Pages:
Author

Topic: My second ZEC + XMR+ ETH thread builds info links thoughts and photos. - page 19. (Read 148003 times)

legendary
Activity: 1894
Merit: 1087
the righ has been running smooth for weeeks but now my hash rates are all over
legendary
Activity: 1894
Merit: 1087
can someone please help me,

my GPU temps are thortling (one minute theyre 75 degrees then theyre 60 degrees)

im loosing hashrate too

what the heck is going on I have 3 rigs and this is only happening on one

there is decent airflow and spacing, not sure what I am doing wrong

:S


List all three mobos
List all three psus
List all gpus

How do you set the gpus?  Flashed bios or msi afterburner?

It could be one bad card on that mobo.
If could be one bad slot.
It could,be one bad riser.

2 of your rigs (with different cards)

and also a third 6 card rig

no fashed bios all stock

hashrates are fucking terrible 5mh to 20 mh per card
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
can someone please help me,

my GPU temps are thortling (one minute theyre 75 degrees then theyre 60 degrees)

im loosing hashrate too

what the heck is going on I have 3 rigs and this is only happening on one

there is decent airflow and spacing, not sure what I am doing wrong

:S


List all three mobos
List all three psus
List all gpus

How do you set the gpus?  Flashed bios or msi afterburner?

It could be one bad card on that mobo.
If could be one bad slot.
It could,be one bad riser.
legendary
Activity: 1894
Merit: 1087
can someone please help me,

my GPU temps are thortling (one minute theyre 75 degrees then theyre 60 degrees)

im loosing hashrate too

what the heck is going on I have 3 rigs and this is only happening on one

there is decent airflow and spacing, not sure what I am doing wrong

:S
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
I ordered a higher quality  psu to test the panda miner


I got an evga 1600 g2 due on fri and the panda miner due on fri.


If this miner comes with a Cantonese or Mandarin Chinese gui only I am going to freak.

I will post in this thread and in the Panda thread

https://bitcointalksearch.org/topic/pandaminer-b3-pro-new-batch-releases-with-only-800-1750-1668461



the gear is in Cincinnati

lol... if its GUI is not available in English then I guess that will tell us all what we need to know about this company  Wink

"GUI" is likely standard windows 10 or 7, then just change the Windows pack language to English.... I am writing this with a grin on my face.... I think a foreign "GUI" adds more excitement to the review  Grin

Btw, EVGA-1600w-G2 (Gold) is a very good PSU, I have 8 of these in my farm for the 390s rigs, its expensive but it doesnt let me down - its power cord alone are the ones used in data centres. Its bigger brother 1600w-P2 (Platinum) is even more costly - could buy a decent 480 for that price.

Yea I went strong on the GPU  I have owned the 1600watt T2 the 1600 watt P2  the 1600 watt G2  all are solid consumer gpus

all should be able to run this unit
legendary
Activity: 1834
Merit: 1080
---- winter*juvia -----
I ordered a higher quality  psu to test the panda miner


I got an evga 1600 g2 due on fri and the panda miner due on fri.


If this miner comes with a Cantonese or Mandarin Chinese gui only I am going to freak.

I will post in this thread and in the Panda thread

https://bitcointalksearch.org/topic/pandaminer-b3-pro-new-batch-releases-with-only-800-1750-1668461



the gear is in Cincinnati

lol... if its GUI is not available in English then I guess that will tell us all what we need to know about this company  Wink

"GUI" is likely standard windows 10 or 7, then just change the Windows pack language to English.... I am writing this with a grin on my face.... I think a foreign "GUI" adds more excitement to the review  Grin

Btw, EVGA-1600w-G2 (Gold) is a very good PSU, I have 8 of these in my farm for the 390s rigs, its expensive but it doesnt let me down - its power cord alone are the ones used in data centres. Its bigger brother 1600w-P2 (Platinum) is even more costly - could buy a decent 480 for that price.
sr. member
Activity: 600
Merit: 261
Do someone have ever had a moded RX470 stucked at 21MH/s (Eth) instead of the usual 27MH/s expected ?
All my Rx470 are between 26 and 27 , but two of the are stucked at 21 , no matter what clock I choose.
Maybe someone has already met this situation ?

What is the hashing rate with original rom? Does changing frequencies with original rom affects on hashing rate?

Yeah, that's along the lines of what I was thinking too ps_jb.  Mainly because at least most of my rx 470's get right around 21 on ETH right out of the box. Sure, this could be coincidence but I would at least take a crack at reflashing them with original bios then reflashing with the strap-1500 mod to see if that helps. Also, are you 100% sure that you modded these GPU's bios (were they ever running at 27)?  I learned first hand that trying to flash bios with multiple GPUs attached can be a bit tricky, so maybe you re-flashed a GPU that had already been modded instead of modding the intended one(s)?  Just a thought.
sr. member
Activity: 600
Merit: 261
I ordered a higher quality  psu to test the panda miner


I got an evga 1600 g2 due on fri and the panda miner due on fri.


If this miner comes with a Cantonese or Mandarin Chinese gui only I am going to freak.

I will post in this thread and in the Panda thread

https://bitcointalksearch.org/topic/pandaminer-b3-pro-new-batch-releases-with-only-800-1750-1668461



the gear is in Cincinnati

lol... if its GUI is not available in English then I guess that will tell us all what we need to know about this company  Wink
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
I ordered a higher quality  psu to test the panda miner


I got an evga 1600 g2 due on fri and the panda miner due on fri.


If this miner comes with a Cantonese or Mandarin Chinese gui only I am going to freak.

I will post in this thread and in the Panda thread

https://bitcointalksearch.org/topic/pandaminer-b3-pro-new-batch-releases-with-only-800-1750-1668461



the gear is in Cincinnati
hero member
Activity: 742
Merit: 500
Do someone have ever had a moded RX470 stucked at 21MH/s (Eth) instead of the usual 27MH/s expected ?
All my Rx470 are between 26 and 27 , but two of the are stucked at 21 , no matter what clock I choose.
Maybe someone has already met this situation ?

What is the hashing rate with original rom? Does changing frequencies with original rom affects on hashing rate?
newbie
Activity: 64
Merit: 0
Do someone have ever had a moded RX470 stucked at 21MH/s (Eth) instead of the usual 27MH/s expected ?
All my Rx470 are between 26 and 27 , but two of the are stucked at 21 , no matter what clock I choose.
Maybe someone has already met this situation ?

If you change the PCIE slot, is it the same?
hero member
Activity: 501
Merit: 500
Do someone have ever had a moded RX470 stucked at 21MH/s (Eth) instead of the usual 27MH/s expected ?
All my Rx470 are between 26 and 27 , but two of the are stucked at 21 , no matter what clock I choose.
Maybe someone has already met this situation ?
hero member
Activity: 742
Merit: 500

nice  I am getting the 8 card miner from panda on dec 30th
https://bitcointalksearch.org/topic/m.17335548

the machine and your software may be a match


That are the great news! Cool thing to play around New Year Eve Smiley

legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
Is their any resource that explains the main contributing factors to the difficulty of a given coin?  My impression as a newbie to all of this is that it is either solely or mostly driven by the volume of miners pointed at a given coin.  For example... price of XMR goes through the roof, everybody and their brother switch their rings to mine XMR in the last 2 days and bam! difficulty rises 20% in that time period. Similar thing when ZEC came out and most people started mining it... both XMR & ETH difficulty bottomed out.  Is this the only main driving factor or are their other things to consider?

Well.. That is one factor and the last factor is reward. Every X amount of months / years, the reward halves.

As Hotmetal stated,  the overall network hashrate is adjusted every period [variable per coin] to ON AVERAGE yield the same amount of coins.  If the network hashrate increases by 20% because everyone's brother switches over, you can expect ON AVERAGE that your payout is going to decrease by 20%.  Each pool will have a "local" average on top of the network average.  I think this (variance from expected payout) is much more pronounced with coins that have a longer block time (a.k.a. bitcoin [10 minutes!]).  With coins with a very short block time (a.k.a. ETH [15 seconds!]) the variance is naturally much less.  I hope I explained this well enough.

Switching programs do have a downside  if everyone uses one all edges are lost.
and they can cool off a hot coin.

I recommend 30 to 60% on a switcher and 70 to 40% on your hope for a hot coin.
I also plug zec as  the market cap is tiny  well under 25 million when I last looked.  

I could pump the coin and thats saying a lot as I sure as fuck am not spending 100,000k on a pump,but no coin has this  tiny market cap except for  zec.

I've figured out a pretty novel way of doing predictions (without using difficulty or market value) which could prove extremely profitable if you're mining and holding a coin for up 4 weeks (worst case).
However there are other things to finish up before diving into that. I've done most of the most profitable auto-switching code server side already .. Now to start working on server API and Linux / Windows side code.

nice  I am getting the 8 card miner from panda on dec 30th
https://bitcointalksearch.org/topic/m.17335548

the machine and your software may be a match

and btc coins are pushing 1000 usd which seems to be helping all  the alts.

sr. member
Activity: 378
Merit: 250
Is their any resource that explains the main contributing factors to the difficulty of a given coin?  My impression as a newbie to all of this is that it is either solely or mostly driven by the volume of miners pointed at a given coin.  For example... price of XMR goes through the roof, everybody and their brother switch their rings to mine XMR in the last 2 days and bam! difficulty rises 20% in that time period. Similar thing when ZEC came out and most people started mining it... both XMR & ETH difficulty bottomed out.  Is this the only main driving factor or are their other things to consider?

Well.. That is one factor and the last factor is reward. Every X amount of months / years, the reward halves.

As Hotmetal stated,  the overall network hashrate is adjusted every period [variable per coin] to ON AVERAGE yield the same amount of coins.  If the network hashrate increases by 20% because everyone's brother switches over, you can expect ON AVERAGE that your payout is going to decrease by 20%.  Each pool will have a "local" average on top of the network average.  I think this (variance from expected payout) is much more pronounced with coins that have a longer block time (a.k.a. bitcoin [10 minutes!]).  With coins with a very short block time (a.k.a. ETH [15 seconds!]) the variance is naturally much less.  I hope I explained this well enough.

Switching programs do have a downside  if everyone uses one all edges are lost.
and they can cool off a hot coin.

I recommend 30 to 60% on a switcher and 70 to 40% on your hope for a hot coin.
I also plug zec as  the market cap is tiny  well under 25 million when I last looked. 

I could pump the coin and thats saying a lot as I sure as fuck am not spending 100,000k on a pump,but no coin has this  tiny market cap except for  zec.

I've figured out a pretty novel way of doing predictions (without using difficulty or market value) which could prove extremely profitable if you're mining and holding a coin for up 4 weeks (worst case).
However there are other things to finish up before diving into that. I've done most of the most profitable auto-switching code server side already .. Now to start working on server API and Linux / Windows side code.
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
Is their any resource that explains the main contributing factors to the difficulty of a given coin?  My impression as a newbie to all of this is that it is either solely or mostly driven by the volume of miners pointed at a given coin.  For example... price of XMR goes through the roof, everybody and their brother switch their rings to mine XMR in the last 2 days and bam! difficulty rises 20% in that time period. Similar thing when ZEC came out and most people started mining it... both XMR & ETH difficulty bottomed out.  Is this the only main driving factor or are their other things to consider?

Well.. That is one factor and the last factor is reward. Every X amount of months / years, the reward halves.

As Hotmetal stated,  the overall network hashrate is adjusted every period [variable per coin] to ON AVERAGE yield the same amount of coins.  If the network hashrate increases by 20% because everyone's brother switches over, you can expect ON AVERAGE that your payout is going to decrease by 20%.  Each pool will have a "local" average on top of the network average.  I think this (variance from expected payout) is much more pronounced with coins that have a longer block time (a.k.a. bitcoin [10 minutes!]).  With coins with a very short block time (a.k.a. ETH [15 seconds!]) the variance is naturally much less.  I hope I explained this well enough.

Switching programs do have a downside  if everyone uses one all edges are lost.

and they can cool off a hot coin.

I recommend 30 to 60% on a switcher and 70 to 40% on your hope for a hot coin.

I also plug zec as  the market cap is tiny  well under 25 million when I last looked. 

 I could pump the coin and thats saying a lot as I sure as fuck am not spending 100,000k on a pump,but no coin has this  tiny market cap except for  zec.
hero member
Activity: 615
Merit: 500
Is their any resource that explains the main contributing factors to the difficulty of a given coin?  My impression as a newbie to all of this is that it is either solely or mostly driven by the volume of miners pointed at a given coin.  For example... price of XMR goes through the roof, everybody and their brother switch their rings to mine XMR in the last 2 days and bam! difficulty rises 20% in that time period. Similar thing when ZEC came out and most people started mining it... both XMR & ETH difficulty bottomed out.  Is this the only main driving factor or are their other things to consider?

Well.. That is one factor and the last factor is reward. Every X amount of months / years, the reward halves.

As Hotmetal stated,  the overall network hashrate is adjusted every period [variable per coin] to ON AVERAGE yield the same amount of coins.  If the network hashrate increases by 20% because everyone's brother switches over, you can expect ON AVERAGE that your payout is going to decrease by 20%.  Each pool will have a "local" average on top of the network average.  I think this (variance from expected payout) is much more pronounced with coins that have a longer block time (a.k.a. bitcoin [10 minutes!]).  With coins with a very short block time (a.k.a. ETH [15 seconds!]) the variance is naturally much less.  I hope I explained this well enough.
sr. member
Activity: 378
Merit: 250
Is their any resource that explains the main contributing factors to the difficulty of a given coin?  My impression as a newbie to all of this is that it is either solely or mostly driven by the volume of miners pointed at a given coin.  For example... price of XMR goes through the roof, everybody and their brother switch their rings to mine XMR in the last 2 days and bam! difficulty rises 20% in that time period. Similar thing when ZEC came out and most people started mining it... both XMR & ETH difficulty bottomed out.  Is this the only main driving factor or are their other things to consider?

Well.. That is one factor and the last factor is reward. Every X amount of months / years, the reward halves.
sr. member
Activity: 600
Merit: 261
Merry Christmas everyone. Just wanted to take a moment to thank all of you on this thread who have been very kind, patient and helpful educating a mining newbie like me over the past 2 months. I am still not where I want to be but have learned a lot and hope to continue that and leverage it into a profitable 2017.

Here's a quick update on my never-ending battle to get 6 GPU's running on my 2nd rig... yesterday, I took another shot at strap-modding the bios of my MSI 470s.  My first attempt about a week ago to mod GPU #0 didn't work for some reason and I had to flash it back to original settings.  This time, I unplugged my orphan Nitro, rebooted, and flashed dev #1 with the modded bios.  Reboot went fine but upon checking DevMgr, found the dreaded error 43 on presumably the card i flashed. Tried running the 6xGPU mod a few times but no dice, so I re-flashed it with original bios, rebooted, plugged the Nitro back in I'm back to "normal".

What i don't get is why the exact strap-mod I used on the exact same GPUs on my other rig (same W7) doesnt wanna work on this rig... 0 for 2 now. I'm sure I am doing it right but this rig doesn't wanna take it.  The only real difference is the Mobo between the 2 systems.  Anyway, I'm at the point where it probably makes sense to strip this thing down, run the driver cleaner, then re-install the GPUs one by one, maybe flashing each as I go.

But, then I got to thinking... if I'm going to basically re-build this thing, I might as well install Windows 10, since I have heard from numerous people that this is a better OS for anything over 4 GPUs.  And, it looks like I can still get an upgrade for free so probably worth the shot.  This time I am going to try plugging the 1st GPU (and monitor) into slot #2 (x16) since starting with slot #1 last time may have been causing some of my issues. Hopefully, W10 will let me get 6 GPUs running and modded. That should get me about 40% more hash-rate on XMR vs what that rig is currently doing.

So, I got my 6th 470 for rig #2 so will probably take a crack at moving to W10 and rebuilding this thing today.  This is my plan but I wanted to ask for opinions as to whether or not this is over-kill.  Note that current status is W7, max of 5 GPUs, 2 unsuccessful attempts to strap-mod bios on 2 separate GPUs (exact ones that I have done 4x on my other rig).

1- Power down and unplug all but 1 GPU. Sole GPU will be plugged into PCIe slot #2 (x16), monitor plugged into GPU via HDMI cable.
1a- I suspect this might not work given the issues i have had with slot #2, so will plug into slot #1 initially if needed, then try slot #2 as primary again after W10 install.
2- Boot up and upgrade to Windows 10 before doing anything else
3- Load Crimson drivers from AMD site for 1 plugged in GPU
4- Test GPU in Claymore miner, probably ETH since it shows individual GPU speed, unlike XMR miner
5- Flash that GPU with bios strap-mod, overclock & under-volt, then re-test in ETH miner to ensure speed increase
6- After ensuring all is working correctly, power down, unplug GPU from slot #2 and plug in another GPU into that slot
7- Repeat all of the steps above except for #3 since the drivers will already be on the machine
8- After all steps done for all 6 GPUs, start adding GPUs 1 at a time until up to 6 (hopefully)
9- If I get the error 43 code at GPU #5, install the 6xGPU registry mod app

This is obviously going to take the better part of a full day.  The thing that could save significant time would be to skip the 1 at a time method and add GPUs incrementally after getting the first 1 to work. This would make bios flashing a bit tougher, but I am under the impression that windows should add GPU's device numbers in order, so maybe this could save some time. Meaning, get GPU#1 working after flashing bios mod on device #0, then add the 2nd GPU and flash device #1, etc, etc.

At this point, I am no planning to mess with the bios on the orphan Nitro. If I can get 6GPUs and the 5 MSI's bios modded on this rig, I will be very happy.

Anyone see anything wrong with this approach or have any suggestions?


Just wanted to update the status my move to Windows 10 on my 2nd rig, in case someone already here or who finds it in the future was grappling with the same issues as I was.

Firstly, I decided to go with the "add each card incrementally" approach instead of swapping out 1 by 1 to mod the bios then replug them all 1 by 1... and this worked, at least for the most part.

After upgrading to W10 (took a bit over an hour), the main thing I did differently from the W7 build was start with the x16 slot #2 on this mobo.  In fact, I did this even before updating windows (had only 1 card plugged into the x16 slot) and it actually worked for the first time on this rig!  So, after installing W10, just left it there and it worked fine.  From there, I just added GPUs 1 by 1, testing with Claymore ETH miner, then flashing bios, then rinse and repeat.  Was really worried before turning it on with GPUs #5 & 6 plugged in, but W10 had absolutely no issue with this whatsoever! So, I now have MSI 470 4GB Gamings in slots 1-5 and the Sapphire Nitro 470 8GB in slot #6.

The one unexpected thing I did run into was this... when copying the ROM from my newest MSI GPU (just opened &received a few days ago), I noticed that under memory type it actually showed Samsung, not Hynix as my other 8 MSI 470's showed.  The other interesting thing is that this GPU actually got about 2 Mh/s more on ETH out of the box than my other MSIs.  But, since I wasn't sure if that Samsung vs Hynix memory could affect the 1500 memory strap, I just left it un-modded, so it is running about 4 MH/s slower on ETH (over-clocked & under-volted) compared to the other MSIs on this rig. Does anyone know if this will affect the memory strap mod at all?

So... bottom line is, testing on ETH miner, I switched over to Claymore XMR 9.6 and actually getting almost exactly as expected, a bit over 4Kh/s (~700 each from the 4 modded MSI and ~600 each from the un-modded MSI and un-modded Nitro). So, very happy with this so far and maybe a bit of room to tweak for further improvement.

No, all totaled with 10 GPUs and 3 i7 K's, I am getting about 7.6 Kh/s on XMR, using about 1,260 watts total. At current rates, about a bit over $10 profit/day (not counting BurstCoin mining). This is the best it's been in the 3-4 months I've been involved in mining so hoping it lasts a while  Wink
hero member
Activity: 742
Merit: 500
I was just at buysolar's house.

send him pm's with any solar questions.

Here is what I can offer.   Batteries are costly and die fast.  You need a lot of them and like a can of gasoline they are stored energy (i.e fire + explosion hazard)

Buysolar and I are in a good state (New Jersey) with good solar energy rules he is tied into the grid.  So we sell excess in the day and buy back power at night.

Some states are terrible with grid setups some are better.



Battery only makes sense if you can score batteries cheaply.  Plus place them in a place that does not burn down.


Another thing to think of is hail storms.  NJ  can go 10 years without a hail storm.  I personally have live here for 24 years  and have not seen a storm with dime sized hail.

I am sure many midwest states and central states are laughing at that.  I had a friend in Kansas that has seen hail bigger then golf balls more then 10 times in her life.  I would think that size hail would be hard on a panel.

Thank you, Phil and buysolar!

LiFePO4 batteries are not flammable and can be purchased from China relatively cheap (all we know that it means to buy from CN, but it is really good price). These lithium phosphate batteries lifetime is around 10 years.

I thought about selling excess of electricity to grid - but rates in NM are not very good.

Hail state around us is Colorado. Those guys are really bad with that!
Pages:
Jump to: