Author

Topic: PhoenixMiner 6.2c: fastest Ethereum/Ethash miner with lowest devfee (Win/Linux) - page 479. (Read 784950 times)

full member
Activity: 224
Merit: 102

Quote

   Please check the IP addresses of the pool servers you are connecting with after the reconnect. Some of them are really overloaded and the latency may become very large. We will include stats about the pool latency in the future. When you reconnect to the same pool URL, the load balancing servers often connect you to a different server with new IP address that may be less loaded and hence the different latencies.


I am connecting with all my miners to the same eth-eu1.nanopool.org pool. As I wrote, sometimes this latency problem is solved by closing the miner a running it again. With the same .bat file to the same pool. It looks like your miner dos not like my NVidias Smiley
full member
Activity: 357
Merit: 101
not OC just basic run and it breaks after the 5th share here is the error again and again, installed 390 driver

CUDART error in CudaProgram.cu:422 : unspecified launch failure (4)
GPU3 search error: unspecified launch failure
CUDART error in CudaProgram.cu:422 : unspecified launch failure (4)
GPU4 search error: unspecified launch failure
CUDART error in CudaProgram.cu:422 : unspecified launch failure (4)
GPU6 search error: unspecified launch failure
CUDART error in CudaProgram.cu:422 : unspecified launch failure (4)
GPU2 search error: unspecified launch failure
CUDART error in CudaProgram.cu:422 : unspecified launch failure (4)
GPU1 search error: unspecified launch failure
CUDART error in CudaProgram.cu:422 : unspecified launch failure (4)
GPU5 search error: unspecified launch failure
CUDART error in CudaProgram.cu:422 : unspecified launch failure (4)
GPU7 search error: unspecified launch failure
Thread(s) not responding. Restarting.

using command line :

setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
PhoenixMiner.exe -nvidia -pool us1.ethermine.org:4444 -pool2 eu1.ethermine.org:4444 -wal mywallet.rig -proto 3 -eres 0 -lidag 2

still not working, can anyone help on this?
   Unfortunately, we can't reproduce it here. Does it happen with older versions too and what kind of cards you are using? This error happens when the program is trying to run the search kernel after the DAG was already successfully created, so it is even more puzzling.

An issue with 2.6;

A machine that works as expected with 2.5d, with 2.6 after it creates the DAG file and starts mining the main thread is somehow reported as not responding although it is still mining. I've setup the reboot.bat file so the machine goes into reboot loop after that.
....
   Thanks for reporting this issue and for the log - it is obvious that the main thread is not dead but the watchdog decided that it is, so we'll see what could be causing this. The restricted net access is not to blame for this as even if you stop the Internet access entirely no thread should freeze.


I try PhoenixMiner 2.6b with recommended above parameters, on Farm1 PhoenixMiner didn't start and show the same error "clCreateCommandQueue (-6)" but on second Farm PhoenixMiner started up and started working. After restart PhoenixMiner on Farm2 I recive error "clCreateCommandQueue (-6)" again Sad
I try restart PhoenixMiner and MSI Afterburner several times and change parameter GPU_FORCE_64BIT_PTR from 0 to 1 and backward - i have achieved success several times (PhoenixMiner
 started UP sometimes)..

What i noticed, when Phoenix starts up normally Phoenix define 4GB VRAM, when Phoenix can't starts, it define only 3GB VRAM

Log when Phoenix can't start:
17568:15:09:59.153: main No CUDA driver found
17568:15:09:59.477: main Available GPUs for mining:
17568:15:09:59.477: main GPU1: Radeon RX 570 Series (pcie 1), OpenCL 1.2, 3 GB VRAM, 32 CUs
17568:15:09:59.477: main GPU2: Radeon RX 570 Series (pcie 2), OpenCL 1.2, 3 GB VRAM, 32 CUs
17568:15:09:59.477: main GPU3: Radeon RX 570 Series (pcie 4), OpenCL 1.2, 3 GB VRAM, 32 CUs
17568:15:09:59.477: main GPU4: Radeon RX 570 Series (pcie 5), OpenCL 1.2, 3 GB VRAM, 32 CUs
17568:15:09:59.477: main GPU5: Radeon RX 570 Series (pcie 6), OpenCL 1.2, 3 GB VRAM, 32 CUs
17568:15:09:59.491: main ADL library initialized

