Topic: Consolidated Litecoin Mining Guide for 5xxx, 6xxx, and 7xxx GPUs - page 2. (Read 624200 times)
December 28, 2013, 09:58:04 PM
I have an old "mini" box with a wi-fi card, and I built it to run Apple OSX about 5 years ago. It only has a single CPU without a fan (about 1.0 Ghz), but do you think if I installed Ubuntu on it I could run my BTC cube miner off of it and free up my Litecoin box? I think you can run mining_proxy and it won't even mess up your meager video card and such, on an old box like that. It would just be the terminal window running.
December 28, 2013, 08:45:24 PM
Thanks to you all for the input, especially Ivanlabrie, I finally got it to work best with this, after doing extensive research:
./cgminer --scrypt --url=stratum+tcp://xxxxpool.com:3333 --userpass=xxxx:x -I 13 -g 2 -w 256 --lookup-gap=2 --temp-target=75 --auto-fan --no-submit-stale --gpu-engine=1070 --gpu-memclock=1800 --gpu-powertune=20
Some of the other parameters are redundant, even thread-concurrency & shaders, but I got it to mine steady (the R9 280X / 7970). They say every rig is slightly different so you have to play around.
It now mines Litecoins at a steady ~650 KHs rate.
What I didn't mention was that when trying also mining BTC on an external Block Erupter (mining_proxy.py), and even though it's external, the console and system memory takes a hit from that and dropped my LTC mining down quite a bit to under 500 again. Then I tried using the -q (quiet) option on the mining_proxy (which quiets but doesn't eliminate all output), so now I'm back to the 610+ range.
I only have 4 gig memory installed on that box right now, but will have 8 Gb soon, so hopefully that will improve things.
./cgminer --scrypt --url=stratum+tcp://xxxxpool.com:3333 --userpass=xxxx:x -I 13 -g 2 -w 256 --lookup-gap=2 --temp-target=75 --auto-fan --no-submit-stale --gpu-engine=1070 --gpu-memclock=1800 --gpu-powertune=20
Some of the other parameters are redundant, even thread-concurrency & shaders, but I got it to mine steady (the R9 280X / 7970). They say every rig is slightly different so you have to play around.
It now mines Litecoins at a steady ~650 KHs rate.
What I didn't mention was that when trying also mining BTC on an external Block Erupter (mining_proxy.py), and even though it's external, the console and system memory takes a hit from that and dropped my LTC mining down quite a bit to under 500 again. Then I tried using the -q (quiet) option on the mining_proxy (which quiets but doesn't eliminate all output), so now I'm back to the 610+ range.
I only have 4 gig memory installed on that box right now, but will have 8 Gb soon, so hopefully that will improve things.
December 28, 2013, 04:41:42 PM
what is your card setup to get 1422k/hs? Two 5970s?
Yes mine is running stable too. One 5970 and one 5850 yielding ~1,125 khs. My 5970 seems to like a little OC'ing (gpu 835/mem 1100), likes 8192 TC, and also i'm not sure whether it likes 17 or 18 intensity better. 17 = more stable but 18 may get a little more hash? more volatile tho.
My only concern will be when i get a 6xxx plugged in i'm sure things will go haywire. Will have to do a -d X command to work on the 6xxx card solo, then do a gpu-fan command when running them all in tandem to identify who is who, then hopefully my batch file will make it work. Because I have no idea how to run a successful .conf file like ivan mentioned.
Yes mine is running stable too. One 5970 and one 5850 yielding ~1,125 khs. My 5970 seems to like a little OC'ing (gpu 835/mem 1100), likes 8192 TC, and also i'm not sure whether it likes 17 or 18 intensity better. 17 = more stable but 18 may get a little more hash? more volatile tho.
