I am sooo tempted to buy that D8p and a 1080 from you;  but I really need to keep biting my tongue and wait for the press release in a few days/weeks about the new generation of hardware.......

Ill be left with one last GTX960 shortly; and at home that will leave me with just it plus my 1070 until I buy more cards.....

if the old board is bad/damaged, I may be able to figure out where the missing link is and add botch-wires to fix the broken/open traces.... but something inside my head says its probably a bad solder on a choke/capacitor or something in the link from the molex/SATA to the PCIE slots (their solder flow job is definitely minimal).... either way, if its just power missing, it should theoretically be repairable...  Having a functioning D8P board, I could compare traces to find the fault.


A powered riser didn't allow the machine to see a card plugged into it?   If so; This itself would probably be the worst case because the PCIE lines themselves would be bad;  but wince the power didn't even make it to the PCIE slot, I am willing to gamble that that my main theory is the issue at hand...

www.cryptomined.com sells the D8P, as well as fully enclosed (and rather nice might I add) cases specifically for the D8P.   I think they also have a case that can handle all 12 GPU on this board.

This is what I bought.
Amazon Link

I figured ~22.50 per fan was worth it.... to get it without the ebay fuss.

assuming you are BTC as payout type:

c=BTC,myr-gr

should be all fine from there.

if you want a worker name you can see on the wallet page:

workername1,c=BTC,myr-gr

Well, I saw nice gains ~16 hours ago up to a few hours ago... and now a run of bad luck....    It seems to stay more consistent than NIST5 so far... im planning on keeping my miners on xevan for now until things change.   XLR seems to be up in value recently as well.

That alert is just because; for whatever reason, the batch was unable to contact the hosting server for the algo name....   may have been your connection, may have been tripod... im unsure exactly what at the moment... but it should have started to show up correctly again.... did it?   My machines are currently able to access it fine.

At least that's progress on figuring it out.

do you have expeirence with an ohmmeter and doing trace continuity checks?

I can take some measurement son my board, and see if yours correlate.....   maybe we can find the component that has a bad solder (what I would expect to be the case)



Oah and another good thing to try:
use a powered riser on each one of the three slots with a single video card;  see if its the slot power (which is what it seems) or if its the slot PCIE lanes that are knackered.

have you tried just one gpu;  rotating slots one at a time to see if its just the slots causing the problem?

What cpu? (just for knowledge)


Have you tried any BIOS setting changes?   I'm pretty sure its supposed to work right out of the box.... so not too sure about this being of help.  But who knows, maybe.

If I can help you find a solution, I will.

I have a populated board with no video cards to provide as a test-case if need be.

are you powering pairs of the molex/SATA plugs independent on their own cables, and all of them?  I know they are specifically for powering most of the 16x and all of the 1x slots.

I know GPU slot 1 (and possibly 2) share with the CPU from what following the traces on the board have shown me...  Its been a few weeks since i've looked at it.

PRoblem with those is if you don't temperature control them, you get condensation on the board like none-other.

Once I finish making a desiccant housing for it;  Ill be re-mounting my TEC to my WWAN adapter used for my home internet.  When I am uploading a file, or streaming;  it gets stupid hot.  The TEC kept it's whole housing at 36-38*f; but condensation buildup was insane.    The heatsync on the other side of the TEC came from an old AMD x64 machine, and was passively cooled.  I need a desiccant box to control it and run it that way any further.

I would think that using water cooling and having an internal TEC;  no radiator at all... could be an interesting product line if it can be made efficient and small enough.  IIRC the power draw from a TEC isn't near as much as lets say a typical heat pump.

Hey spotswood;  I would like one of these.  I want the single stacked version with a boxed frame all the way around;  so I assume its the original with all support beams.  I just want it to be a cube of sorts for when its not in a rack case.

If you have an easy way for it to mount into a 4U slot;  that's an option I would like added.  I don't mind if the cards are facing 90* in the rack mount;  I plan to have active cooling running from front to back of the rack anyways, so should be fine.  I just want to be able to mount fans across the backs of the cards like your renders;  but don't want a double-high unit like phil got.

If we are on the same page about what i want, send me a PM with a BTC address and a BTC amount to cover it being shipped to 95037;  I'll transfer some funds your way when I get the PM =)

found them and started them?  or code42?

good to average right now.   Things can change though... just gotta keep an eye on what people are talking about; and watching the algos on zpool stats to see where the bulks of hashrate jump to now and then.

Takes a while to figure out.

gotcha.  Makes more sense now.

My only guess is your return traffic is being blocked/filtered on the way back somehow;
or there is a problem after the node 198.27.73.218

the discord reply does make sense.  Hopefully it irons out.

Yep; can't say which one, but  they were upgrading; and I became the disposal service....  The contractor was originally instructed to destroy the units and throw in the dumpster....  I ended up with many amps and now these speakers by getting in the middle of the process.  I may have the same sort of thing on a 20+ screen location happen as well.  if that's the case, im looking at ~90 more amplifiers and ~250 more speakers........ that being the case we would take out a loan on a warehouse to store them locally.... and be at the flea market every weekend.  Just sell one pair and it was more than worth the trip to try and sell them.   Given the brand/quality,  easy sell for sure.

We already have a guy who came to inspect them so he could talk to his other guy and make a bid on a lot of 36 of em.  We expect him to lowball us to ~50/unit;  but I think we are going to hold out for an absolute low of 120 for bulk (target 160-180), and we are going to be listing and holding onto many to get what we can out of em.

Given the brand and quantity;  should be quite a bit cash.  My business partner is taking his % in speakers for himself;  and the ones that sell through our ebay, he will get a %, but all in all;  pretty much easy $.

Im not gonna count it until I have it, but given how much I am owed by my buddy, and how easy these things will sell;  I'm sure I should be able to deck out 1-2 rigs of new hardware right when it releases.  All I need to do now is get a spotswood frame for my D8P and i'm off to the races.

im holding out for the new gen of GPU....

This just fell in my lap the other night;  so it should definitely fund the new hardware purchase.

76x   EV SL10-2V...

https://imgur.com/a/gFdPR

Are you using a VPN service?  I know some botnets have been getting banned, so if they are using a public VPN or gateway;  could be banned from that type of things.  Can't say for certain at all about that though.  This line is pure guesswork.

the first photo looks appropriate;  for a lack of connection to the pool.

The second photo is common when the pool is processing payouts, DB maintenance, etc....   Give it a few minutes and it usually responds fine after.  I assume it loaded for you in the past.

third photo:  don't ping the port, just ping the URL.

Also do a traceroute on it.  It will provide more information on the problem.  If you are not able to see it;  your DNS may have it blocked or filtered.   a traceroute should answer some questions.

I'm just gonna say two words and leave a little link.

Voluntary Compliance

This will only go to reinforce how corrupt our government is; and how twisted some people can make the actual definition of two words be 100% the opposite of its english definition.

*edit*  mind you there is no federal law or statute that requires you to pay a tax, or imposes a penalty for not paying said tax.   It's the tax code;  that does impose these things;  which the tax code itself is 100% not a federal law or statute.  It's a code of rules 100% in its own realm written by a private corporation.

... say it uses 150w mining 'n' algo with 'x' settings....  it varies depending.

Algos like skien take a LOT less power than lets say NIST5.   This is one of the reasons last year and the year before skien was soo damned profitable.  Not many people saw it that way.

When you calculate your rig's power requirements; you must be sure that the rig can not exceed those requirements due to a misconfiguration, new algo, etc.  If you build for what the card is capable of, you have one major issue that can't creep up on you.

We had machines that were running the edge of what they could provide power for.   Worked fine for ~ a year.   Then;  all sorts of strange problems started happening;  and still happen.  Adding a second power supply helped immensely, but it still has random GPU dropouts and odd issues every once in a while.

yep;  looks correct.
 Are you just running it as a single algo?  If so;  no need to have a timeout setting.  4 sec will only help spam the port if anything goes wrong;

are you running it in a looped batch to relaunch it if it closes out for some reason?  This is a good practice, example below:

tribus.bat

if it won't launch and you want to read the error;  just add pause between the ccminer call and the goto call like this:
tribuspaused.bat

The pause will make it wait for a key press before starting over.