i get about 610 kH/s (LTC) -> thats ok! the problem is the generated heat of the first one in the main pcie slot. while at all other 7950's the fans spins with about 2200 rpm to hold a target of 75°C, the first one runs with 3600 rpm (100%) and can't get below 77°C. same happens, when i use a card from another slot.
any ideas?

Are you saying whatever card is in the first slot gets hotter? I had problems with the cards running hot, so I stuck a small fan to blow right on the card (at the back of the computer, blowing air 90 degrees from the MB). The air it blows ends up in the front part of the case where it leaves. The fan is about 12" in size and blows more air than typical case fans, but is not much noisier than such, just something from the local department store. It dropped the temperature of one card that could barely cool itself (fan runs almost on max) by 8C and now I don't hear this jet engine sound

. Also means the tower has no side panel... but oh well.

itsjustme

member

Activity: 85

Merit: 10

Quote from: os2sam on April 15, 2013, 09:25:19 PM

Quote from: itsjustme on April 15, 2013, 07:25:16 PM

Quote from: os2sam on April 14, 2013, 08:59:15 AM

Quote from: itsjustme on April 14, 2013, 08:48:03 AM

additionally: cgminer "freezes" (cannot even quit it with Q, have to close the windows clicking on the X) once I switch users on windows 8 Sad

Don't switch users. Or if you do stop CGMiner and then reopen it in the other user account.

ouch, can't do this at all Sad

no way around this?

I've never used Win8. But in WinXP/7 when you switch users the background user looses access to the GPU since the foreground desktop has it.

Search through this thread and you'll see it talked about several times.

poclbm can handle it properly: it keeps mining even when i'm logged in with different users :|
is there a way to achieve the same using cgminer? :/

Richy_T

legendary

Activity: 2576

Merit: 2267

1RichyTrEwPYjZSeAYxeiFBNnKC9UjC5k

I think you can run multiple instances of cgminer and have a different pool for each. Just disable whatever GPU is supposed to be going to the other pool.

Or you could do what I do when I need to mine a little on another pool for whatever reason, disable the GPU and run the miner directly.

Marrs

member

Activity: 112

Merit: 10

Quote from: Melfice on April 16, 2013, 04:58:19 PM

Quite a simple question. What is the way to configure and run cgminer so that gpu 0 will work with pool 0 and gpu 1 with pool 1, but not both gpus with the same pool? Preferably command line configuration.

If your hardware supports IOMMU, you could theoretically create two Xen VMs, assign a single gpu to each VM, and run cgminer in each VM with separate pool addresses.

Google it if you care enough to try it, but it's probably more trouble than it's worth. ;-)

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: Melfice on April 16, 2013, 04:58:19 PM

Quite a simple question. What is the way to configure and run cgminer so that gpu 0 will work with pool 0 and gpu 1 with pool 1, but not both gpus with the same pool? Preferably command line configuration.

Cannot.

Melfice

newbie

Activity: 40

Merit: 0

Quite a simple question. What is the way to configure and run cgminer so that gpu 0 will work with pool 0 and gpu 1 with pool 1, but not both gpus with the same pool? Preferably command line configuration.

cimpex

sr. member

Activity: 588

Merit: 252

Quote from: Richy_T on April 16, 2013, 04:14:44 PM

Either airflow or, perhaps more likely, a heat source near slot 1. On my board, the CPU is right next to slot 1.

i'm using riser cables. all cards have the same spacing to each other and get the same air.

Richy_T

legendary

Activity: 2576

Merit: 2267

1RichyTrEwPYjZSeAYxeiFBNnKC9UjC5k

Either airflow or, perhaps more likely, a heat source near slot 1. On my board, the CPU is right next to slot 1.

cimpex

sr. member

Activity: 588

Merit: 252

Quote from: Richy_T on April 16, 2013, 03:52:56 PM

Quote from: cimpex on April 16, 2013, 03:42:31 PM

i have a problem with my gigabyte 7950 rig. clocked at 1150/1400 i get about 610 kH/s (LTC) -> thats ok! the problem is the generated heat of the first one in the main pcie slot. while at all other 7950's the fans spins with about 2200 rpm to hold a target of 75°C, the first one runs with 3600 rpm (100%) and can't get below 77°C. same happens, when i use a card from another slot.
any ideas?

win 7 64-bit (dedicated rig, no other software installed than drivers, winrar, realvnc and trixx)
amd driver 13.1
sdk 2.8
cgminer 2.11.4
no display plugged in, no dummyplugs

thanks!

Poor contact between the heatsink and GPU?

I'm assuming you're not running a tower? I was having "heat rising" issues for a while.

the heat problem is in the first pcie slot only! it doesn't matter which of the 5 7950 is in pcie slot 1, everyone gets hot while in every other slot they run "cool".

Krak

hero member

Activity: 591

Merit: 500

Quote from: cimpex on April 16, 2013, 03:42:31 PM

i have a problem with my gigabyte 7950 rig. clocked at 1150/1400 i get about 610 kH/s (LTC) -> thats ok! the problem is the generated heat of the first one in the main pcie slot. while at all other 7950's the fans spins with about 2200 rpm to hold a target of 75°C, the first one runs with 3600 rpm (100%) and can't get below 77°C. same happens, when i use a card from another slot.
any ideas?

win 7 64-bit (dedicated rig, no other software installed than drivers, winrar, realvnc and trixx)
amd driver 13.1
sdk 2.8
cgminer 2.11.4
no display plugged in, no dummyplugs

thanks!

I noticed the same thing with my dual GPU system. Now that I'm only running one GPU in it, I can't even hear the fan anymore.

Richy_T

legendary

Activity: 2576

Merit: 2267

1RichyTrEwPYjZSeAYxeiFBNnKC9UjC5k

Quote from: cimpex on April 16, 2013, 03:42:31 PM

i have a problem with my gigabyte 7950 rig. clocked at 1150/1400 i get about 610 kH/s (LTC) -> thats ok! the problem is the generated heat of the first one in the main pcie slot. while at all other 7950's the fans spins with about 2200 rpm to hold a target of 75°C, the first one runs with 3600 rpm (100%) and can't get below 77°C. same happens, when i use a card from another slot.
any ideas?

win 7 64-bit (dedicated rig, no other software installed than drivers, winrar, realvnc and trixx)
amd driver 13.1
sdk 2.8
cgminer 2.11.4
no display plugged in, no dummyplugs

thanks!

Poor contact between the heatsink and GPU?

I'm assuming you're not running a tower? I was having "heat rising" issues for a while.

cimpex

sr. member

Activity: 588

Merit: 252

i have a problem with my gigabyte 7950 rig. clocked at 1150/1400 i get about 610 kH/s (LTC) -> thats ok! the problem is the generated heat of the first one in the main pcie slot. while at all other 7950's the fans spins with about 2200 rpm to hold a target of 75°C, the first one runs with 3600 rpm (100%) and can't get below 77°C. same happens, when i use a card from another slot.
any ideas?

win 7 64-bit (dedicated rig, no other software installed than drivers, winrar, realvnc and trixx)
amd driver 13.1
sdk 2.8
cgminer 2.11.4
no display plugged in, no dummyplugs

thanks!

FreedomCoin

hero member

Activity: 675

Merit: 507

Freedom to choose

it seems that CGminer is splitting the load between my two cards so running each one at half speed. When i run one GPU at a time with the -d command i can get my left video card to run full speed. But when i try to run them together the feedback states the mhash but when ending a session i see it was just split between the two cards.. When i try -d 0 and or -d 1 only my left card seems to want to work, and crossfire is not enabled.

This is on an Alienware M17xR2 Laptop with two 5800 Junipers, they should be running at about 185 mhash each

See pics in link. my output is:

Code:

./cgminer -o stratum.ozco.in:80 -u user -p pass -I 8 -d 0 -o stratum.ozco.in:80 -u user -p pass -I 8 -d 1  --gpu-engine 820,820 --gpu-memclock 1000,1000 --temp-target 80,80 --temp-overheat 90,90

See screenshots...

Joshwaa

hero member

Activity: 497

Merit: 500

Quote from: Richy_T on April 15, 2013, 10:10:01 PM

Quote from: Joshwaa on April 15, 2013, 10:04:53 PM

Ill give it a try. But I am on 13.1 and I have 4 other rigs on 13.1
Report back in a bit.

Was cgminer running fine for you on 13.1? For me, it was giving a blank screen and hashing but producing dead blocks (then again, I did try jumping straight to the beta drivers first thing). I may try upgrading again later but it'll be a while.

Yes running fine on 13.1 and one is running 13.2 Beta7, both work well.

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: avirdoz on April 15, 2013, 02:33:20 PM

Hello. I'm using Xubuntu and cgminer 2.11.4 with 3x 7950. Help me please how can i control voltages from console if it's possible? I try do it with cgminer but no effect. Thanks.

Cannot control voltage on 79xx in linux.

bitpop

legendary

Activity: 2912

Merit: 1060

stratum+tcp://[2001:41d0:2:4f54:1189:1377:71d:8cf7]:3333 didnt work either. Its ok though. No hurry like you said.

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: bitpop on April 15, 2013, 11:47:23 PM

Maybe I need stratum+tcp6?

Nope that does not work. Try getting the actual ipv6 ip address and put it in in full notation bracketed etc. I recall only testing it when I first put in stratum support and it worked with ipv6 since it uses libcurl to set up the sockets and libcurl should support ipv6. Personally I don't think it's worth the effort when all pools are still on ipv4 and ipv6 remains slower.

bitpop

legendary

Activity: 2912

Merit: 1060

Maybe I need stratum+tcp6?

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: ralree on April 15, 2013, 11:24:12 PM

Quote from: -ck on April 15, 2013, 07:47:19 PM

Quote from: ralree on April 15, 2013, 07:17:42 PM

I've encountered a bug with the latest github trunk compiled for linux 32-bit:

Update 2: Just tried reverting to tag v2.11.4. It does not seem to have the issue. Not sure what changed.

Code was in flux with a stratum socket closing bug that tried to close fd 0 till yesterday. How recently did you merge with master?

It was quite a recent checkout - 2 days ago maybe...

Yah and as I said it was only fixed yesterday.

bitpop

legendary

Activity: 2912

Merit: 1060

I do

I just didn't include it here. Can anyone else successfully use ipv6?

Quote from: kano on April 15, 2013, 11:14:42 PM

Quote from: bitpop on April 15, 2013, 10:53:21 PM

Why isnt cgminer working with:

"pools" : [
{
"url" : "stratum+tcp://ipv6.stratum.bitcoin.cz:3333",

A ping resolves fine, my ipv6 is fine.

You can't have a comma at the end, you need a username and password, and you have to close the brackets ... Cheesy

Topic: OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.0 - page 371. (Read 5805537 times)