Whether or not the laptop heats up enough to cause damages depends on the specific unit. A recent laptop should have no problem running the necessary software to mine with 1 graphics card without placing much strain on the CPU.

Bus speed is generally not an important factor for the purposes of mining. (Note the negligible change, if any, in mining performance when swapping between PCIe x16 v3.0 and PCIe x1 via risers.)

Basic guidelines with multi-PSU rigs:
Use one PSU to power the board's main connection (dual PSU cables shouldn't wire all the connections from both 20/24-pin main connectors; it should only have all the cables for one connector, then the other should only have the green pin and one ground connected). Use this same PSU to power the 4/8-pin CPU connection(s) on the motherboard. Generally this PSU should also be used to power the storage device, but that is not necessarily a requirement. Ensure that the other PSU only deals with graphics cards and risers.

Ensure that you distribute the load such that neither one of the PSUs is over 70-80% of its rating. Despite the "Add2PSU" name, you're not truly getting the effects of a single PSU; appropriate load distribution remains your responsibility.

If you're like me and choose not to use an Add2PSU device or dual PSU cable, ensure that the secondary PSU is powered on before hitting the boot switch to turn on the primary (connected to the motherboard) PSU.

As always, but especially in this case, use quality PSUs. Adding more PSUs in a nonredundant configuration, as is typical, can add more points of failure. For example, if EITHER unit fails, it may result in downtime for the entire rig. (If the primary PSU fails, this is the obvious result. However, if the secondary PSU were to fail, it may or may not be the case that the rig can continue mining with the graphics cards connected to the primary PSU. Oftentimes, a power failure to some of the cards can cause the mining software/drivers/OS to throw an error, downing the entire system.)
Using quality PSUs is especially important when each represents its own possible point of failure for your rig. Look for quality OEMs such as Seasonic, FSP, and Superflower. (HEC, CWT, Sirfa/Sirtec are definitely lower on the totem pole, but better than no-name units. I personally don't use these for mining rigs, where the are under heavy loads 24/7, but I've used their units for home use and they usually do the job.)

First, it's more common to refer to one "rig" as one CPU/mobo/RAM set/etc. and all associated cards hosted on that individual system. I'm going to interpret your 30-60 number as the overall "size" of your theoretical mining operation, regardless of how many rigs are used to host that number of graphics cards.


The "best" distribution is a matter of available hardware (not everyone can buy a 19-slot board and 6x P106 cards right now) and personal preference.

Running more GPUs on each rig may give you a decent advantage in price and operating costs. Keep in mind that riser costs almost always scale exactly with the number of cards you have, and PSU costs do to a similar extent (except for extremes, such as 1-2 cards per rig, since each rig always requires a minimum of 1 PSU), so those will not be the main areas for savings.

However, in costs for the motherboard, memory, CPU, and storage device, you can save. (The motherboard slightly less so, since motherboards that can support more GPUs are generally more expensive.)


On the other hand, having more cards per rig can start to put too many eggs in one basket. If any part of the rig "base" fails, for example, a CPU, then you have more cards down until you can repair it (whether that requires buying a new replacement board, or just finding the time to swap out for an available spare, although if you have spares readily available for all rig base components, we'd have to ask why you don't make some extra rigs, since purchasing spares defeats the point of saving on the initial investment, and not using them only saves a negligible amount of power, compared to what a 30-60 card operation would require).

Metroid's post had a little truth to it. If one card throws an error, it can bring down the mining software/driver/OS/etc. This definitely isn't always the case, but it can happen. (The most common example is a dual-PSU rig, where the secondary PSU (which only runs some of the GPUs) fails and manages to down the entire rig.)


Setup headaches are also something to consider. With a good motherboard and a uniform set of GPUs, I generally haven't had scaling problems between 6-12 GPUs on Ubuntu. In both cases, setting up drivers has been an all/nothing thing, rather than an issue of getting all 12 GPUs recognized. However, others have reported much bigger headaches, and I can definitely imagine the 19-slot Asus board being a pain, especially since it still requires 6x P106 "mining-oriented" cards in order to make full use of the board's potential.

In general, as you add more GPUs per rig, things get harder at different rates. For example, going from 5 to 6 might not be as hard of a jump as getting from 13 to 14. (This also depends on the motherboards you use, setup will generally get more difficult once you try exceeding the number of PCIe slots by throwing in PCIe splitters and/or M.2 adapters, but YMMV all the same.)

How you weight/prioritize the above factors is a matter of personal preference.

If you want to avoid online scammers, sell them locally for cash via Craigslist or something of the sort, meeting only at public places. Seller protection is relatively easy here. (Buyer protection, not so much, since most buyers don't lug around a test system to confirm the cards are working.)

Chances are however, that eBay is fine, if you've been a long-term seller with good feedback. Requesting signature confirmation, sending the buyer a tracking number, and recording a video of yourself putting the card in the box before it's taken by the carrier is probably sufficient for most realistic cases. (Note that your video doesn't verify the card is in working condition, so there's a small chance someone could claim it's defective and send back a different, broken, unit, so ensure that you have proof of the serial number...)

