What BIOS mods did you need to make to get it to work?  Not even getting mine to boot yet with that many.

Why wouldn't this just be someone moving a ton of coins to another address to ensure they get their Bitcoin Cash like everyone else is doing?

After the fork, there are two separate coins.  You can deposit either Bitcoin from a Bitcoin wallet, Bitcoin Cash from a Bitcoin Cash wallet, or you can separately deposit both...no different from depositing Bitcoin and Ethereum.  You only get "free" Bitcoin Cash if you own Bitcoin before the fork and you have the private keys or the 3rd party (e.g. Kraken) that has your keys provides you a way to get access to your Bitcoin Cash.  If your 3rd party is something like Coinbase/GDAX, then they are saying that you won't get your Bitcoin Cash.  Even 3rd parties that announce that you will get access to your Bitcoin Cash could potentially mess up if they haven't thoroughly vetted and tested all processes, so there's risk there.  If you have your private keys, then you control all the risk and you only have to deal with your own mistakes (i.e. not properly securing your keys, etc.) and won't be victim to someone else's.  The sage advice people are giving to make sure you control your private keys assumes that you are capable of properly securing them.  If that's not the case, storing them with a 3rd party wallet or exchange could potentially be safer.

One thing some people often fail to do is separate mining and investing/trading.  They are two completely different activities.  There are various exchanges and services for converting one coin into another.  Maybe people are too confused or lazy to figure out how to convert crypto or think it's too much trouble.  It isn't and the transaction fees are really not that significant compared to what you would lose by mining a currency with a good future, but bad mining profitability.

Choose which coin to mine based on the max value you can convert to fiat currency.  That should be your only consideration.

Then you can convert your mined coins or cash into whichever coins you want to INVEST in based on what you think the future of those coins.

Simple example:

Assumptions:
One day of mining (today)
Ether price today = $200
Zcash price today = $200
Ether price next year = $400
Zcash price next year = $250
Mining Ether rate today = $17/day
Mining Zcash rate today = $20/day

If you "mine and HODL" ETH for 24 hours, you'd get $17 in ETH and it would be worth $34 next year.
If you "mine and HODL" ZCASH for 24 hours, you'd get $20 in ZCASH and it would be worth $25 next year.

HOWEVER, if you MINE ZCASH for 24 hours and convert it to ETH, then you'd end up with $40 next year.

THAT'S what you need to do: MINE what's most profitable NOW and INVEST your mined coins in whichever coin you think will increase the most in the future.

Current mining rates on these two coins may be pretty similar, but they could potentially be VERY different.  In fact, the best coin to invest in may not even be minable at all.

Good luck figuring out which coin will increase the most in the future.  As anyone can tell you, it's just another form of gambling...not much different from the equity, commodity, and fiat currency markets.

$33.50 for brand new MSI Gaming X RX 580 8GB:
http://www.ebay.com/itm/MSI-Radeon-RX-580-DirectX-12-RX-580-GAMING-X-8G-8GB-256-Bit-GDDR5-ETHEREUM-MINER-/322596953925?hash=item4b1c46df45:g:b88AAOSwmoBZbHwD

Looks like an anomaly due partly to some noob ending their auction after 2am with no min bid.  Live and learn. lol

Same item sold for $499 yesterday:
http://www.ebay.com/itm/MSI-Video-Card-RX-580-GAMING-X-8G-8GB-GDDR5-256Bit-PCI-Express-DL-DVI-D-HDMI-/222574172803?epid=781800439&hash=item33d2745a83:g:g70AAOSwjHdZYBU3

Locally, Craig's List doesn't have much except for one guy selling for $400.

Tools (afterburner, precision, gpu-z, etc.) are better in Windows, so if you're not good figuring things out in Linux, it might be better for you to pay for Windows license and go that route.  In general, Linux is more stable, more secure, requires less fewer resources (e.g. RAM) to run well, and (particularly wrt Win10) much better for privacy.

If you're doing BIOS mods on AMD cards, then the tools are also better on Windows, although that doesn't mean you also have to use Win for mining.

First things first.  If you get new cables, make sure they are designed for your specific PSU model.  Even within the same manufacturer, different models can be wired differently and still fit perfectly.  Particularly with SATA cables, but may also be an issue with PCIe cables.  Many people have fried SSDs by using SATA cables interchangeably.

A lot of people use SATA cables to power the risers (if you're using risers), particularly with lower powered cards (like nvidia 1060).

I wouldn't recommend putting more than 2 cards on the motherboard or more than two cards on a power cable unless you like pressing your luck or playing with fire (literally/figuratively).

Haven't heard of people using splitters, but if you're already powering two cards on one cable, I wouldn't push it.

Thanks.  How much system RAM do you have?

I have more m.2 to pcie converters arriving this week along with another 4GB RAM so I'll be able to do more testing very soon.

What issues (if any) did you encounter plugging HDMI monitor into other cards?

Did you disable other things like sound, serial i/o, etc?

Do you have any automated startup routines to undervolt and/or overclock (using nvidia-smi and nvidia-settings)?

To get 8 GPU working in Windows 7, you need to make some changes to the registry to address the code 43 warning.  Problem I had is when I went down to 7 GPUs and back up to 8, the registry hack no longer worked.  I didn't spend too much more time on it before switching to Linux which doesn't have that problem.  Check out the NVIDIA forum post below and try that and it may help.  Supposedly not necessary for Windows 10.

https://devtalk.nvidia.com/default/topic/738070/cuda-setup-and-installation/8x-gpu-gtx-issue-under-windows/post/4208565/#4208565

@car1999

OK, I made a weak try to get 9 GPUs running in Win7 (using 2 M.2 to PCIe converters on ASUS Prime Z270A) and it didn't work, but the method I tried wouldn't have gotten 8 to work, so I'll probably need to revisit a little more to be sure.

I then moved one of the M.2 converters to the MSI Z170A Gaming M5 to get an 8th GPU there, but the BIOS wouldn't recognize anything on either of the M.2 slots.  I suppose the converter could be bad since it didn't work in either system.

Anything special you needed to do to get the M.2 to PCIe converter to work on the Gaming M5?  What BIOS version are you using and what are your BIOS settings for PCI and monitor?

I got 8 cards working in Windows 7.  That was an adventure.  You have to manually update the registry to pull that one off.  Windows 10 doesn't have that issue.  I haven't tried 9 cards in Win7, but I've heard it's not possible.

On the Linux side (XUbuntu 16.04), I've been having some issues with the X server configuration.  I currently have 7 cards running well there and have been focusing on overclocking and undervolting the nvidia cards and I think I've got a good, stable script working now.  I only have two M.2 converters and installed them both in my Win7 rig (was the only working rig I had at the time, so why not?

Here are the next things on my list to test (order subject to change):

• Add 9th card to Win7 rig and see if the rumors are true that 9 won't work.  My registry edits to get 8 cards working make me wonder if those same edits will also get 9 to work.
• Add 8th and 9th card to Linux rig and see if the motherboard supports that.  I recall something about the 2nd M.2 on some motherboards not supporting PCIe and only support storage and/or memory.
• Convert Win7 rig to Linux if it will give the ability to run 9 cards.

FYI...that Win7 rig is on an ASUS Prime Z270A mb (7 PCIe slots and 2 M.2 slots just like the MSI).  The Linux rig has the MSI Z170A Gaming M5 mb.  Obviously, I might have different results with the second M.2 slot on these rigs.  If one or both of these boards allow me utilize both M.2 slots, then I'll order more, but I have what I need to do the appropriate testing.

Thanks!

Sorry for quoting the whole thing, but I wanted to comment on several areas.  The CoolBits option opens up certain settings to be modified.  The 28 refers to activating 3 different bits (16+8+4).  I think they are for voltage mods, clock mods, and something else.  There are also a couple other bits (2 and 1) so you could do CoolBits up to at least 31.  I think those may be for features on older cards.  In any case, they're not important and 28 should be fine.  24 might even work...I forget what that one bit was for, but 28 certainly won't hurt anything.  You can google the NVIDIA CoolBits option to see exactly what each bit does.

nvidia-settings was giving me some grief when I was trying to do some of this before.  I found equivalent settings in nvidia-smi and played with the max power level, but Claymore didn't like what I was doing and had problems.  I think default power was 120W and even lowering that to 110W caused problems.  It could be that nvidia-smi is just outdated and no longer really supported.  I'll take another look at the nvidia-settings command line arguments.  Assuming I get them working correctly on the command line, I'll try to get them integrated into the xorg.conf file and hopefully that will get everything configured automatically at startup.

BTW, on the "GPUMemoryTransferRateOffset[3]" setting, is it correct to assume that the "3" refers to the performance level (ranges 0 to 3) and that this would only have an effect when the card is in that mode?  If so, it SHOULD be in that top level whenever you're mining, but I did notice in the GUI that the cards tended to stay in level 2 for some reason.  I'll play with it and report what I find.

If anyone knows the proper syntax to put this in the xorg.conf, please post.  I believe it's in the DEVICE section (there's one DEVICE section for each GPU) and begins with "Option" and then the setting name followed by the desired value.  I currently have a line for each device that sets CoolBits to 28.  I think it looks like:

"Option"  "CoolBits" "28"

...or something pretty close to that as I recall.

Can you share how you did this with nvidia-settings?  Did you use GUI or command line?  Do you have examples of command line or xorg.conf file entries?  I had limited luck with this in Linux.  I assume you're using latest nvidia drivers or an older version?

What are you using to overclock 10xx cards in Linux and what hash rate are you getting with that?  Are you putting things in xorg.conf that you can share or are you changing in the GUI nvidia-settings?  I changed in the GUI, but it didn't have any effect.  Or are you executing on the command line?

Thanks.  I actually got it working just with steps 1 and 4.  I already had all the others set except for 3 which I didn't do and doesn't seem to be necessary.  I'm using Linux in UEFI mode.  6 cards worked fine.

I didn't have a 7th riser available (getting more tomorrow), so I plugged a 7th card directly into the last PCIe slot and it recognized it, but ran into an issue with X-server.  However, I was able to use a tty terminal and that worked fine and miner recognized all 7 cards.  Using a riser for the 7th card might fix that issue.  If not, I might just need to try plugging monitor into one of the cards.  I tried plugging it into the 7th card, but that didn't work.

I have some M.2 to PCIe converters coming this week, so I'll try adding an 8th and 9th card and see if that works.  In theory, 9 cards should work in Linux, but I'll find out soon if this motherboard has any problems with that many.

Well, I got a Skylake Celeron (G3900) CPU running on this Gaming M5 board and no change.  Still boots to a blank screen whenever I enable 4G/crypto option. No cards installed.  What did people do to get this to work?

Do any of you use Linux mining with the 1060?  If so, what do you use to overclock?

When using powered risers, you have power inputs on the riser as well as on the top of the card.  Has anyone tested how much is drawn through each?  Is one significantly more than the other?  I want to use SATA power cables on the one that pulls less power.  If one pulls a lot less, then I might consider using a SATA power cable for 3+ GPUs, but I don't want to do that if it's going to be too much to handle.  This would likely be on a rig with lower powered cards like 1060s, so it seems like it may be feasible.  Currently I'm using no more than two SATA connections per cable and they don't seem to be getting warm.

Using Xubuntu 16.04 and Claymore 9.5 with a bunch of GeForce GTX 1060 3GB cards (multiple brands), focusing on ETH single mining.  Trying to optimize performance to get some more hashes out of these cards.  There are Windows tools for making the necessary adjustments, but not so much for Linux.  Are there any shell commands or tools or Claymore settings that people have used to increase performance on these cards?

I have same question.  Trying to enhance performance on 1060's in Linux to get a few more MH/s.  Don't have the same tools as Windows, but hoping there are shell commands or Claymore switches that can be used.  Anyone able to provide some guidance or clues on this?