My only concern will be when i get a 6xxx plugged in i'm sure things will go haywire. Will have to do a -d X command to work on the 6xxx card solo, then do a gpu-fan command when running them all in tandem to identify who is who, then hopefully my batch file will make it work. Because I have no idea how to run a successful .conf file like ivan mentioned.
yes im running two of 5970. actually i can get about 1520khs but it is risky. and crashing too often.
1422khs running stable and no crash anymore about a day. hope it wont.
actually i got headaches to make this rig stable. got many errors, many crashes, cards was closer to be burned.
im noob and running this shits made me stress. but with researching threads, getting help from experienced users. it s now okey.
i also started to use cgwatcher it has some options about restarting cgminer if a gpu crashes. dead or sick gpu it also restart cgminer. so when u r sleeping or at outside it can help.
also it has some features about temp. protect. if your card gets 85+ temp. its auto lowering -I
etc etc etc.
i got many crashes with high oc'ing so i lowered core clock and memory clock. I also advice u to not make much OC.
December 28, 2013, 03:25:41 PM
what is your card setup to get 1422k/hs? Two 5970s?
Yes mine is running stable too. One 5970 and one 5850 yielding ~1,125 khs. My 5970 seems to like a little OC'ing (gpu 835/mem 1100), likes 8192 TC, and also i'm not sure whether it likes 17 or 18 intensity better. 17 = more stable but 18 may get a little more hash? more volatile tho.
My only concern will be when i get a 6xxx plugged in i'm sure things will go haywire. Will have to do a -d X command to work on the 6xxx card solo, then do a gpu-fan command when running them all in tandem to identify who is who, then hopefully my batch file will make it work. Because I have no idea how to run a successful .conf file like ivan mentioned.
Yes mine is running stable too. One 5970 and one 5850 yielding ~1,125 khs. My 5970 seems to like a little OC'ing (gpu 835/mem 1100), likes 8192 TC, and also i'm not sure whether it likes 17 or 18 intensity better. 17 = more stable but 18 may get a little more hash? more volatile tho.
My only concern will be when i get a 6xxx plugged in i'm sure things will go haywire. Will have to do a -d X command to work on the 6xxx card solo, then do a gpu-fan command when running them all in tandem to identify who is who, then hopefully my batch file will make it work. Because I have no idea how to run a successful .conf file like ivan mentioned.
December 28, 2013, 01:45:09 PM
5970 won't show the specific GPU stats in CG miner 3.7.2 on my win8 machine, just a black screen. Only can see summary stats if I do a '-T' command in the batch file, but this is a pain as I can't optimize very easily. Anyone else run into this issue and how were you able to resolve?
Thanks
Thanks
same here. i couldnt find a way...
i m using -T to see details. it only for 5970s
Thankfully I'm not the only one.....although it sucks that you are dealing with this too. Out of curiosity, what drivers/SDK and OS are you using? I'm on win8.1 32bit, 13.11 v9.5 beta driver, and latest (2.9? sdk). At the end of the day I think it may be a driver issue but when I tried installing some 12.x drivers my system started going bezerk.
windows 7 64bit home premium
app sdk 2.9
ati cata. 13.9
actually it is running stable. just should work with -T
-I 18 --thread-concurrency 8000 -g 1 -w 256 -T
i have an other rig with 5x 7950s and it runs smooth. no need to add -T
as i said its fully about 5970s. i dont think its about drivers. because i tried many different drivers.
only that drivers worked stable for me.
December 28, 2013, 01:07:32 PM
5970 won't show the specific GPU stats in CG miner 3.7.2 on my win8 machine, just a black screen. Only can see summary stats if I do a '-T' command in the batch file, but this is a pain as I can't optimize very easily. Anyone else run into this issue and how were you able to resolve?
Thanks
Thanks
same here. i couldnt find a way...
i m using -T to see details. it only for 5970s
Thankfully I'm not the only one.....although it sucks that you are dealing with this too. Out of curiosity, what drivers/SDK and OS are you using? I'm on win8.1 32bit, 13.11 v9.5 beta driver, and latest (2.9? sdk). At the end of the day I think it may be a driver issue but when I tried installing some 12.x drivers my system started going bezerk.
December 28, 2013, 09:28:17 AM
5970 won't show the specific GPU stats in CG miner 3.7.2 on my win8 machine, just a black screen. Only can see summary stats if I do a '-T' command in the batch file, but this is a pain as I can't optimize very easily. Anyone else run into this issue and how were you able to resolve?
Thanks
Thanks
same here. i couldnt find a way...
i m using -T to see details. it only for 5970s
December 27, 2013, 08:03:30 PM
Conf file I made isn't doing any better.....any suggestions? Honestly have never used the conf file before, just batch file.
(this is for testing out a solo 5970)
{
"pools" : [
{
"url" : "http://stratum.SWEETPOOL.ORG:3333",
"user" : "USER.WORKER",
"pass" : "WORKERPASS"
}
]
,
"intensity" : "11,11",
"thread-concurrency" : "8192,8192",
"vectors" : "1,1",
"worksize" : "256,256",
"kernel" : "scrypt,scrypt",
"gpu-fan" : "0-75,0-75",
"gpu-vddc" : "1.088,1.088",
"gpu-engine" : "0-820,0-820",
"gpu-memclock" : "1100,1100",
"gpu-memdiff" : "0,0",
"gpu-powertune" : "0,0",
"temp-cutoff" : "80,80",
"temp-overheat" : "90,90",
"temp-target" : "70,70",
"api-port" : "4028",
"expiry" : "120",
"gpu-dyninterval" : "7",
"gpu-threads" : "1",
"log" : "5",
"queue" : "1",
"retry-pause" : "5",
"scan-time" : "60",
"scrypt" : true,
"temp-hysteresis" : "3",
"shares" : "0",
"api-listen" : true,
"api-network" : true,
"api-port" : "4028",
"kernel-path" : "/usr/local/bin"
}
(this is for testing out a solo 5970)
{
"pools" : [
{
"url" : "http://stratum.SWEETPOOL.ORG:3333",
"user" : "USER.WORKER",
"pass" : "WORKERPASS"
}
]
,
"intensity" : "11,11",
"thread-concurrency" : "8192,8192",
"vectors" : "1,1",
"worksize" : "256,256",
"kernel" : "scrypt,scrypt",
"gpu-fan" : "0-75,0-75",
"gpu-vddc" : "1.088,1.088",
"gpu-engine" : "0-820,0-820",
"gpu-memclock" : "1100,1100",
"gpu-memdiff" : "0,0",
"gpu-powertune" : "0,0",
"temp-cutoff" : "80,80",
"temp-overheat" : "90,90",
"temp-target" : "70,70",
"api-port" : "4028",
"expiry" : "120",
"gpu-dyninterval" : "7",
"gpu-threads" : "1",
"log" : "5",
"queue" : "1",
"retry-pause" : "5",
"scan-time" : "60",
"scrypt" : true,
"temp-hysteresis" : "3",
"shares" : "0",
"api-listen" : true,
"api-network" : true,
"api-port" : "4028",
"kernel-path" : "/usr/local/bin"
}
December 27, 2013, 04:36:45 PM
Touche'. I read 290. Need...more...sleep.
In all seriousness, anyone have any idea on how to get the individual GPUs to read out in CGMiner on a 5970? (per my above post) This has me baffled. Its hashing well now (~750 total) but when I toss in a 6xxx this weekend i'm sure optimizing is going to be a bee-hatch.
In all seriousness, anyone have any idea on how to get the individual GPUs to read out in CGMiner on a 5970? (per my above post) This has me baffled. Its hashing well now (~750 total) but when I toss in a 6xxx this weekend i'm sure optimizing is going to be a bee-hatch.
Just use a conf file, and use commas...you can figure it out easily that way.
