New automatic update 1330 eith new drivers and new kernel make all my rigs containing rx 480/580 not working. It starts mining but after 1-2 minutes it shows thah there is problem wit riser ot too muvh undervolt ant it restarts...
Even clean rigs wigh only 480 or 580 not mixed with other card do not work. Please help.

I have several rigs using simplemining. One of the rigs have 6 VEGA gpu. I would like to ask is there a way to to set different memory timings for each vega cards using amdmemorytweak in OC section? I tryed -d function but it not working may be.

I found myself the answer

Gpu: 01234 - RX580, RX580, VEGA64, VEGA64, RX580

amdmemorytweak --i 0 --REF 30
amdmemorytweak --i 1 --REF 30
amdmemorytweak --i 2 --CL 20 --RAS 30 --RCDRD 13 --RCDWR 12 --RC 44 --RP 14 --RRDS 3 --RRDL 6 --RTP 5 --FAW 12 --CWL 8 --WTRS 4 --WTRL 9 --WR 14 --REF 65000--RFC 249
amdmemorytweak --i 3 --CL 20 --RAS 30 --RCDRD 13 --RCDWR 12 --RC 44 --RP 14 --RRDS 3 --RRDL 6 --RTP 5 --FAW 12 --CWL 8 --WTRS 4 --WTRL 9 --WR 14 --REF 65000--RFC 249
amdmemorytweak --i 4 --REF 30

Rx580 31,5 mh, vega64 51,7 - stable

RX 580 Sapphire Pulse 4gb Elpida
Timing: 777000000000000022AA1C00AC615B3CA0550F142C8C1506006004007C041420CA8980A9020004C 01712262B612B3715
Memory: 2100Mhz - STABLE, 2125 and 2150 also OK but power consumption is rising and MHS change is not so much
Core: 1145Mhz
Voltage: 900Mv
Hash rate: 31.50 to 31.55

GIGABYTE 990FXA-UD3 Rev.4 and rev.5 have 6 PCIE slots and work perfect!!! Used by me with 4 X 7990 and two 970 GTX! All 10 GPU work perfect.

New miner 8.0, bios mod and R9 290x tri-x - 315 mhs Thats my record. I have several 290x tri-x but only one of them gives me this result. Other between 295 and 305. Core 1100 / Mem. 1375 / voltage + 6 / temp. 82-84C. 320mhs is possible wih core 1140 but needs more air

There is no reason to discuss further. You have to add AVERAGE HASHRATE GRAPH to BETARIGS!!!

What about 1 hour extra time i gave to that client? This conversation in not about money as you understand! First 4 hours the rig is really working 250 to 300 mhs, but last 2 hours the speed was full 350! If you have no graph how can we calculate real situation?

This is messege from that renter -

"Hi Alexbg21!

Filo1992, the renter of your rig #7468 ▂▅▇█▓▒░░░ARig-24/7-335Mhs+/Stable/Monitored░░░░▒▓█▇▅▂, has sent you a message!

thanks for more 25%

You can reply to this message by going to the rental message page at http://www.betarigs.com/rig/7468/job/205128/messages and also, if the rental is still active, at the rig page at http://www.betarigs.com/rig/7468
Cheers!
The FinalHash Team @ BetaRigs "

What i have to say now. WELCOME AGAIN?!!!

After 25% penalty without any normal graph etc... Will stop leasing my rigs in BETARIGS. ▂▅▇█▓▒░░░ARig-24/7-350Mhs+/Stable/Monitored░░░▒▓█▇▅▂ rig have almost 100% rental performance index in mining rig rentals!!! You will loose your clients if you work that way!

Palit JETSTREAM GTX 970 - Latest release 12 by SP_ OC + 155Mhz on Core and + 50 on Memory - 8300-8400mhs! Will donate for sure! Use A MINER and kernels for AMD and have 290 Tri-x with 8000mhs stable with 1000mhz core! 1025 Mhz give me 8200 but driver crash occure some times! Also there is something strange incorporated in this miner i feel Also some 30-40 watt + on each VGA and also temerature + 5-10 degree ...

Palit GTX 970 - X11 - Latest 9 release by SP_ : 7900 Mhs

Palit JETSTREAM GTX 970 X11 slightly OCed - 7600 - 7700 Mhs

Latest ccMiner release 1.4.6-tpruvot (Oct 26th 2014) Overclocked Palit GTX 970 - X11 7400 Mhs

Ofcourse i dont need, but may be somebody else will really need so this info here is usefull in any case!

It was already tested. Problem is not 6 GPU but more than 8 GPU. 14.4/14.6/14.7/14.9 - tested. Problem is in amd kernel. amdkmdag.sys - ATI Radeon Kernel Mode Driver and amdkmdap.sys - AMD multi-vendor Miniport Driver causes this problem.

Here is the answer Not so long to read! Latest posts are about mythical 8 gpu limit on windows!

https://bitcointalk.org/index.php?topic=712228.msg9132557#msg9132557

PALIT GTX 970 JETSTREAM non-clocked gives me 5600-5700 Khs on X11 wiht ccminer. Overclocking gave makes 6100-6150 Khs. Is this normal results or it can be clocked more?
---------------------------------------------------------
What about cudaminer? Which one is better to use?

https://imageshack.com/i/exWBHEs3j

https://imageshack.com/i/kqcGXV1Hj

PALIT GTX 970 JETSTREAM non-clocked gives me 5600-5700 Khs on X11 wiht ccminer. Overclocking gave makes 6100-6150 Khs. Is this normal results or it can be clocked more?

Still no success with more than 8 AMD GPU on windows. I think kernel file makes problem, but success with 4 X 7990 and two GTX 970 plus CPU mining - all 45 Mhs on X11 !!! 8 GPU limit per windows is MYTH. This is absolutely not true!!! These days i will buy pcie expander card and will try to add three more GTX 970 In windows 8.1 i did not connect any monitor like it was it windows 7. Everithing works perfect.

8+GPU works only on GIGABYTE 990FXA-UD3 Rev. 4.0 (Revision 1.2/2.0/3.0 and GYGABYTE 990FXA-UD5 which i tested have bios issue for 8+GPU). Asrock H81PRO BTC - only 7 GPU!!! I think this board have bios bug or something like that. Till now GIGABYTE 990FXA-UD3 Rev. 4.0 is best for support multiple GPU but do not try to switch more than 10 GPU - South bridge will die !!! 2 of my board are death now, trying to switch on 11th GPU

https://imageshack.com/i/exWBHEs3j

https://imageshack.com/i/kqcGXV1Hj

-------------------------------------------
windows 8 gpu limit
8 gpu limit
Maximum Gpus windows 8
Windows GPU limit
-------------------------------------------

As i understood problem is in these two files:

amdkmdag.sys   c:\windows\system32\drivers\atikmdag.sys
amdkmdap.sys   c:\windows\system32\drivers\atikmdap.sys