Mining hardware scammers are very creative, so there's not really any guarantees, but I think there are still steps to make transactions reasonably safe over eBay.

Note that the PCIe "multiplier" devices are hit-and-miss. Even if you assume your unit has no defects, those devices cannot guarantee that all of the physically available slots will work. You may get all, some, or even none of the extra slots when using one of those.

Without exclusive access to your private keys, the coins do not really belong to you. Essentially, an online wallet service just gives you an IOU for some number of some type of coin. The hope is that the user may "cash in" the IOU and spend it as they please, making the online wallet effectively functionally equivalent to wallet clients where the user has full control over their private keys. Perhaps the service never breaks from this illusion, but perhaps they will. Either way, the IOU is just that, an IOU that is backed solely by the reputation of the issuer. Such is a risk one takes with anyone that offers a third-party wallet service, be that Mt. Gox, WeEx, BTC-e, NiceHash, Coinbase, etc.

Any wallet provider that claims to fully insure user balances in the event of a major hack or other incident is probably full of BS. These types of disclaimers are fairly standard. To be frank, even if NiceHash didn't make mention of the risk associated with storing BTC with them, that'd be no reason to trust them.

The fifth card does not have to be the same model, or even use the same GPU. You could add an AMD GPU, and it's relatively easy to put together on Windows, but it can be difficult to juggle the drivers on most Linux distros.

The one you can choose will largely be restricted by available power. I think you could get by with another 1060 (count your connectors, too).
Finding a good spot to mount another card is up to you, though.

The PSU-end of the modular connections has NO widely-used standards. Various brands from the same company will often differ in pinouts.

The 6-pin connection on the end of the PSU cannot be assumed to be equivalent to a 6-pin PCIe connector (and almost definitely isn't).

You COULD look up a pinout diagram for your specific unit (or make one with a multimeter), then fashion your own cable together (keep in mind that you can't expect to draw as much current, because the "peripheral" connections will likely not have as many 12V lines as the dedicated 6-pin connections, since PCIe power doesn't require 5V that is required for molex, or the 3.3V line that is sometimes used for SATA [it's sometimes excluded, as it's almost never used]). Otherwise, stick with using the modular lines of 4-pin molex power connectors as-intended, and use 2x molex to 1x 6-pin adapters (don't power too many risers off of the same line of connectors, obviously.)

There's no need to try to start both PSUs at the same time.

The secondary PSU can be forced on first, then the primary PSU can be started normally via the main switch connected to the motherboard.

Nobody can tell you when a specific unit will fail, but there are tips that will generally extend the life of your hardware.

Keep core temps low (subjective, but generally below 80C for 24/7 operation. Although some have stories of running cards at core temps near 90-100C, none of them would truly recommend that practice in general). Do note that you generally have diminishing returns on this (for example, cooling from 60C to 55C probably offers smaller lifetime benefits than 80C down to 75C).

Avoid unnecessary strain on the fans (no need to run fans at 100% just to get below 60C, or whatever). Use cheap box fans to supplement, but avoid causing turbulence with the integrated fans. Also remove dust buildups as-needed (varies based on the air quality of your mining location)

Avoid excessive overclocking. For example, excessive memory overclocks can cause artifacting, which is a symptom of an unstable overclock that may cause permanent damage, and this symptom might not be noticed without having a display connected to each graphics card.

Use a quality PSU. A PSU made by a reputable OEM (particularly Seasonic, FSP, SuperFlower, and some others) generally has a better chance of consistently providing clean power to your cards, rather than flaky no-name PSUs which may have more irregularities.

I have 7700 and 7800 series cards from 2012 that still work fine today (used for mining from mid-2012 to early-2015, then fired up again in 2017.). A good friend of mine still has some 5000 series cards from 2010 that are still operational (but not actually used for mining anymore).

When supply exceeds demand, retailers generally rely on maintaining competitive prices to ensure that other retailers don't secure more sales. In a way, they are somewhat desperate for cards to sell. While, in the current state of the market, it is still true that lower prices are favored by buyers, the need for a retailer to set a price in the ballpark of the MSRP is greatly reduced because they can be confident that enough sales can be secured to deplete their available stock at higher prices, since competition between buyers is much more significant than competition between retailers when demand greatly exceeds supply. (The question becomes "Why should they sell to you?" rather than "Why should you buy from them?")


Keep in mind that production of GPUs and graphics cards was originally designed for a market where the average user has 1-2 desktop PCs, each with 0, 1, maybe 2 discrete graphics cards each (3-4 way SLI/crossfire are uncommon, due to the diminishing returns offered and generally poor support overall). Average users are expected to upgrade their graphics cards no more than once per new generation of GPUs, often much less frequently, and some users purchase prebuilt gaming systems.

Of course, production under that model always maintained a surplus, since you could always purchase a recent graphics card at or below MSRP at almost any time.

Now, increase the popularity of mining:

A typical mining rig has somewhere on the order of 6-12 cards, to balance out the desire for squeezing the most of the purchase of one rig base (motherboard/CPU/storage device/memory/etc.), and avoiding having too many eggs in one basket + setup headaches. Even if just 1% of the knowledgeable PC-building community were to average 3-4 rigs each, demand would increase from the previous model significantly, likely far beyond the aforementioned surplus of cards. (After all, maintaining a ridiculously large surplus of cards would be a huge waste, so there was no reason AMD/Nvidia and their card manufacturing partners would constantly be prepared for huge increases in demand.)


Regardless of what Nvidia might tell its partners about favoring gamers over miners, a seller gets paid the same amount regardless of what the buyer chooses to do with the card (neglecting warranty services). If sellers strategically set prices just below the threshold at which mining appears* to be unlikely, then the inflated price allows the seller to receive the greatest revenue from their limited stock.

*The appearance of the threshold is more important than what it may be in reality, since uninformed buyers may attempt to purchase cards at absurd prices, if they choose to believe that they will achieve a significant return on their investment before their hardware becomes obsolete.

You could get an adapter that would allow you to connect a VGA display to a graphics card.


Alternatively, if your rig runs Windows, switching your rig to use the iGPU for display should be easy; just go into the BIOS, set the rig to default to the integrated graphics, connect to the integrated graphics, and you should be good.

If your rig runs Linux, I'd recommend getting an adapter instead of trying to run drivers for both the integrated graphics and the discrete GPUs. It's possible, but a real pain.

If you haven't setup your second rig yet, use the TV with HDMI until you've finished setting it up, then make the necessary changes to use the iGPU with the other monitor.

What mining software are you using?

If you're certain all individual cards and risers are working, try adding them one-by-one, starting with one connected to the physical PCIe x16 slot closest to the CPU (use this one for video, as well).
Assuming your BIOS is currently at default settings, ensure that you have enabled 4G decoding, and have PCIe link speed to GEN1 or AUTO.

Clear CMOS (either use the appropriate jumper, or remove the battery and replace it after ~10 seconds).

Then attempt to boot with ONE card, directly installed into the physical PCIe x16 slot closest to the CPU. Can you get your system to boot with this configuration?

You could get away using smaller PSUs if you bought 3 PSUs and did 4 cards per section of PCIe slots (perhaps 1x 1000W and 2x 850W PSUs, assuming you keep overclocking to a minimum, and your CPU is reasonable for a dedicated GPU-mining rig). However, this would make expanding your rig in the future more of a pain than if you purchased 2 higher-end PSUs, and left the third section free.


If you're always careful to run the cards below full TDP, then perhaps you can get away with predictions showing 90% use (with 1080s having a 180W TDP, this would be 1080W for 6 cards, placing a 1200W PSU at 90% of its rating*). I wouldn't personally recommend banking on your ability to keep your mining system from hitting the full TDP, but some people do it successfully.
*This did not include the nonnegligible draw from the CPU and other necessary components. 2x 1200W PSUs wouldn't be enough for a 24/7 rig pushing 12x 1080s at 100%.


If you want to play it safe and only use 2x ATX PSUs, try for 2x 1400-1500W. Generally, you want the sum of all TDPs for components connected to each PSU to be somewhere in the ballpark of 70%-80% of the rating, for safe 24/7 operation.

If fans are running at 100%, or close to 100%, it won't last long. Laptop fans are typically much smaller than the fans on desktop graphics cards, so in order to achieve similar airflow, they must spin at higher RPMs. This causes additional wear.

Laptops were not meant to be pegged 24/7. Removing it from the case and adding your own fan setup could make it a better plan, but I doubt it's worth it.

In general, if you have a laptop that's recent enough to offer a reasonable hashrate, and you don't mind gradually ruining it by mining, it's better to sell the laptop and put the money towards a real rig, or put it into coins.

If I recall correctly, the 750W G3 is done by Super Flower, which is one of the top-tier PSU OEMs. For a 1070 and 1080ti (and nothing wacky with your CPU/mobo/RAM/fans/drives), 750W should do the trick, assuming it has all of the connectors you need.

Temperature and power use aren't the only limiting factors when overclocking.

There are also safe limits for core and memory clocks, for every unit. Off the top of my head, excessive memory clocking can lead to artifacting (I've seen some users with Samsung memory get close to +600-650 stable OC, but check with your own hardware), which is a sign that you are pushing your card too hard.

Whether or not is something you should check experimentally, although there are some trends with certain algorithms. For example, with equihash, I haven't noticed gains between +100 and +300 mem on my 1060s and 1080tis, but people have had mixed results.

If it's a closed loop system, it will take years before any loss is noticeable. The vast majority of any evaporated water (or other liquid) would remain in the loop, to be *mostly* condensed at the radiator. Unlike the plated copper, water (or other preferred coolants) has a relatively high specific heat capacity. This allows it to absorb quite a bit of energy vs. the temperature increase that you observe.  In short, making the water reach the necessary temperatures to evaporate a significant amount of it isn't easy.
(Also note that the pressure conditions within a closed loop don't make a great environment for mass evaporation of the liquid.)

Custom cooling loops allow you to throw in your own reservoir(s), so you have the option of adding fluid, should you need any over time. "Closed" loops, by definition, do not allows this.