December 26, 2013, 09:55:21 PM
Touche'. I read 290. Need...more...sleep.
In all seriousness, anyone have any idea on how to get the individual GPUs to read out in CGMiner on a 5970? (per my above post) This has me baffled. Its hashing well now (~750 total) but when I toss in a 6xxx this weekend i'm sure optimizing is going to be a bee-hatch.
In all seriousness, anyone have any idea on how to get the individual GPUs to read out in CGMiner on a 5970? (per my above post) This has me baffled. Its hashing well now (~750 total) but when I toss in a 6xxx this weekend i'm sure optimizing is going to be a bee-hatch.
December 26, 2013, 07:32:23 PM
why only intensity 13? Should be able to increase that and subsequently your k/hs. I'd go with at least 16 (after your GPU/TC settings are dialed)
Because it's a 280X (or a 7970 rebrand with a better bios and ram), they like i 13 and g 2 dude. :p
December 25, 2013, 11:50:47 PM
why only intensity 13? Should be able to increase that and subsequently your k/hs. I'd go with at least 16 (after your GPU/TC settings are dialed)
December 25, 2013, 03:57:31 PM
Regarding the above post, all is working well, but after 2 days, now I'm only averaging in the upper 500KHs range. The GPU stays at 72-75 C, and upon a restart in the console, it goes back to the same range, not higher as originally. What would be the best thing to tweak without making it hotter? For example, thread concurrency > 8192?
Try this conf:
"intensity" : "13, 13",
"worksize" : "256, 256",
"kernel" : "scrypt",
"lookup-gap" : "2,2",
"thread-concurrency" : "22336,22336",
"shaders" : "2048,2048",
"gpu-fan" : "30-90,30-90",
"gpu-engine" : "1070,1070",
"gpu-memclock" : "1500,1500",
"gpu-memdiff" : "0,0",
"gpu-powertune" : "20,20",
"temp-cutoff" : "86,86",
"temp-overheat" : "78,78",
"temp-target" : "70,70",
"api-listen" : true,
"api-port" : "4028",
"auto-fan" : true,
"auto-gpu" : false,
"expiry" : "120",
"gpu-dyninterval" : "7",
"gpu-threads" : "2",
December 25, 2013, 02:45:24 PM
Regarding the above post, all is working well, but after 2 days, now I'm only averaging in the upper 500KHs range. The GPU stays at 72-75 C, and upon a restart in the console, it goes back to the same range, not higher as originally. What would be the best thing to tweak without making it hotter? For example, thread concurrency > 8192?
December 24, 2013, 01:57:56 PM
I just built a new rig yesterday with a single R9 280X on Ubuntu Saucy (13.10) - bgfminer didn't do the trick, so I installed cgminer 3.7.2 (the "built" version) into my home folder and with the help of all the posts above I put together a script. If you see anything to improve on, let me know, but it keeps hashing just above 700KHs and now about 17 hours later without a restart, it's still going +/- 660KHs. And the GPU temp stays under 75 C.
Of course I also installed fglrx for Ubuntu and did the initialization, reboots, etc.-- It does include the amdcccle app now.
Wanted to share the results:
export GPU_MAX_ALLOC_PERCENT=100
./cgminer --scrypt --url=stratum+tcp://MYCHOSENPOOL.com:3333 --userpass=WORKER.1:x --url=stratum+tcp://MYCHOSENPOOL.com:3333 --userpass=WORKER.1:x --failover-only --shaders=2048 --intensity=10 --thread-concurrency 8192 --auto-fan --vectors=1 --worksize=256 --lookup-gap=2 --temp-target=75 -I 13 -g 2 -w 256 --gpu-powertune 20 --gpu-memclock=1500
I don't want to go any hotter than 75, so I guess this is the best to expect, and I'm happy so far.
Of course I also installed fglrx for Ubuntu and did the initialization, reboots, etc.-- It does include the amdcccle app now.
