Anyone here with knowledge of CPU's and PCIe channel utilization? I ask because @Blackhawk67 over at the z.cash forum is claiming that he is getting an extra 70-100 sol/s per gpu out of his 1080ti's when mining with both cpu and gpu. Based on his post it looks like he is getting about 830 sol/s per gpu, but when just mining with gpu's he's back down to normal range. I suspect it has something to do with cpu and pcie channel utilization, but my knowledge only goes so far..
https://i.imgur.com/D3jtIyV.png
https://forum.z.cash/t/nvidia-gtx-1080-ti-gpus/16765/98
Topic: EWBF's CUDA Zcash miner - page 104. (Read 2164327 times)
i build my first nvidia ring but doesn't loook great till now
5x 1060 3G tested nicehash for benchmark i only get 200 on zec mining
and lol ether mining does 2 mh ? wtf is that 
i tested multipoolminer but could not done the benchmark driver always crash
do we need to bios mod ?
5x 1060 3G tested nicehash for benchmark i only get 200 on zec mining
and lol ether mining does 2 mh ? wtf is that i tested multipoolminer but could not done the benchmark driver always crash
do we need to bios mod ?
A Good 750 watt PS is plenty for a pair of GTX 1080 ti - most of them are only 250 watt TDP, though I suspect a few of the "high end" ones might be 300 or a bit more.
Oddly enough, my Aorus has 2 x 8 pin connectors but only has a stock TDP of 250 watts - PLENTY of headroom there.
Rig also has one of the lower-end 3-fan Gigabyte 1080ti cards, and eats 570 peak watts AT THE WALL with a mild overclock and no TDP adjustment on both cards.
I've not had the time to play with it much, though - and it's not running ZEC at the moment anyway, I'm doing something else with it that pushes it harder (more like ETH) for testing purposes.
Oddly enough, my Aorus has 2 x 8 pin connectors but only has a stock TDP of 250 watts - PLENTY of headroom there.
Rig also has one of the lower-end 3-fan Gigabyte 1080ti cards, and eats 570 peak watts AT THE WALL with a mild overclock and no TDP adjustment on both cards.
I've not had the time to play with it much, though - and it's not running ZEC at the moment anyway, I'm doing something else with it that pushes it harder (more like ETH) for testing purposes.
Well I am running a corsair CX750, and adding the --pec option reveals my 1080ti aorus xtreme cards pulling about 290-300 watts each (max stable oc settings), so that only leaves me with 150 watts for the mobo, processor (celeron - 50watts), ssd, etc..
Hi there , I've tried to download your miner at google docs but when I try to download it I got the following error : miner contains a virus and cannot be downloaded . Made an exception and somehow downloaded it but my antivirus (Nod32) also spiked a virus and deleted the files . Is it safe to download and install it ?
A Good 750 watt PS is plenty for a pair of GTX 1080 ti - most of them are only 250 watt TDP, though I suspect a few of the "high end" ones might be 300 or a bit more.
Oddly enough, my Aorus has 2 x 8 pin connectors but only has a stock TDP of 250 watts - PLENTY of headroom there.
Rig also has one of the lower-end 3-fan Gigabyte 1080ti cards, and eats 570 peak watts AT THE WALL with a mild overclock and no TDP adjustment on both cards.
I've not had the time to play with it much, though - and it's not running ZEC at the moment anyway, I'm doing something else with it that pushes it harder (more like ETH) for testing purposes.
Oddly enough, my Aorus has 2 x 8 pin connectors but only has a stock TDP of 250 watts - PLENTY of headroom there.
Rig also has one of the lower-end 3-fan Gigabyte 1080ti cards, and eats 570 peak watts AT THE WALL with a mild overclock and no TDP adjustment on both cards.
I've not had the time to play with it much, though - and it's not running ZEC at the moment anyway, I'm doing something else with it that pushes it harder (more like ETH) for testing purposes.
A stupid question, if I have 6 gpu in the miner I have to put cuda_devices 0 1 2 3 4 5? Or do not need to put it? I ask you why I do not know the cuda function.
Use the cuda_devices parameter only if you do not want to run all the GPUs.
i.e. --cuda_devices 0 2 if you want to use only 1st and 3rd GPU.
Here are my settings for 2 MSI GTX 1080Ti's: I'm getting about 770 Sol/s with both of them, 780 and over when not using the computer.
Core Clock: 1974
Mem Clock: 5702 or 11404
Core Voltage: 950mv
Temperature: 65-67
Fan Speed: 43-48%
Power consumption: around 250w per card according to EWBF's miner
3.1 Sol/watt
It's been stable so far and is running very cool and quiet. :smiley:
Also, i'm using the Voltage/Frequency Curve Editor in MSI Afterburner to manage my overclock and voltage.
Damn, those are some great stats.
I OCd them roughly to your stats (even though I didnt edit curves, can you pls send your curve screenshot?),
and I'm getting around 730-740 MAX, I dont think I can squeeze a lot more out of them.
I guess I'm happy with 740, even though that 780 sounds so sweet.
That sounds already better.
Here is the screenshot:
https://i.imgur.com/dilvdDI.png
Aha, interesting curve, you think it's the most optimal approach?
I got pretty close to your numbers now, without playing with the curve. 745-750 tops...
That is a great result IMHO, I'll leave it like this till tomorrow, and maybe play around with it some more.
I believe with future versions of the miner we could see 800 sols.
I'm currently doing some more testing and it seems like one of my cards can do 800 Sol/s already
it's bouncing between 785 and 805. Core is set to 2012 and memory to 5899 with 993mv so a bit more power.Your rig closely resembles mine (2 1080ti's right?) What power supply are you running? I'm thinking 750w is not enough since one of the cards will dip below 600 sol/s if I push it too hard, requiring a restart.
A stupid question, if I have 6 gpu in the miner I have to put cuda_devices 0 1 2 3 4 5? Or do not need to put it? I ask you why I do not know the cuda function.
Can this miner do Zencash as well?
Any equihash Version 0.3.4b released.
Many new features on the way.
Web page screenshot: https://s4.postimg.org/opjzemkwt/Screenshot_20170611_220048.png
Many new features on the way.
Web page screenshot: https://s4.postimg.org/opjzemkwt/Screenshot_20170611_220048.png
EWBF, in your next release can you please include timestamps in the log files?
Ok, it will be a local date.
EWBF,
Can you also look at making the miner re-check for connection to the primary pool at a time interval if it does divert to a secondary pool? Right now if the primary pool goes down, it will mine on the alternate pool until the miners is restarted, even if the primary pool comes back on line.
Can this miner do Zencash as well?
Might be dummy question..
Does it make diffencence in performance, if for example have two 3x10X0 rigs than those same six cards are on one motherboard.
Does it make diffencence in performance, if for example have two 3x10X0 rigs than those same six cards are on one motherboard.
You have to go with risers to run more than 4 cards on one motherboard.
Performance FOR MINING is not noticeably affected, as mining is a low-bandwidth application.
Such a rig can be a LITTLE more efficient on a per-card basis as you're only using one CPU, one set of RAM, one HD/SSD/USB stick.
Such a rig tends to be noticeably harder to get set up, get working well, get overclock undervolt etc set for each card to optimise it.
Stability seems to be somewhat worse, and risers themselves sometimes cause issues.
Cost on a "per GPU" basis will tend to be a HAIR lower - you pay for the risers and need additional power connectors to power the risers, but you save on MB/RAM/CPU and such SOMETIMES depending on the MB.
Good cooling will tend to be easier to set up, as the cards can be spaced out more.
There are applications where using a 1x PCI-E connection to a GPU is a noticeable to major performance bottleneck, but mining isn't one of those.
(It's a very minor issue on ETH/ETC/offshoots when you first start up a rig, and once every few days when a new DAG epoch is entered, but even THERE it's not a significant issue - you lose some seconds 'cause the DAG takes longer to load but after that it's no issue at all).
@opposumi - thanks - you have just fixed my problem.
I reviewed settings and there is somethig called AiProtection - it's kind of asus FW. Disabled it and every mining is working.
My topology is:
provider <-> asus router (nat) <-> main switch <-> room switch <-> miner
finally problem is fixed
Thank You
I reviewed settings and there is somethig called AiProtection - it's kind of asus FW. Disabled it and every mining is working.
My topology is:
provider <-> asus router (nat) <-> main switch <-> room switch <-> miner
finally problem is fixed
Thank You
No problems, good to hear
@opposumi - thanks - you have just fixed my problem.
I reviewed settings and there is somethig called AiProtection - it's kind of asus FW. Disabled it and every mining is working.
My topology is:
provider <-> asus router (nat) <-> main switch <-> room switch <-> miner
finally problem is fixed
Thank You
I reviewed settings and there is somethig called AiProtection - it's kind of asus FW. Disabled it and every mining is working.
My topology is:
provider <-> asus router (nat) <-> main switch <-> room switch <-> miner
finally problem is fixed
Thank You
today updated firmware to latest one.
p2p not blockig
network services disabled
qos disabled
same like it was
p2p not blockig
network services disabled
qos disabled
same like it was
does log show anything odd?
Your network topology?
operator router --> asus (with nat?) --> miner?
if so, can you connect miner directly to operator router to verify if miner can then connect or not?
today updated firmware to latest one.
p2p not blockig
network services disabled
qos disabled
same like it was
p2p not blockig
network services disabled
qos disabled
same like it was
hard to say what causing it...
That is the one thing why i hate consumer grade network devices
is there any p2p blocking rules or similar enabled, spi and such..
...latest firmware too?
Looked user manual, not much settings to fiddle
4.6.4 Network Services Filter - try to disable if not allready so...
and QOS, that should not make much differenc imo, but try disable (to be sure that asus qos is not so crappy, that it chokes miner traffic to death)
That is the one thing why i hate consumer grade network devices
is there any p2p blocking rules or similar enabled, spi and such..
...latest firmware too?
Looked user manual, not much settings to fiddle
4.6.4 Network Services Filter - try to disable if not allready so...
and QOS, that should not make much differenc imo, but try disable (to be sure that asus qos is not so crappy, that it chokes miner traffic to death
)
This is what I already managed to know, but how to find the clue?
That is a point here
That is a point here
Sounds firewall issue, or router is doing something odd.
Hi guys,
this is the first my first on this site.
I was searching for about one day.
I just started with mining, and at the begining I have an issue that I'm trying to solve
I have win 7 64bit with GTX 1070.
I started with wallet, than I choose Cuda Zcash miner and it up with follwoing setup:
Than I see such an error
I was trying to investigate the problem and I tried ping this server, than to tracert it.
The connection is ok
It was done over my regullar office network.
Than I had tried to disbale network and run USB LTE modem. - It's working fine!!!
Than I restarted on teh network and same problem.
Then I put USB LTE modem to the router and the problem still occurs.
So there is a question how the CUDA zhash miner cooperate with router? My router is ASUS RT-AC68U. There is latest version of merlin software.
There is disabled FW.
What is interesting when I try to connect telnet to the pool server - the connection is established.
So anyway. Could give me some hints how to solve my topic?
Thanks in advance
BR
yacenty
this is the first my first on this site.
I was searching for about one day.
I just started with mining, and at the begining I have an issue that I'm trying to solve
I have win 7 64bit with GTX 1070.
I started with wallet, than I choose Cuda Zcash miner and it up with follwoing setup:
Code:
miner --server eu1-zcash.flypool.org --port 3333 --user t1P1e9PEoWc7yexNCxssAsRZj11L5sagRgJ.rig1 --pass x
Than I see such an error
Code:
+-------------------------------------------------+
| EWBF's Zcash CUDA miner. 0.3.4b |
+-------------------------------------------------+
INFO: Current pool: eu1-zcash.flypool.org:3333
INFO: Selected pools: 1
INFO: Solver: Auto.
INFO: Devices: All.
INFO: Temperature limit: 90
INFO: Api: Disabled
---------------------------------------------------
ERROR: Stratum subscribe timeout
I was trying to investigate the problem and I tried ping this server, than to tracert it.
The connection is ok
It was done over my regullar office network.
Than I had tried to disbale network and run USB LTE modem. - It's working fine!!!
Than I restarted on teh network and same problem.
Then I put USB LTE modem to the router and the problem still occurs.
So there is a question how the CUDA zhash miner cooperate with router? My router is ASUS RT-AC68U. There is latest version of merlin software.
There is disabled FW.
What is interesting when I try to connect telnet to the pool server - the connection is established.
So anyway. Could give me some hints how to solve my topic?
Thanks in advance
BR
yacenty
Jump to: