If you are going to be running powered USB risers than you do not need to use the extra power connectors to the motherboard. Those extra connectors are for cases when users are going to use non-powered ribbon type PCIe extension cables to run the GPUs off of, so then yes the power will need to come from the motherboard. Using the powered USB risers negates this need as the GPU draws the power instead from the riser board.

Even if you use the one full size slot directly to power one GPU you are ok without using them as there is plenty of power from the normal power connectors for one GPU and the motherboard/CPU components. With all that said, it also does not hurt to use the extra connectors if you have the cables to spare. I run several of these boards and all the GPUS are using powered USB risers so I do not bother with the on-board 4-pin molex connectors and everything has been working fine for the past 12+ months.

Mining may continue to be "profitable" but not by the same metric people have been using for the past 12-18 months.

Times have been good for anyone getting into mining for basically the past two years as rising coin prices have more than offset the accompanied increase in hash-rate and thus difficulty. Now, the good times in mining are finally coming to an end and we are going to be looking at digging in for at least 24-36 months to see any reasonable profitability return.

This is especially true for those looking to buy new mining hardware right now. If you have had your rigs for awhile and you already earned your investments back, you can weather the upcoming dry months a lot better than new investors, as your only consideration is if your mining profits is greater than your electrical costs. Those who need to earn an ROI on top of operating expenses will obviously have a tough time of it, especially as computer hardware is still selling at 200-300% above its MSRP.

To see what is coming you can simply pull up a BTC chart from approximately January 2014 to January 2017. It took BTC roughly 3 years to recover from its November 2013 high of over $1200 and I expect the market to play out similar to that time period. BTC, ETH, ZEC, BCASH and the others will thrash around for another 6 months, just as it took time for BTC to bottom out in the fall/winter of 2014. If this pattern repeats we are looking at 2021 before everything recovers to recent ATH levels.

So if you can tough it out for another 3 years, pay utilities (electricity) out of your own pocket and keep mining and holding, you may indeed make some hefty profits. However, if you are barely scraping by, have bills to pay, have high priced electricity, or the worst-case scenario have recent credit card hardware purchases to pay back, I think you may be in for a tough go of it.



I modified a chart to show one possibility in relation to BTC, although I think it can apply to the whole cryptospace in general. In red I labeled possible similarities to the 2014-2017 time period to try and help myself put one possible scenario into perspective. If the market does indeed play out this way, we will have a long wait before mining truly returns.

ASICs could be around the corner, as if there is profit to be made someone will go after it, and Bitmain the rumored manufacturer certainly has the know how and experience.

Anyway, one thing I have been wondering is if Nvidia and/or AMD aren't starting to run mining farms of their own and helping to run-up both the retail price of GPUs  (less supply as they are using more of them internally) and also driving the hashrate (thus difficulty levels) way up resulting in less income for everyone else. If you think about it this is what Bitmain was accused of doing earlier on with BTC then scrypt coins, so it is certainly possible.

If this were to be the case, then even without the arrival of the alleged Ethash ASICs, home mining would quickly come to an end as any new chip technologies and/or driver optimizations for mining would be kept confidential by the large GPU manufactures.

I agree that server PSUs are the way to go, both for the value and for their longevity. Server PSUs were actually designed and built for this kind of 24/7 operation, whereas some of the lesser consumer PSUs assume a typical low stress 8 hour day usage pattern. As a bonus, at 240V operation, server PSUs are often just as efficient, but at a small fraction of the cost, as the Platinum or Titanium consumer PSUs

While both types obviously work for mining, I think server PSUs are the best value because of the above reasons. Also, the breakout board can cost as much as the used server PSU, so oftentimes new miners are put off by this. Just realize that if a used PSU dies, all your wiring and breakout boards are probably fine and you can simply swap out the failed PSU with another $20 used one versus having to RMA the consumer version you bought new as well as the re-wiring of everything.

At the current Ethereum difficulty levels, the coin price could go as low as $150 and still be at break even for electricity for most people. This of course works well for those who have paid for rigs, but those who just bought in the past 60 days they can say good bye to any quick ROI forecasts for the time being.

So I think the answer would depend on when you bought the majority of your mining gear, as people with paid for gear can probably keep mining as long as the daily profits exceed their electrical costs. Those with more recent purchases, especially considering they probably paid quite a premium for their GPUs may want to consider selling off while they can still get a decent price. If the coin price continues to drop and stays low for a month or more, I feel people will then start looking to get out, again with the more recent entrants the first to go.

The way Ethereum, Bitcoin, and other cryptos are plunging in value, you may not even need to wait that long. Already mining profitability is down to about $1.50 day for a RX580, meaning about a year and a half payback at current GPU and coin prices. If Ethereum drops under $500 for any length of time, i think recent mining investors will start unloading their overpriced gear while they can get any value out of it.

People who manage to buy their gear last year when prices were close to MSRP are doing ok, but any people who bought gear more recently at 150% + premiums will be hurting.

Basically, I think now is the time to wait to get back into mining and not be trying to chase down some overpriced GPU's that may or may not earn back their cost.




You already said you do not like Nvidia, but their GTX1060 cards are very power efficient if that is your main concern. I have several and they run cool and quiet and you can run a rig with quite a few of them with a standard 1000 watt PSU.

Looks like when they said "The exact date of reimbursement for old balances will be announced by January 31, 2018" they actually meant on January 31, 2018, assuming they don't push the announcement off even further.

Even if everything they told us thus far is on the up and up, you would think they would have a plan in place by now and details they could release to the public. All this secrecy and shady back-room dealing doesn't really lend confidence to everyone who they still owe BTC to.

Also, did they ever release the name of the "appropriate authorities" they reported this supposed crime to? Myself, I still think it smells of an inside job, but I guess we will find out tomorrow (hopefully).

No, I do not think you are over thinking it at all. We deserve to know what happened and right now anyone's guess is as good as any other.

Here is one thought, it could have been an inside job of one or more people who were preparing to close up and exit scam everyone. Remember this happened right as BTC was crossing the $10k barrier. In a short amount of time it reached close to $20,000 when the stolen BTC was moving around.

Who is to say they didn't sell a lot of that BTC between $15,000 and $19,000. Now that the price has corrected down to $10,000 again, they could buy back all the stolen BTC they sold, keep the $5-9k per BTC profit for themselves and sometime in the future mysteriously be able to return all the missing BTC to the users.

Why not keep it all might you ask? Think about it, it would be far less likely for users to press ahead with lawsuits as well as any possible criminal charges if users are made whole again.

I think a lot of people are expecting something to be announced. I too have a small amount locked up there and I also will not show my face there again unless they reimburse us. I am surprised people went back to mining with them already before they even announcing a payback plan.

Myself I will stay away until they repay us, and even then would probably be cautious for about 6 months after. I will say the instant and free Coinbase transfer was a good start on their part on preventing further issues going forward, but it should be automatic once a minimal threshold is reached and not a manual process.

Why?

If you wait 6 months your 1 Bitcoin will probably be worth $20,000 again or even more. Why sell your Bitcoin when it is down 40% off of its highs and furthermore invest this loss into high priced mining gear?

Most PC mining components are going for nearly 2x their MSRP right now, so you will further compound your losses by converting your 1 BTC into overpriced mining gear that depreciates in value over time.

Finally, for mining to even work out long-term and for you to obtain your ROI, the prices of crypto's will need to continue to climb. They need to climb just to stay ahead of the new hash-rate that is being added daily, so your mining gains are offset by everyone else competing, not to mention the threat of ASICs, your coins going down in value, or other threats.

All of this just to maybe mine enough coins in the hopes that crypto prices climb enough in next 10-12 months to get your 1 BTC back? Sure you may also have the mining equipment at the end of this time, which is now a year old, but in that same year your 1 BTC might double in value or more, while you are just basically back to break-even of your original investment, only in USD value. Since we are assuming the BTC price doubled, you would still be playing catch up in crypto terms as you would only have earned 0.5 BTC at that point, so you would actually still be down 50% in BTC.

Basically, entering into mining right now is not a good idea. You are about a month or so too late, as the time to get in was when your 1 BTC was worth nearly $20,000 and the prices of mining components were maybe only 25% inflated at that time. Now that BTC has risen and fell back to $12,000, and the mining components have doubled in value versus their previous (modest?) 25% premium, you will be getting less value for your BTC no matter how you look at it.

I have used this strategy myself numerous times, but it is always more like gambling when the coins are not featured on the profitability sites. All coin mining is somewhat a gamble anyway, but these new coins can go either way.

So while you can sometimes find a new coin and get on-board before the public and catch a rising star early by mining lots of coins, you can also end up with wallets full of dead coins. For the most part, once a coin gets delisted from an exchange its effective life is pretty much over, especially if it was only listed on one exchange to begin with. So keep this in mind when selecting new coins to mine, especially before they have even hit one exchange.

I can also recommend the Ubiquiti EdgeRouter as it is a pretty powerful commercial grade router at a SOHO price level. Much better than anything you will get from Linksys, Netgear and the like, especially with that many rigs.

I run a number of rigs myself and also think just statically assigning them an IP address is the best option. My internal network IP scheme I use a /23 block (512 addresses) and keep the lower half of the block for my non-mining devices, and the upper half for mining devices. This way I could name my miner according to their IP address. So my rig named miner01 would get the address 192.168.1.1, miner02 would use 192.168.1.2, and so on. So between setting them up as static addresses not needing DHCP and having a simple naming and IP addressing scheme, this works out well for managing larger (50+) rig type environments.

By segmenting your network this way and with a capable router such as the Ubiquiti, you can even go a step further and apply strict firewall rules on your mining segment to further protect your network. Since you miners only need to connect to a handful of Internet based servers, a handful of firewall rules will help keep any unwanted traffic from going in/out to your rigs. I would recommend applying some rules to your normal LAN traffic as well, but that is getting a bit outside the topic.

If you took the time to actually read his post he didn't use the Kill-a-watt at 240V, he only used it at 120V.

He took the 240V readings off his 240V Tripplite PDU. This is why he was inquiring about the accuracy of them (kill-a-watt) as he assumes the Tripplite PDU reading is fairly accurate.

This is really getting crazy. This would be like a car manufacturer refusing to honor the warranty as their ideal of "proper" usage is only for little old ladies who only drive a few miles a week, such as 2 miles to church on Sunday, 1 mile to the grocery stone on Monday, 5 miles to the Dr. on Wednesday and 2 miles for bingo on Friday. So you show up at the shop and tell them you drove your brand new car to your relatives house 2 states away and back last week and now your engine is knocking they claim you used in in an application that they deem is inappropriate and you are now on you own to fix it.

As I stated before, they gladly are raking in the increased profits, AMD and Nvidia's stock are both up as well as many of their downstream partners such as PNY, but they are refusing to service the very people bringing in the profits. As someone already stated, if you do need warranty work the last thing I would tell them is you use the card for mining. Also don't try and RMA a bunch of cards at a time.

I also do not see how they can consider mining as an inappropriate use of GPUs. Every retail box you look at has OpenCL or Cuda support listed right on it, and mining is a simple application utilizing one of those standards. With the increase in their profits due to mining they should be putting a few more dollars into the production of their cards and not worrying about reducing the warranties.

This is kind of like biting the hand that feeds them as they love the profits coming in from mining interests but are mad about the uptick in warranty requests. Most failures are due to cost savings from using slightly less capable PCB components, usually capacitors or voltage regulators and such, and spending a couple of dollars extra during manufacturing would all but eliminate most of these problems.

Also ensuring adequate application of thermal compound between the heatsinks and heat producing components would help, but the point is a few tweaks and spec changes would eliminate most of these warranty problems from the start. The problem is they got so used to shipping mediocre products (as even intense gaming rarely comes close to the demands of mining) for so long that now when people actually find a way to utilize the cards power in its entirety, they start blaming misuse instead of addressing the root cause. Once again, their attitude is even more ironic as it is the very "misuse" they are complaining about that is bringing in the windfall profits they are now enjoying.

When this does blow over, with attitudes like this they will not only have alienated miners, but also the gamers they claim they are so concerned about.

I have a couple of kill-a-watt meters and have tested them against a decent Fluke multi-meter and I would say the readings are usually within 5% of each other and the multi-meter. This would mean up to 50 watts discrepancy at 1000 watts, or 45 watts using your 900 watt reading.

So I would agree that there is something else going with your 180 watt difference besides the accuracy of the kill-a-watts. As you said, the switch to 240 may explain 30-40 watts, but not the entire 180 watt difference you are measuring. Maybe the PDU is displaying its reading in output wattage where the kill-a-watt would be monitoring the input wattage. So combining the increased 240 v efficiency with not counting the PSU's inefficiency (probably 8-10%) would get you closer to explaining the difference in the measured values.

The problem with these ROI predictions is expecting this pace of price increases (which is offsetting the difficulty rise) to continue for the next 10-12 months. Maybe it will, maybe it will not.

Consider that 12 months ago in January 2017 the price for one Ethereum was around $10. The past year it has risen more than 100X to $1,000+ in price.

This is why the ROIs have worked out for miners, although I would venture to guess that had someone invested $1,600 back then for a standard 6 GPU rig instead into Ethereum directly they would have come out way ahead buying 160 Ethereum coins than they earned via mining.

So getting back to the topic, for your 10  month ROI calculation to work, you are counting on the price of Ethereum to continue to rise. A 10x rise would mean Ethereum would need to hit $10,000 where another 100x rise means $100,000. So if you do believe the price is going to continue to rise to keep offsetting the ever increasing difficulty, you would still be better off investing in the coin directly rather than buying an overpriced GPU.

BTW, I just used Ethereum as an example as you would probably mine a Equihash coin such as Zcash for a 1080, but the premise is the same as all the coins have acted similarly.

This!

You have one and possibly two very real fire hazards going on until you correct this. As has been pointed out the Kill-a-watt is only rated at 15 amps at 120V, and even then that is for a maximum snapshot reading and not meant for continuous 24/7 operation with that much current running through it.

The other potential hazard as Elder III poitned out is if you are running this through a 15 amp circuit. You can check the breaker panel to see if the outlet you are using is connected via a 15 or 20 amp breaker. In the US all circuit ratings are required to be derated to 80% of maximum capacity for continuous operation, which mining most definitely qualifies.

I had this issue before too and the "solution" is to remove the hot air that naturally rises in your room. While you do need some intakes, trying to cool this intake air is not as important as just simply removing the heat.

The solution I found that worked for me, both in wintertime and during 100 degree plus summer days is to move the exhaust fans as high as you can. Going through the roof would be ideal, but at a minimum the height they need to be at is above the height of your highest rig.

It may seem strange, but even 100 degree F (38C) intake air will cool your GPUs as long as you have adequate exhaust ventilation at or above the maximum height of your rigs. It is best for the air intakes to be positioned lower toward the ground, but even if they are higher the main factor was keeping the exhaust fans high.

You will also need to calculate the CFM of air movement that you will need. I only run one out of three exhaust fans I have during the winter, but each of my fans is 20" and can move up to 4000 CFM on high. I went with the shuttered wall mount fans and had to frame in the openings, but ever since I put them in all my heat issue went away as my rigs will run around 60-70 degrees C even on 100 degree F (38C) days. In the winter they run a mild 30C.

Here are similar fans to what I bought, not the exact same as I bought mine from a local building supply company but very similar. http://www.airdistributor.net/p-414-ves-environmental-solutions-brand-3-speed-shutter-mount-direct-drive-agriculturalindustrial-wall-exhaust-fan-cfm-range-880-4874-sizes-12-thru-24.aspx

I think you are on the right track with the inversion idea, however you would need to have a real time unit doing it as it would be impossible to have a recording keep up with the constant changes in sound over time as factors such as; air temperature, pressure, humidity, the fans mechanical aging, how many people are in the room, rearranging of furniture, adding removing rigs, and on and on would influence the sound being produced at any given moment.

I used to play around with electronics when I was younger and I recall it wasn't that hard to make a audio inverter. There are some professional speakers that will also invert a signal, but perhaps you could head to your local guitar or music store and see if they have something ready to go to save you a bit of time and effort. However, even with this setup it will depend on where you are in the room compared to your miner and speaker placements for the maximum damping to have effect, as elsewhere in the room or perhaps further out in the house, the sound waves will reinforce each other thus amplifying the effect.

In the end, while it may work, it may just be easier to try and relocate the miners to a room you are not using or to a removed garage or shed if possible, or as you suggested but didn't want to explore a sound isolation box.