Wanted to share the results:
export GPU_MAX_ALLOC_PERCENT=100
./cgminer --scrypt --url=stratum+tcp://MYCHOSENPOOL.com:3333 --userpass=WORKER.1:x --url=stratum+tcp://MYCHOSENPOOL.com:3333 --userpass=WORKER.1:x --failover-only --shaders=2048 --intensity=10 --thread-concurrency 8192 --auto-fan --vectors=1 --worksize=256 --lookup-gap=2 --temp-target=75 -I 13 -g 2 -w 256 --gpu-powertune 20 --gpu-memclock=1500
I don't want to go any hotter than 75, so I guess this is the best to expect, and I'm happy so far.
December 20, 2013, 12:42:51 AM
5970 won't show the specific GPU stats in CG miner 3.7.2 on my win8 machine, just a black screen. Only can see summary stats if I do a '-T' command in the batch file, but this is a pain as I can't optimize very easily. Anyone else run into this issue and how were you able to resolve?
Thanks
Thanks
December 16, 2013, 04:30:56 PM
just want to share my settings:
VTX3D HD7950 Boost X-Edition
Edit: it is a VX7950 3GBD5-2DHXV3
570Kh/s
GPU 170-180Watts
Best settings: 1100/1250/1.088V 570kH/s 3.00 kH/s/W
Not tried to go further down with Voltage, since flashing the bios only works with voltage 1.088V or 1.250V and I had hard time trying to get flashing going to 100%.
Bios tuned like described here.
VTX3D HD7950 Boost X-Edition
Edit: it is a VX7950 3GBD5-2DHXV3
570Kh/s
GPU 170-180Watts
Best settings: 1100/1250/1.088V 570kH/s 3.00 kH/s/W
Not tried to go further down with Voltage, since flashing the bios only works with voltage 1.088V or 1.250V and I had hard time trying to get flashing going to 100%.
Code:
"intensity" : "19",
"vectors" : "1",
"worksize" : "256",
"kernel" : "scrypt",
"lookup-gap" : "0",
"failover-only" : true,
"shaders" : "0",
"thread-concurrency" : "22272",
"gpu-engine" : "880-1100",
"gpu-fan" : "20-85",
"gpu-memclock" : "1250",
"gpu-memdiff" : "0",
"gpu-powertune" : "0",
"gpu-vddc" : "1.088",
"temp-cutoff" : "95",
"temp-overheat" : "90",
"temp-target" : "66",
"auto-fan" : true,
"auto-gpu" : true,
"api-mcast-port" : "4028",
"api-port" : "4028",
"expiry" : "120",
"gpu-dyninterval" : "7",
"gpu-platform" : "0",
"gpu-threads" : "2",
"hotplug" : "5",
"log" : "5",
"no-pool-disable" : true,
"queue" : "1",
"scan-time" : "60",
"scrypt" : true,
"temp-hysteresis" : "3",
"share" : "0",
"vectors" : "1",
"worksize" : "256",
"kernel" : "scrypt",
"lookup-gap" : "0",
"failover-only" : true,
"shaders" : "0",
"thread-concurrency" : "22272",
"gpu-engine" : "880-1100",
"gpu-fan" : "20-85",
"gpu-memclock" : "1250",
"gpu-memdiff" : "0",
"gpu-powertune" : "0",
"gpu-vddc" : "1.088",
"temp-cutoff" : "95",
"temp-overheat" : "90",
"temp-target" : "66",
"auto-fan" : true,
"auto-gpu" : true,
"api-mcast-port" : "4028",
"api-port" : "4028",
"expiry" : "120",
"gpu-dyninterval" : "7",
"gpu-platform" : "0",
"gpu-threads" : "2",
"hotplug" : "5",
"log" : "5",
"no-pool-disable" : true,
"queue" : "1",
"scan-time" : "60",
"scrypt" : true,
"temp-hysteresis" : "3",
"share" : "0",
Bios tuned like described here.
December 12, 2013, 12:37:35 PM
Hello
I need some help to get more speed out of my 7950.
my setup is:
xfx 7950 ghost (dual fan) -+ 630khps gpu 0
xfx 290 (locked) -+870khps gpu 1
config:
sems like that my 290 is also voltage locked, so i will try another bios later to undervolt.
temps
7950: stable 64
290: stable 78
the cards are inside a case, so will lower the fan-speed when my risers arrive. Any idea's? if needed i can open the window in the room where the miner is, to lower the temps even more (I live in COLD Denmark)
I need some help to get more speed out of my 7950.
my setup is:
xfx 7950 ghost (dual fan) -+ 630khps gpu 0
xfx 290 (locked) -+870khps gpu 1
config:
Code:
"api-allow" : "W:127.0.0.1",
"api-listen" : true,
"api-port" : "4028",
"expiry" : "120",
"hotplug" : "5",
"kernel-path" : "/usr/local/bin",
"log" : "5",
"queue" : "1",
"scan-time" : "60",
"scrypt" : true,
"shares" : "0",
"auto-fan" : true,
"gpu-threads" : "1",
"gpu-dyninterval" : "7",
"gpu-engine" : "1100,1000",
"gpu-fan" : "65-88,65-90",
"gpu-platform" : "0",
"gpu-memclock" : "1500,1500",
"gpu-memdiff" : "0,0",
"gpu-powertune" : "-20,11",
"gpu-vddc" : "1.030,0.990",
"intensity" : "20,20",
"temp-target" : "80,80",
"temp-overheat" : "85,85",
"temp-cutoff" : "92,92",
"temp-hysteresis" : "3",
"vectors" : "1,1",
"worksize" : "256,512",
"lookup-gap" : "2,2",
"shaders" : "0,0",
"thread-concurrency" : "21712,32765",
"no-pool-disable" : true
}
"api-listen" : true,
"api-port" : "4028",
"expiry" : "120",
"hotplug" : "5",
"kernel-path" : "/usr/local/bin",
"log" : "5",
"queue" : "1",
"scan-time" : "60",
"scrypt" : true,
"shares" : "0",
"auto-fan" : true,
"gpu-threads" : "1",
"gpu-dyninterval" : "7",
"gpu-engine" : "1100,1000",
"gpu-fan" : "65-88,65-90",
"gpu-platform" : "0",
"gpu-memclock" : "1500,1500",
"gpu-memdiff" : "0,0",
"gpu-powertune" : "-20,11",
"gpu-vddc" : "1.030,0.990",
"intensity" : "20,20",
"temp-target" : "80,80",
"temp-overheat" : "85,85",
"temp-cutoff" : "92,92",
"temp-hysteresis" : "3",
"vectors" : "1,1",
"worksize" : "256,512",
"lookup-gap" : "2,2",
"shaders" : "0,0",
"thread-concurrency" : "21712,32765",
"no-pool-disable" : true
}
sems like that my 290 is also voltage locked, so i will try another bios later to undervolt.
temps
7950: stable 64
290: stable 78
the cards are inside a case, so will lower the fan-speed when my risers arrive. Any idea's?
if needed i can open the window in the room where the miner is, to lower the temps even more (I live in COLD Denmark) 
December 10, 2013, 03:06:04 AM
Hello, i need a help please 
I have a 5970 + 7970
i need settings for cgminer on 5970
7970 = 680kh/s ...I think that is good
5970 = 350kh/s... What settings do you recommend please ?
thx
I have a 5970 + 7970
i need settings for cgminer on 5970
7970 = 680kh/s ...I think that is good
5970 = 350kh/s... What settings do you recommend please ?
thx
December 07, 2013, 02:57:24 PM
@Roger100
port number should be 3334, not 3335.
port number should be 3334, not 3335.
Jump to: