Author

Topic: [ANN] profit switching auto-exchanging pool - www.middlecoin.com - page 199. (Read 829922 times)

member
Activity: 95
Merit: 10
It would be the driver set found here then?

http://www.techpowerup.com/downloads/2196/amd-catalyst-13-1-software-suite-vista-7-64-bit/mirrors

I'll be running 6x R9 270.  I have 20 rigs, and if I could get 6 cards on them, it would increase my utilization.  I have some R9 280X and R9 270X as well though.

The file I used when installing on my 6 GPU rigs is 13-1_vista_win7_win8_64_dd_ccc_whql.exe (149,984 KB).
member
Activity: 84
Merit: 10
 I have been unable to get 6 cards working on a system despite trying BAMT/Win8/Win7/SMOS for like 2 days.  I had to settle with 5 cards per rig, which ultimately made me have more rigs because of not being able to have 6 cards in one.

Not directed at me, but I'll respond anyway.

The biggest surprise I had to get 6 gpus running in one rig, was that I had to use a presence detect short on slots that were described as not needing them (mb z77a g65). Oh and if you do use presence short, dont f**k up your pcie slots by forcing some extra wire into them. Simply cut & connect the appropriate wires in the riser, much simpler and less invasive. Connect B17 to ground (B16 or B18 will do just fine, no need to go to A1).

Use a win98 boot usb stick with "atiflash -i" to see if they are detected by 'the hardware'. If they are, but do not show up in windows, or report error 43, it's a driver thing. If they do not show up for "atiflash" you'll probably need a presence short. No amount of driver juggling will help you in that case.

Which OS if you don't mind me asking? I have 10 rigs with that motherboard running 6 cards on each under BAMT, and no shorts were needed.


Win 8, but that shouldn't matter as they weren't even detected by atiflash without the shorts.

Mind you, I am using 1x-1x risers for all 6 cards so perhaps that is different from your setup.
member
Activity: 110
Merit: 10
It would be the driver set found here then?

http://www.techpowerup.com/downloads/2196/amd-catalyst-13-1-software-suite-vista-7-64-bit/mirrors

I'll be running 6x R9 270.  I have 20 rigs, and if I could get 6 cards on them, it would increase my utilization.  I have some R9 280X and R9 270X as well though.  My 270X rigs have the MSI Z77A-GD65.  I'll have to look into the short via the riser.  I use 16x -> 16x risers for the 16x slots, and 1x -> 16x for the 1x slots, although I have 3 of the Pro BTC boards which will be using 1x -> 1x.  Thanks for the pointers!
sr. member
Activity: 504
Merit: 254
 I have been unable to get 6 cards working on a system despite trying BAMT/Win8/Win7/SMOS for like 2 days.  I had to settle with 5 cards per rig, which ultimately made me have more rigs because of not being able to have 6 cards in one.

Not directed at me, but I'll respond anyway.

The biggest surprise I had to get 6 gpus running in one rig, was that I had to use a presence detect short on slots that were described as not needing them (mb z77a g65). Oh and if you do use presence short, dont f**k up your pcie slots by forcing some extra wire into them. Simply cut & connect the appropriate wires in the riser, much simpler and less invasive. Connect B17 to ground (B16 or B18 will do just fine, no need to go to A1).

Use a win98 boot usb stick with "atiflash -i" to see if they are detected by 'the hardware'. If they are, but do not show up in windows, or report error 43, it's a driver thing. If they do not show up for "atiflash" you'll probably need a presence short. No amount of driver juggling will help you in that case.

Which OS if you don't mind me asking? I have 10 rigs with that motherboard running 6 cards on each under BAMT, and no shorts were needed.
member
Activity: 84
Merit: 10
 I have been unable to get 6 cards working on a system despite trying BAMT/Win8/Win7/SMOS for like 2 days.  I had to settle with 5 cards per rig, which ultimately made me have more rigs because of not being able to have 6 cards in one.

Not directed at me, but I'll respond anyway.

The biggest surprise I had to get 6 gpus running in one rig, was that I had to use a presence detect short on slots that were described as not needing them (mb z77a g65). Oh and if you do use presence short, dont f**k up your pcie slots by forcing some extra wire into them. Simply cut & connect the appropriate wires in the riser, much simpler and less invasive. Connect B17 to ground (B16 or B18 will do just fine, no need to go to A1).

Use a win98 boot usb stick with "atiflash -i" to see if they are detected by 'the hardware'. If they are, but do not show up in windows, or report error 43, it's a driver thing. If they do not show up for "atiflash" you'll probably need a presence short. No amount of driver juggling will help you in that case.
member
Activity: 95
Merit: 10
RickJamesBTC,

   What OS are you using?  I have been unable to get 6 cards working on a system despite trying BAMT/Win8/Win7/SMOS for like 2 days.  I had to settle with 5 cards per rig, which ultimately made me have more rigs because of not being able to have 6 cards in one.

I'm running four 6 GPU rigs. Here is a guide:

https://bitcointalksearch.org/topic/m.5111084

Let us know how you get on.

Just editing to add the details of my 6-GPU rigs rigs..

 - Windows 7 Pro (completely up to date)
 - AMD Catalyst 13.1 from TechPowerUp
 - Gigabyte GA-990FXA-UD3
 - 6x R9 280X on ribbon risers (4 powered)
member
Activity: 84
Merit: 10
Balanceboost, going to be a good payout day, nice
member
Activity: 110
Merit: 10
RickJamesBTC,

   What OS are you using?  I have been unable to get 6 cards working on a system despite trying BAMT/Win8/Win7/SMOS for like 2 days.  I had to settle with 5 cards per rig, which ultimately made me have more rigs because of not being able to have 6 cards in one.
full member
Activity: 160
Merit: 100
Noticed the coin diff this morning is 135.  That isn't doge.  Wonder what we are mining and if we are giving doge a break.
hero member
Activity: 798
Merit: 1000
Do you need two different btc address for two miners?

Nope, you don't have to.   Most just use one address for multiple rigs.
sr. member
Activity: 350
Merit: 250
Do you need two different btc address for two miners?
hero member
Activity: 798
Merit: 1000
On topic: Really large amounts of unexchanged building up now  Shocked

It's all Doge piling up with the Jamaican Bobsled Team donations and transactions Mon....

Feel the rhythm! Feel the rhyme! Get on up, its bobsled time! COOL RUNNINGS!

member
Activity: 73
Merit: 10
IT Consultant
Im having 760 kh/s at 7970 (< 0.3% rejects) (which is same as 280x) using config which can be found on this page http://cgminer-config.com/. I've made that page a while ago so people could share theirs configs and vote for best ones ;P

That config is crazy imo. GPU fan at 95% is a big no-no for me. The fan on my hottest card run at 55% for 82 C.

On topic: Really large amounts of unexchanged building up now  Shocked
newbie
Activity: 21
Merit: 0
using asia server atm
try to go for high thread-concurrency + low intensity
getting 1~% reject

280x @ 680kh/s:-
thread-concurrency : 28672 (2048 * 14)
gpu-thread : 1
intensity : 18
gpu-engine : 1100
gpu-memclock : 1500

7950 @ 590kh/s :-
thread-concurrency : 25984 (1792 * 14.5)
gpu-thread : 1
intensity : 18
gpu-engine : 1050
gpu-memclock : 1250


Woah you sure can run more optimized than that.

We run our 280x @ 725kh/s with 0.5% reject rate
thread-cuncurrency: 8196 (with Zuikkis kernel)
gpu-thread: 2
intensity: 13
gpu-engine: 1020
gpu-memclock: 1500


Im having 760 kh/s at 7970 (< 0.3% rejects) (which is same as 280x) using config which can be found on this page http://cgminer-config.com/. I've made that page a while ago so people could share theirs configs and vote for best ones ;P

using smos-linux (modded BAMT) on my rig
does not play nice with 280x
but 7950 works like charm

ill try for ubuntu/debian os this weekend

edit: and thanks for the links, will give them a try
hero member
Activity: 798
Merit: 1000
I wonder what I am doing wrong.  Today was a decent day, but just over .01 per Mh/s for me.  A far cry to the .015 that many are reporting.  I have been connected to the pool since it's inception and am always between 22 and 24 Mh/s.

http://www.middlecoin.com/reports/18Y3Dhw1EdFTfzpVe4bKKhPgsQYuFWERLT.html

I have 13 rigs with a total of 42 cards active - all either 7950's or R9 280x's.  Hottest card is 83, most in the mid to high 70's.  CGRemote shows a total of 12 hardware errors and many of these rigs have been weeks without a reboot.  The 7950's hash at about 630-670 while the R9 280's are between 690-735.  While I haven't done the math on all 42 cards, the ones I glanced at are at least 85% WU vs. Hash Rate.  Most of the 7950's are undervolted a bit and so are a few of the R9 280's.

Can anyone think of anything that I might be missing?





Are you connected to Middlecoin.com or one of the beta servers? Try connecting to the USEast.middlecoin.com server if you're in the USA, or eu or Asia server if not.

Try to get those Reject rates down too.  Tweak your intensity and thread concurrency.
newbie
Activity: 15
Merit: 0
using asia server atm
try to go for high thread-concurrency + low intensity
getting 1~% reject

280x @ 680kh/s:-
thread-concurrency : 28672 (2048 * 14)
gpu-thread : 1
intensity : 18
gpu-engine : 1100
gpu-memclock : 1500

7950 @ 590kh/s :-
thread-concurrency : 25984 (1792 * 14.5)
gpu-thread : 1
intensity : 18
gpu-engine : 1050
gpu-memclock : 1250


Woah you sure can run more optimized than that.

We run our 280x @ 725kh/s with 0.5% reject rate
thread-cuncurrency: 8196 (with Zuikkis kernel)
gpu-thread: 2
intensity: 13
gpu-engine: 1020
gpu-memclock: 1500


Im having 760 kh/s at 7970 (< 0.3% rejects) (which is same as 280x) using config which can be found on this page http://cgminer-config.com/. I've made that page a while ago so people could share theirs configs and vote for best ones ;P
copper member
Activity: 199
Merit: 14
Ok well ill have a play
My cards are quite new though, not much info to google around with
x1 gigabyte windforce edition r9290
x1 sapphire tri-x r9290

both quiet and cool
the sapphire hashes about 50k more than the gigabyte using the below

--gpu-fan 100 --temp-target=80 --temp-overheat=85 --intensity 19 --vectors 1 -g 1 -w 512 --thread-concurrency 32756 --lookup-gap 2 --gpu-memclock 1500 --gpu-engine 1000 --gpu-powertune 20
 --api-port 4028 --expiry 120 --failover-only true --log 5 --scan-time 60 --temp-hysteresis 3

Ill see what happens when i turn down the intensity
member
Activity: 73
Merit: 10
IT Consultant
using asia server atm
try to go for high thread-concurrency + low intensity
getting 1~% reject

280x @ 680kh/s:-
thread-concurrency : 28672 (2048 * 14)
gpu-thread : 1
intensity : 18
gpu-engine : 1100
gpu-memclock : 1500

7950 @ 590kh/s :-
thread-concurrency : 25984 (1792 * 14.5)
gpu-thread : 1
intensity : 18
gpu-engine : 1050
gpu-memclock : 1250


Woah you sure can run more optimized than that.

We run our 280x @ 725kh/s with 0.5% reject rate
thread-cuncurrency: 8196 (with Zuikkis kernel)
gpu-thread: 2
intensity: 13
gpu-engine: 1020
gpu-memclock: 1500
newbie
Activity: 21
Merit: 0
Must be my config then.
I've got a great hash rate around 880 each card with 820-840 WU but my rejects are huge.


using asia server atm
try to go for high thread-concurrency + low intensity
getting 1~% reject

280x @ 680kh/s:-
thread-concurrency : 28672 (2048 * 14)
gpu-thread : 1
intensity : 18
gpu-engine : 1100
gpu-memclock : 1500

7950 @ 590kh/s :-
thread-concurrency : 25984 (1792 * 14.5)
gpu-thread : 1
intensity : 18
gpu-engine : 1050
gpu-memclock : 1250
legendary
Activity: 2716
Merit: 1094
Black Belt Developer
Must be my config then.
I've got a great hash rate around 880 each card with 820-840 WU but my rejects are huge.


try lowering intensity: a bit less hashrate but better WU, which is what counts.
Jump to: