My cards I test was AIO - reference liquid cooling with hbm2 temp no more than 60 degrees. Memory can work up to 1190mhz without rude crash but hashrate is fake both in xmr and eth
Hash rate is fake where exactly, in miner, pool or both?--ypsi
Topic: Any word on amd vega hash rates? - page 18. (Read 202743 times)
--ypsi
heya is there anyone here have stock powerplay costum like this
Name=Vega56
GPU_P0=852;905;0
GPU_P1=991;905;0
GPU_P2=1084;905;0
GPU_P3=1138;905;0
GPU_P4=1150;905;0
GPU_P5=1202;905;0
GPU_P6=1212;905;0
GPU_P7=1407;950
Mem_P0=167;905;0
Mem_P1=500;905;0
Mem_P2=800;905;0
Mem_P3=950;905
Fan_Min=3000
Fan_Max=4900
Fan_Target=75
Fan_Acoustic=2400
Power_Temp=85
Power_Target=0
MV 905 both mem PO,P01, P02 Thankyou
Name=Vega56
GPU_P0=852;905;0
GPU_P1=991;905;0
GPU_P2=1084;905;0
GPU_P3=1138;905;0
GPU_P4=1150;905;0
GPU_P5=1202;905;0
GPU_P6=1212;905;0
GPU_P7=1407;950
Mem_P0=167;905;0
Mem_P1=500;905;0
Mem_P2=800;905;0
Mem_P3=950;905
Fan_Min=3000
Fan_Max=4900
Fan_Target=75
Fan_Acoustic=2400
Power_Temp=85
Power_Target=0
MV 905 both mem PO,P01, P02 Thankyou
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
Wouldn't put much stock in these numbers... these are better than I get with Vega FE's and all my 56/64s.
The newest driver baby... I have my memory clocked at 1185
You should carefully examine your results on the pool. My experience with reference samsung-based vegas is that they have a lot compute errors after 1150mhz hbm clock producing some kind of fake hashrate.
There was another user mentioning that the errors come from leakage from the silicon in high temperature as he could take the cards a lot higher in clock with lower temps using a liquid card . Another conclusion is that lower voltage is beneficial because it lowers temperature
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
Wouldn't put much stock in these numbers... these are better than I get with Vega FE's and all my 56/64s.
The newest driver baby... I have my memory clocked at 1185
For Vega 56, 64, RX 580/570:
It's a whole new ballgame with Cast 0.9.2 + latest Adrenaline 18.x.x drivers + cast-xmr-ui graphical user monitoring interface (available via Github, see Cast thread for the link).
- No need for buggy August BlockChain drivers.
- Cast 0.9.2 has a new --ratewatchdog function to monitor hash rate drop offs/ followed by a restart (no need to modify prior scripts from JJ-CircusDad)
- Cast-xmr-ui turns on "compute" mode without slogging through AMD settings one card at a time
Try this out. I think you will be impressed.
I have been mining for a few days with the new Andrenalin drivers and Cast 9.02 and it works like a charm. It also increased the hash rate compared to the previous version combined with old block chain drivers.
Although I still use the "power play" for the registry I get power consumption about 160w per card at the wall, about 1970Hs for the 56 (Bios 64) and 2040Hs for the 64. All reference cards with Samsung memory.
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
Wouldn't put much stock in these numbers... these are better than I get with Vega FE's and all my 56/64s.
The newest driver baby... I have my memory clocked at 1185
You should carefully examine your results on the pool. My experience with reference samsung-based vegas is that they have a lot compute errors after 1150mhz hbm clock producing some kind of fake hashrate.
I found these two videos on you tube. both are fails
https://www.youtube.com/watch?v=Z48TzK529Eo
https://www.youtube.com/watch?v=SvLhEzSyouY
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
Wouldn't put much stock in these numbers... these are better than I get with Vega FE's and all my 56/64s.
The newest driver baby... I have my memory clocked at 1185
I can confirm. Newest drivers work like a charm on Cast.
Plug and play.
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
Wouldn't put much stock in these numbers... these are better than I get with Vega FE's and all my 56/64s.
The newest driver baby... I have my memory clocked at 1185
You should carefully examine your results on the pool. My experience with reference samsung-based vegas is that they have a lot compute errors after 1150mhz hbm clock producing some kind of fake hashrate.
There was another user mentioning that the errors come from leakage from the silicon in high temperature as he could take the cards a lot higher in clock with lower temps using a liquid card . Another conclusion is that lower voltage is beneficial because it lowers temperature
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
Wouldn't put much stock in these numbers... these are better than I get with Vega FE's and all my 56/64s.
The newest driver baby... I have my memory clocked at 1185
What's your voltage and did you unlock SOC?
without higher SoC over 1185 it would just crash.
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
Wouldn't put much stock in these numbers... these are better than I get with Vega FE's and all my 56/64s.
The newest driver baby... I have my memory clocked at 1185
You should carefully examine your results on the pool. My experience with reference samsung-based vegas is that they have a lot compute errors after 1150mhz hbm clock producing some kind of fake hashrate.
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
Wouldn't put much stock in these numbers... these are better than I get with Vega FE's and all my 56/64s.
The newest driver baby... I have my memory clocked at 1185
What's your voltage and did you unlock SOC?
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
Wouldn't put much stock in these numbers... these are better than I get with Vega FE's and all my 56/64s.
The newest driver baby... I have my memory clocked at 1185
For Vega 56, 64, RX 580/570:
It's a whole new ballgame with Cast 0.9.2 + latest Adrenaline 18.x.x drivers + cast-xmr-ui graphical user monitoring interface (available via Github, see Cast thread for the link).
- No need for buggy August BlockChain drivers.
- Cast 0.9.2 has a new --ratewatchdog function to monitor hash rate drop offs/ followed by a restart (no need to modify prior scripts from JJ-CircusDad)
- Cast-xmr-ui turns on "compute" mode without slogging through AMD settings one card at a time
Try this out. I think you will be impressed.
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
Wouldn't put much stock in these numbers... these are better than I get with Vega FE's and all my 56/64s.
The newest driver baby... I have my memory clocked at 1185
In my country vega 64 price start from $750-900usd, and 1080ti start from 850-1000usd, this year not for gamer, maybe 2018-2019, wheb all vga price will crash hard
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
Wouldn't put much stock in these numbers... these are better than I get with Vega FE's and all my 56/64s.
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
what are the clocks and volts?
I get 11200 Hs using a 6 VEGAs 56 rig. More details on the HW/SW installation, etc here:
http://criptobakery.com/?page_id=101
http://criptobakery.com/?page_id=101
I mine with Vega 64 with 2130 hashes on 170W from the wall on Cryptonight.
how do i get vega frontier edition to work properly?
I've tried 1703, 1709, using blockchain drivers, 17.2, 17.1 various registry files but for some strange reason,
I keep getting code 43 when i manually disable enable vega FE in dev mgr or it enables and hashes at 1500h/s
I've tried 1703, 1709, using blockchain drivers, 17.2, 17.1 various registry files but for some strange reason,
I keep getting code 43 when i manually disable enable vega FE in dev mgr or it enables and hashes at 1500h/s
clean install of blockchain drivers on windows fall edition, set 2 threads at 2016
disable and re-enable Vega FE in windows device drivers, set overdriven tool settings to normal params (mem to 1100 to 1107) including fans and mine away
works for me?
Are you using risers?
Not at the moment but I have in the past, 2 FE in a threadripper moboard
(I presume you have seen the geekmark video and howto https://youtu.be/B9_5z6GbwV0 )
how do i get vega frontier edition to work properly?
I've tried 1703, 1709, using blockchain drivers, 17.2, 17.1 various registry files but for some strange reason,
I keep getting code 43 when i manually disable enable vega FE in dev mgr or it enables and hashes at 1500h/s
Which miner do you use? My vegas hashing 2khs on xmr-stak 2.2.0 but only 1.5khs on 2.3.0 version. Any advices are appreciated.I've tried 1703, 1709, using blockchain drivers, 17.2, 17.1 various registry files but for some strange reason,
I keep getting code 43 when i manually disable enable vega FE in dev mgr or it enables and hashes at 1500h/s
P.s. did you saw that video? https://youtu.be/wUrt7DgSiDM
xmr-stak v2,2.2 and v2.3 1960 Mh/s (vega 56 with 64 Bios @ 1408/1100)
Jump to: