Proof: http://zpool.ca/?address=1PHSDYvVp6HpqtuUPocK41DrdeHbbezaeP

Zero disconnects.  I see not what you speak of... at all.

Stop blaming your machine/internet on the pool.....   Thanks.....

a hot power cord is a bad sign.... 

This is why it can be less profitable to multi-algo or profit switch mine.

Also;  wait for the vert share;  it will delta just the same.... depending on how much hash was thrown at it after you left, and how big of a share it was.  I am betting you are taking zero of this into account when assuming how much credit you should be getting.


You obviously do not see the whole picture..........

Sweet, I gave him the heads up.  I'm sure being compatible with all GPU's, these would go hand-in-hand with pico PSU arrangements and rigs with high device density.

I plan to use his power supplies [as needed] to fill out my B250 D8P I got from you.  I already have a pico PSU that appears to be able to work on the board as it was able to handle my I7-920 on a X58 motherboard running with its 3.2Ghz overclock..  (waiting to get a cpu and ram to test the onda with it next but im sure it wont draw too much either).


yeah, L is low voltage;  most likely specced for processors like M's and atoms.

*edit* the key is its either 1.5, or 1.5/1.35v memory.

@phil;

I got a reply from my buddy about cabling for the power supplies;  he can offer a 6 pin to 2x8 pin cable with them instead of 6 pin to 6 pin....   Or would just a six pin to a single 8 pin be best?

I gave him a heads up about the thread and the benefits of a 6->8 pin connector.  He may come here and reply =)

1 separate IEC cable per 750w PSU... so 3x IEC plugs.  The psu's card edge just has one large bank of +12V, or was it two;  I should have gotten a photo of the label.

Im not sure about that whole conversation myself....   Im not sure exactly what happened there.   I thought initially you were replying about mining MBL....  my mistake

MY buddy has a power supply kit selling right now for a reasonable price; so I thought Id share with you guys:

The kit has a 2250W 12V capability;  and are ~$155 shipped.

They are in the form of 3x server modular PSU with breakout boards and 6 pin PCIE pigtails.... He has them arranged for use in antminers, avalons, etc....

https://www.ebay.com/itm/282378575981

LMK what you guys think  about the price and product....   Ive seen his stock;  he bought cases of the PSUs (they all look brand new), the PCIE cables are brand new (not cheap flimsy ones), and the breakout board is professionally made (in china).

Presently hes having them make breakouts for the other 1600w+ PSU's he has been looking into sourcing for similar kits.

I have also given him the heads up about GPU miners on how they need to have 8 pin plugs, and he should consider adding support for the 8 pin plugs, not just 6 pin.

Im not sure what his margins are on these power supplies, but maybe I can do bulk discounts for you or something.... need to check with him if there's interest from one of you guys.

Please do yourself and us the favor and look up what AuxPOW is for coins, and how pools use it in switching code...  Not earning these coins is "wasting hash" compared to earning them with a hash result that would not work on a big coin.

Also read this whole thread;  there are excellent descriptions and analogies tht explain why its good to get these credits that you think are worthless.


The server can't be configured differently because of the way it calculates share % based on submitted shares.   It must delta up the hashrate (poolside) to be sure people arent whipping in with erroneous miner behavior to try and trick the calculations into giving you hashrate that wasn't submitted.

This is your miner ramping up... not the pool.

it takes time to average.  Everyting you read from the miner output is from your machine;  the only thing the pool reports to you is your next workunit, and a response for a share submission.  hashrate on the pool is based on share size/frequency.  The reading from your miner is from your miner's physical performance, not the calculated performance.

If you spam a bunch of low shares, you get low hashrate.  if you get regularly spaced or spam high diff shares, you get a high hashrate....  pools dont trust any hashrates reported by miners.   easy to fake the statistics.

agreed.

And, the other aspect people misconstrue about profit/algo switching;

different algos have different power requirements.  Some stress almost every single register/function/transistor in the GPU; and thus take WAAAAAAAY more power than others that have simpler more repetitive functions and are done with the GPU.

People could falsely count this as stress from profit switching;  but this in turn could be related to the before mentioned;  overclock (most times different over/underclock is needed per algo), power supply/condition, etc etc etc.

Just adding one external fan in my rackminer case sucking air from the gpu/OE riser locations.... has drastically improved my temps in this box.   I have a watercooled cpu, and 3 9-series video cards in it.  And let me tell you;  I can and do use it as a space heater no problem.  It has the OE rosewell fans on it (low speed and quiet), and has a similar fan added to the outside I mentioned above.  Its a touch noisy if I turn on the power supply "fan increase" button.... but, I can live with that.    This sucker heats the whole RV I live in on the ranch.... almost too much heat at times, but as they say:

Winter is coming......

*edit*  The only downside, is if the new series of GPU is incompatible with the old waterblocks....  Which kinda feels like a hit, but when its reused, its a godsend.  Either way, I think watercooling is the smartest avenue in the end.  Just keep a res tank outside;  put a radiator on it if you feel the need.   then just have a pumpable loop you can tap into on the inside of the building.   Or separate the water/device zones if doing it all indoors.

I have an old theater coolant tank/pump I saved from when they closed down a theater I worked at.   I figured just mount it outside and have it routed indoors to my machines.


Just gotta spring for the waterblocks and fittings at that point.  Depending on the flow of the primary pump, you could have many different conceptual configurations for a farm, such as a pressure/dump system...

Its good practice to tell stratum the algos you are mining;  because if you are mining multiple algos you'd add them all to that list, and give them a normalization value as well.

Its just good practice in the end, as the pools code expects to see something there, and you are better off not relying on a default or unsubmitted value that it will then assume.


My nvidia machines are online and mining.....  not sure what you are trying to imply.

switching software does nothing to mining gear.

What effects mining gear the most, is temperature, dust, and power.  The software; just dictates what its doing and when for the most part.

If you run your gear overclocked and hot;  don't expect it to last as long as stock settings and a proper temp environment.

I killed the fans on my GTX980 recently because I closed up the rackminer case with improper ventilation and it was causing the 980 to have 100% fan speeds for a month.  That was enough to kill one of the fans, and the other is about to go;  my replacements should be in the mail tomorrow....   But had I left them in the position to be running at <80% fan speeds;  the fans would still be good.

Thresholds are the same; aside from decred; which is noted on the main page.


generate a new unique BCH address, and select payout in that;  but that is assuming BCH is on the list of coins the pools actually mines.

Crackfoo can confirm.


correct;  you are only selecting your payout coin type.  you mine on an algorythm; not a specific coin.  you only choose what coin you get paid in.

password

or if you wanted a manual difficulty (not recommended unless you know what you are doing):


Cheers.

I will never buy a brand new asic ever again unless it does more than one algo, or is brand new tech securely in my hands at expected release time with no funny business like bitmain and other companies have done....


so like the board I bought from phil;  brand new, untested (mostly,), etc....  but known source and right away delivered.

If I had that in an ASIC that had more than a chance of not getting profitability screwed like normal batch1-2 users... then i may consider....   but... I dont trust asic profitability anymore.

I bought my s3/s5/s7, ran them, then sold for what I bought them for... so.... I got lucky.

Probably disabled;  just yesterday we all had earned a credit of like 69K bitcoins... so... yeah... probably forked and resyncing.

Make sure the board takes m.2.....   It could be msata.... same port, different protocol/pinout.   I know my Onda B250 D8P is labeled for msata, not M.2/NGFF.

As for the front molex/sata ports, I would suspect they have a detect/sense wire on each plug to verify that they are powered (similar to how video cards wont power unless all/key pins see power/gnd on the connector.)..

I still need to get an ECU and memory for my board, so I can't comment... and I do know its not the same D1800 you have, but the construction and design between these two models looks pretty close, so I wouldn't be surprised if they are not all that dissimilar.

I'd probably classify that integrated processor as the same class as an "Atom" or embedded cpu....   What model does the cpu report as?

Ive used these cases, but the ones that are $100-110.  This one it looks like they just removed the front fand and added a crossbar to screw the gpu edge to.....  idk if its worth the extrs $ because I can make my own bracket... but YMMV.

https://www.youtube.com/watch?v=ZmCEXbspWuI

Im just gonna throw this out there.....