Log when Phoenix started normally:
17569:20:12:50.903: main No CUDA driver found
17569:20:12:51.179: main Available GPUs for mining:
17569:20:12:51.179: main GPU1: Radeon RX 570 Series (pcie 1), OpenCL 1.2, 4 GB VRAM, 32 CUs
17569:20:12:51.179: main GPU2: Radeon RX 570 Series (pcie 2), OpenCL 1.2, 4 GB VRAM, 32 CUs
17569:20:12:51.179: main GPU3: Radeon RX 570 Series (pcie 4), OpenCL 1.2, 4 GB VRAM, 32 CUs
17569:20:12:51.179: main GPU4: Radeon RX 570 Series (pcie 5), OpenCL 1.2, 4 GB VRAM, 32 CUs
17569:20:12:51.179: main GPU5: Radeon RX 570 Series (pcie 6), OpenCL 1.2, 4 GB VRAM, 32 CUs
17569:20:12:51.192: main ADL library initialized

My rigs have 2GB RAM Memory (900-800 MB free when all works)
Sorry for my english)
   Something is wrong with your drivers. We've never seen RX 570 to report less than 4 GB of memory or only OpenCL 1.2 (it reports OpenCL 2.0). You probably have to use DDU to remove your drivers then install them again (for example the blockchain drivers) and make sure that they are patched if you have BIOS modded your cards.

I really hope the stale shares issue recoding works. I'm getting 2x or higher more state shares than Claymore.
It's quite obvious to see in the charts from ethermine.org.
   While it will have positive effect, most of these stale shares (if you are using 2.6) are mostly random and depend on the current pool server (some are busier than the others and the latency is worse). We haven't seen anything close to 2x than Claymore over a long period of time.

After download, the antivirus show file has virus
anybody help me?
   It's a false positive, you can safely ignore it (but if your PhoenixMiner.exe file suddenly got deleted by your AV, make sure that it is white-listed). If you want to be sure that the files wasn't tampered with, you can use the hashes in the first post of this theme to check if the downloaded files are the same as the ones we've uploaded.

i have not read the all post, i want to know is possible to use voltage, overclock, underclock directly from the bat file like i used to do with claymore?
   Not yet, we are working on this.

I'm testing this miner about a week on few rigs, but have mixed feelings. Ammount of info in console = great. Readability of these info = very bad (I wrote about it few days ago). Reported hashrate about 1-2 Mh/s higher than Claymore. But. This is just a teoretical number and what matters are accepted shares. And with Phoenix I have less accepted shares last days then in Claymore.
Sometimes it has weird behavior - autostarting miner after reboot gives me sometimes very high accept share time, about twice as high as normal. Quitting miner and starting it again solve this sometimes.
On rig started restarting itself, after running few minutes of Phoenix. (NVidia GPUs)
Tried a lot of testing, -mi settings, but I still don't have a feeling, it is the best miner. Time will show... Keep up good work, idea is nice. Thanks
   Please check the IP addresses of the pool servers you are connecting with after the reconnect. Some of them are really overloaded and the latency may become very large. We will include stats about the pool latency in the future. When you reconnect to the same pool URL, the load balancing servers often connect you to a different server with new IP address that may be less loaded and hence the different latencies.
newbie
Activity: 3
Merit: 0
Im trying to use Phoenix with Awesome Miner...anyone help me set it up?
sr. member
Activity: 756
Merit: 250
ok reluslts of phoneix afer 2 hours and i think is way too much better than claymore

on the left mining with claymore and on the right of the graphical itsphoenix( stable and more shares)


full member
Activity: 224
Merit: 102
I'm testing this miner about a week on few rigs, but have mixed feelings. Ammount of info in console = great. Readability of these info = very bad (I wrote about it few days ago). Reported hashrate about 1-2 Mh/s higher than Claymore. But. This is just a teoretical number and what matters are accepted shares. And with Phoenix I have less accepted shares last days then in Claymore.
Sometimes it has weird behavior - autostarting miner after reboot gives me sometimes very high accept share time, about twice as high as normal. Quitting miner and starting it again solve this sometimes.
On rig started restarting itself, after running few minutes of Phoenix. (NVidia GPUs)
Tried a lot of testing, -mi settings, but I still don't have a feeling, it is the best miner. Time will show... Keep up good work, idea is nice. Thanks
sr. member
Activity: 756
Merit: 250
ok i am testing it
4x 570 4gb itx

116mh phoenix 114mh claymore

pool reported 105mh with claymore
i will update pool reported hashrate about phoenix miner
member
Activity: 443
Merit: 13
Anyone else mining on ethermine and having serious issues with the pool? The Phoenix miner was finding 2 shares for a period of 10mins for the last couple of hours. I restarted the miner and it seems I cannot connect to the pool.
In Claymore, I can connect to the pool, but I'm getting around 1share/10mins Cheesy ?!!?!?
newbie
Activity: 31
Merit: 0
Those asking about OverdriveNTool, here is a quick message about it back earlier.  Some examples on how I use it.  I make my progfiles, very easy and then add the command to the start of each of my.bat files no matter what coin I load up, so it resetsall cards to default and the OCs each card in about 0.5 seconds and then runs miner.  I have.  I have my minspeed rest adn the same in my reboot.bet.  So whenever I load any miner to anypool for any coin, I reset all cards and overclock them at the start.  Very very easy and no better way to do it!

The dev is working those feature into the program but for now, use overdriveNtool by the guy on guru3d

Works great for all overclocking, undervolting and fan control. I mod all my bios with Polaris 1.6 but if I need to tweak anything after I use overdriveNtool

Could you please share a simple guide on how to undervolt with this tool mate? I don't want to mess up my cards Cheesy I don't know if I should even do it, since for the first two weeks my rig is finally stable *fingers crossed* and I don't know if that would cause any instability, that's why I'm asking. It just much simple with those commands to do it and not go in-depth, that's why I asked, I don't mind waiting tho. Electricity is 0.06usd, so 100-150w won't cause me to lose much of the profit. It just bugs me, that the cards are getting hot and the fans, therefore, are working at 60%, producing a lot of noise.
[/quote]

It is super easy, so open the program, you can goggle -command lines to find all the strings you can use.  You should know you frequencies you want to use and I have this in both my windows startup folder and at the start of each coin.bat:

Code:
cd\Users\Rig\Desktop
OverdriveNTool.exe -r1 -r2 -r3 -r4 -r5 -r6 -p1"570" -p2"570" -p3"570" -p4"570" -p5"570" -p6"580"
cd\rig\clay
EthDcrMiner64.exe -epool http://192.168.1.74:8645 -allcoins exp -eworker R0 -dcri 1 -r 1 -minspeed 175
pause


So I cd\ to the ODNT exe and then run -r on each card (be mindful of where you Intel GPU is at and don't call on it), that -r is a reset to default command so it takes each card in this one rig back to factory default, milliseconds lattr is runs the -p1, p2 commands which is load profile (p) on card number (1) "570" profile named 570. You can see I have a separate profile for the single 580 at the 6th slot.

The best part about this is that it resets all cards and then OCs them in half a second.  All of them.  And it does it when windows loads and when reboot.bat kicks in as well as whenever I load a coin to mine, as I start all .bat files with it.  Global reset and OC works great to clear failed GPUs, in literally a second.

The Program itself its pretty simple, load the GUI, enter your settings onto one card, click to save that as a profile, that will create a config file right in the same folder, you can go in there and make other profiles with copy and paste and edit them faster then in the GUI, but either way woek, edit config file or GUI.  That's how you make any number of profiles you want to call with the -pX function.  You can have a profile for each card in your rig if you want.

These are my two profiles in this one rig, the voltages need a little touchup, kind of high:

Code:
[Profile_0]
Name=570
GPU_P0=300;750
GPU_P1=588;765
GPU_P2=976;956
GPU_P3=1065;950
GPU_P4=1130;950
GPU_P5=1192;950
GPU_P6=1233;950
GPU_P7=1268;950
Mem_P0=300;750
Mem_P1=1000;800
Mem_P2=2200;900
Fan_Min=750
Fan_Max=3000
Fan_Target=70
Fan_Acoustic=910
Power_Temp=90
Power_Target=0

[Profile_1]
Name=580
GPU_P0=300;750
GPU_P1=600;769
GPU_P2=918;893
GPU_P3=1167;1050
GPU_P4=1239;1050
GPU_P5=1282;1050
GPU_P6=1326;1050
GPU_P7=1366;1050
Mem_P0=300;750
Mem_P1=1000;800
Mem_P2=2250;900
Fan_Min=1125
Fan_Max=3500
Fan_Target=75
Fan_Acoustic=918
Power_Temp=90
Power_Target=0

And I have found that these 'hold' better then using watt man, afterburner, claymore command line, sapphire, any of them.  Really rock solid.

If I could use Phoenix Miner with geth I would be getting 32 and 34.5 MHS on the 570s and 580s respectively but nope, I am suffering with Claymore still.  haha.
full member
Activity: 729
Merit: 114
Anybody tried to compare speeds with claymore with -dcri 9 vs phoenix miner?  default value of -dcri flag is 30 in claymore and -dcri 8 or 9 value boosts the hashrate when solo mining
Last time I tried claymore v10 still edged out 2.5d.  I will try again today and report.

-dcri 10 make me 166.0 mh/s at 855w
-mi 12 make me 168.7 mh/s at 865w

Thanks.  Just tried a quick test and yup it does give me more than Claymore v10.  Haven't checked on stability or pool side hashrate.

8 x RX570
Claymore: ~236.x Mh/s
Phoenix: ~237.5 Mh/s

6 x RX580
Claymore: ~186 Mh/s
Phoenix: ~187 Mh/s

EDIT:  1 day tests.

Going good so far.  Hash rates are consistent on poolside too.  Power consumption is more or less like Claymore.
member
Activity: 124
Merit: 10
Anybody tried to compare speeds with claymore with -dcri 9 vs phoenix miner?  default value of -dcri flag is 30 in claymore and -dcri 8 or 9 value boosts the hashrate when solo mining
Last time I tried claymore v10 still edged out 2.5d.  I will try again today and report.

-dcri 10 make me 166.0 mh/s at 855w
-mi 12 make me 168.7 mh/s at 865w
full member
Activity: 729
Merit: 114
Anybody tried to compare speeds with claymore with -dcri 9 vs phoenix miner?  default value of -dcri flag is 30 in claymore and -dcri 8 or 9 value boosts the hashrate when solo mining
Last time I tried claymore v10 still edged out 2.5d.  I will try again today and report.
member
Activity: 146
Merit: 10
i have not read the all post, i want to know is possible to use voltage, overclock, underclock directly from the bat file like i used to do with claymore?

you never read the readme file ? or the first post of this thread ?
often not never  Roll Eyes
member
Activity: 124
Merit: 10
i have not read the all post, i want to know is possible to use voltage, overclock, underclock directly from the bat file like i used to do with claymore?

you never read the readme file ? or the first post of this thread ?
member
Activity: 146
Merit: 10
i have not read the all post, i want to know is possible to use voltage, overclock, underclock directly from the bat file like i used to do with claymore?
member
Activity: 124
Merit: 10
After download, the antivirus show file has virus
anybody help me?

read this forum and make your idea yourself.
22 page for a virus !? how much users for this virus !!!?
newbie
Activity: 14
Merit: 0
After download, the antivirus show file has virus
anybody help me?
jr. member
Activity: 170
Merit: 6
I really hope the stale shares issue recoding works. I'm getting 2x or higher more state shares than Claymore.
It's quite obvious to see in the charts from ethermine.org.

newbie
Activity: 6
Merit: 0
"which probably means that you have to increase the size of your page file to at least 16 GB. Let us know if the increasing of the page file size (you have to restart Windows after that in order to apply the changes) doesn't solve the problem."

I had 25Gb pagefile size when i ran PhoenixMiners Sad
  Please try the new version 2.6b. It is possible that the new OpenCL initialization code introduced in 2.5d is not working well with your driver and if this is the case, 2.6b should solve the problem. Another possible way to fix the problem is to create a batch file to start the miner with the following contents (the important part are the setx statements to set the environment variables):
Code:
setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100

PhoenixMiner.exe -clKernel 1 -pool eu1.ethermine.org:4444 -pool2 us1.ethermine.org:4444 -wal 0xd8081f58c43bd84f9a7882f613a9d8947fb31d00.farm1
pause

I try PhoenixMiner 2.6b with recommended above parameters, on Farm1 PhoenixMiner didn't start and show the same error "clCreateCommandQueue (-6)" but on second Farm PhoenixMiner started up and started working. After restart PhoenixMiner on Farm2 I recive error "clCreateCommandQueue (-6)" again Sad
I try restart PhoenixMiner and MSI Afterburner several times and change parameter GPU_FORCE_64BIT_PTR from 0 to 1 and backward - i have achieved success several times (PhoenixMiner
 started UP sometimes)..

What i noticed, when Phoenix starts up normally Phoenix define 4GB VRAM, when Phoenix can't starts, it define only 3GB VRAM

Log when Phoenix can't start:
17568:15:09:59.153: main No CUDA driver found
17568:15:09:59.477: main Available GPUs for mining:
17568:15:09:59.477: main GPU1: Radeon RX 570 Series (pcie 1), OpenCL 1.2, 3 GB VRAM, 32 CUs
17568:15:09:59.477: main GPU2: Radeon RX 570 Series (pcie 2), OpenCL 1.2, 3 GB VRAM, 32 CUs
17568:15:09:59.477: main GPU3: Radeon RX 570 Series (pcie 4), OpenCL 1.2, 3 GB VRAM, 32 CUs
17568:15:09:59.477: main GPU4: Radeon RX 570 Series (pcie 5), OpenCL 1.2, 3 GB VRAM, 32 CUs
17568:15:09:59.477: main GPU5: Radeon RX 570 Series (pcie 6), OpenCL 1.2, 3 GB VRAM, 32 CUs
17568:15:09:59.491: main ADL library initialized

Log when Phoenix started normally:
17569:20:12:50.903: main No CUDA driver found
17569:20:12:51.179: main Available GPUs for mining:
17569:20:12:51.179: main GPU1: Radeon RX 570 Series (pcie 1), OpenCL 1.2, 4 GB VRAM, 32 CUs
17569:20:12:51.179: main GPU2: Radeon RX 570 Series (pcie 2), OpenCL 1.2, 4 GB VRAM, 32 CUs
17569:20:12:51.179: main GPU3: Radeon RX 570 Series (pcie 4), OpenCL 1.2, 4 GB VRAM, 32 CUs
17569:20:12:51.179: main GPU4: Radeon RX 570 Series (pcie 5), OpenCL 1.2, 4 GB VRAM, 32 CUs
17569:20:12:51.179: main GPU5: Radeon RX 570 Series (pcie 6), OpenCL 1.2, 4 GB VRAM, 32 CUs
17569:20:12:51.192: main ADL library initialized

My rigs have 2GB RAM Memory (900-800 MB free when all works)
Sorry for my english)
member
Activity: 413
Merit: 17
Why would anyone want to control their cards clocks via miner, when there are tools like MSI AB?
Laziness.
full member
Activity: 226
Merit: 100
Why would anyone want to control their cards clocks via miner, when there are tools like MSI AB?
Jump to: