Correct the new processor was an I3, and you're right that could have been it as well.  Fortunately the shotgun resolution approach on the mother board/CPU/memory seemed to do the trick!

The rig has actually been doing well.  It's been running for about 2 days now with 0 errors.

One of the regular 1070ti's is running a but higher than the other card [54 degrees vs 44 degrees].  I thought it kind of odd to have that much of a difference.  Could be card placement (54 degree card is close by the processor fan).

Figured that may be a concerning point to keep in mind if I do any overclocking. 

Thanks everyone for the comments/help, I appreciate it!

Going back to see if the $12.00 payout I had coming is still pending

New motherboard/processor and 8 GB Ram installed.The updated rig is doing A LOT better, it has been mining with all three cards for 4 hours or so.

I know 4 hours isn't a huge milestone marker, but when your crashing every 5 mins... that's a win!!!  We'll see how it runs through the rest of the day.

In case anyone is searching forums/google for the MSI Pro Series Intel Z270 DDR4 HDMI USB 3 SLI ATX Motherboard (Z270 SLI PLUS)

I was following a rig guide on the following site. 
http://www.coinminingrigs.com/how-to-build-a-6-gpu-mining-rig/


It was relatively recent, and had comments from other users who tried the build so I thought I would follow that as closely as I could.

For me it seems the issue was the motherboard (at least thus far).  There are some other caveats though to consider
1 - The board was shipped to me in the motherboard box itself, no additional packing (ordered from Amazon).  Could that of damaged the board, maybe.  Personally I would have like to have seen it shipped in another box.
2 - BBT did do a rig build using this board, so perhaps it was just user error
3 - I varied some components from the builds (power supply selection, GPU selection)

- Your experience may vary, but just wanted to put my experience out there in case others are trying to use that board.

Well as I was at an impasse as to making any further changes to the rig.  I just restarted it once, after a quick mining failure.  Then it went through a windows/NVidia driver #$%^ storm    ....... and now I am looking into SMOS again.

I was able to get SMOS up and running.  The third card was dragging in hash rate, then would crash.  I replaced the riser, the same thing happened.  I then just ran two cards and the hash rate was fine.  I decided to move the 3 card down a slot, and unfortunately it kept doing the same.

I'm starting to think it might be the board as it is showing symptoms similar to what others have seen.  Trying a new motherboard/processor tomorrow.

SMOS is pretty slick, certainly a bit more too learn, but doesn't look too daunting.  Fortunately I was able to get up to a testing point to see that there still may be a hardware issue that was lurking.

Switching over to a Gigabyte GA-Z270-HD3 and an i3 processor tomorrow.

Tried SMOS last night.  To far in the rabbit hole as it is to really start diving into that as well.

I upgraded the Ram from 4GB to 8GB.  Debating on if I should put the other 4GB in a also (kind of old fashioned with installing matching pairs though)
I changed the bus speed back down to 32, and the PCI type to Gen 1

Still messing with Ethereum, need to get other wallets/pools setup and figured out.

All 3 cards are coming up with Windows boot regularly now.

The system was able to mine with the BBT miner set (Claymore Dual on nanopool) for 35 mins.

Moving in the right direction, but running out things I know to trouble shoot.



I've read this motherboard tends to have issues for whatever reason, but I would guess there's probably +/-'s for most boards.  I went ahead and picked up a Gigabyte Z270-HD3 mother board and i3 processor incase I want to scrap it and start over, but I feel it's coming along....just need a few more kinks worked out.

Thanks, think I am going to jump straight to trying the smos.  I don't understand the strings I need to enter to well but will try and take a stab at it.  Hopefully the issues were just windows related and this can fire up quickly.

If not I can always throw the SSD be back in and pick up where I left off.

I tested each card and riser in their proposed slot individually.  Each ran by itself for 1 hour without any failures.  So kind of ruling out any hardware failure type issues.

All 3 cards are in the system, and load up quickly and available.  I am getting video from 1070ti in the primary x16 PCIe slot

The system now locks up when mining around the 5-6 min mark.  I tried the VertCoin miner real quick, and while it ran a bit longer the system froze eventually as well.

Not that I could see in the BIOS settings

Another day of testing.

I decided to change the power for the risers to SATA (2 risers per), that seems to have helped upon boot.  Not sure if the 8 to 6 pin plugging into the VGA ports of the power supply was messing things up, however all 3 cards come up faster/properly at boot.

I've been messing around with the Gen1/Gen 2 option, as well as the PCI bus speed.

I increased the page file.

I ran the system with 1 card and it ran fine for an hour, currently trying out another card/riser solo as well to test it.  Currently making sure there are no card/riser issues currently just to completely rule that out.

I updated the drivers to the latest (I think I was a version back as I installed them from a USB drive).

The issue just seems to be when I add cards to the system.  When I start mining with 2 cards, the system hard freezes about 5 mins in or so.

Trying to get 2 cards to work first then will try to add the third. 

It really does seem to be a driver issue / software issue, but can't seem to pin point it.  Still just working on ethereum at the moment.  Need to get another pool up to test.

I have not tried another miner yet, I will add that to the troubleshooting list.

I did try the cards individually/risers individually, adding them in one at a time.  The issue starts on anything higher than 1 card.

I'll leave the system up for a while just to see if there are any issues.  So far the only time it has failed is when it is mining.

I'll try turning on the integrated graphics as well.



Thanks for the page file suggestion, I'll give that a shot.

You think he's fine..... and then he's shoving crayons up his nose.... pretty much the way my first rig build is acting.

My first post to the forums, been searching through the forums over the past few weeks, reviewing YouTube for build ideas.

I decided upon the following gear:

EVGA SuperNOVA 1600 T2
Kingston ValueRAM 4GB 2400MHz DDR4 Desktop Memory
MSI Pro Series Intel Z270 DDR4 HDMI USB 3 SLI ATX Motherboard (Z270 SLI PLUS)
Gigabyte GeForce GTX 1070 Ti GAMING 8G (GV-N107TGAMING-8GD) x 2
Gigabyte GeForce GTX 1070 Ti GAMING 8G (GV-N107TGAMING-8GD)  OC Edition x1
Open Air Case for Crypto Currency Mining Rig Frame
Leboo 6-Pack PCIE Riser VER 007 PCI Riser 6 PIN 1x to x16 Powered Riser Adapter Card w/ 60cm USB 3.0 Extension Cable & PCI-E to SATA Power Cable - GPU Riser Adapter
Samsung 850 EVO - 120GB - 2.5-Inch SATA III Internal SSD (MZ-75E120B/AM)
Celeron G3900 2.90Ghz 2M LGA1151 2C/2T
Windows 10 64 bit

So I started out with the one 1070 Ti OC edition, testing it in an old PC rig, getting into mining a bit, Nice Hash, Awesome Miner, BBT's miner pack, and was running the one card fine in those tools.

For the power I understand that it is overkill.  I chose it as it would be a good power supply to use should I upgrade in the future or build another PC, and it had quite a few more connections and I wanted to stay away from SATA powering the risers.  The video cards take a single 8 pin, I have those connecting to the VGA ports on the PSU.  The risers are being power on a 6 pin splitter plugged into the VGA port on the PSU.  As I bought a 6 pack of risers, I swapped them around and out a bit, and still getting the same type of errors.

It seems to not be liking multiple cards.  I can tell when it loads properly.  All three cards will stop their fans prior to Windows loading.  When they don't the cards will show up under the device manager with a Code 43 error.  Most the time it is always the 3rd GPU spot, however sometimes it's GPU's 2 and 3, and occasionally just 2.  I can turn them back on by disabling and enabling them in Device Manager.


Here are some things I have done so far.  I'm still researching what it may be but thought it may be time to see if anyone may have some suggestions.  Unfortunately I'm not ready for Linux yet.  I understand that would be the route to go eventually, but that would be an additional greenfield for me at this time.

BIOS updated to latest version
Enabled Windows 10
4G Decoding enabled
PCI bus updated to 96 Mhz
PCI Sub system Started out with Gen 1, but changed it to Gen 2

No Over-Clocking
Updated to latest Nvidia driver
Currently using Integrated Grapics as default
PCI Simple Communications Controller Driver updated

Page File set to 16000, max 16000
Power set to performance
Set to sleep never

I have not turned off Windows update yet.

So... when the stars align and all three cards are ready to go when Windows loads up, it gets a bit all over the place.  I am using Awesome Miner, mining Ethereum on Nanopool using the Claymore Dual miner (dual mode disabled)

It connects, cards are getting around 25-27 hashrate, temps are all good and then:

- Windows Freezes
- Windows BSOD
-  GPU 1, 0, 2 GpuMiner cu_k1 failed 4 , unspecified launch failure
       (and similar failure warnings, ening in WATCHDOG: GPU Error

So that's where I am at so far.  Any thoughts or suggestions would be appreciated.