Zcash Mining equipment

I have $20,000 to invest and would like to know what the best equipment to mine equihash would be for my dollar. I want to be at 20,000 hash or higher. What would you guys recommend? Is there a better currency?

Spend the $20K on Zcash coins and not mine. We're in the middle of a significant price correction.

4 Likes

I understand but want to mine as well. The new types of algorithms are messing with my head they seem to change all the time making equipment that was profitable useless.

At the moment, if you are paying for electricity, its not worthy to buy a mining machine.

Since you have awesome budget, make a rig of 1080Ti GPU to mine Zcash.

I was looking at 8 of those per rig for 5840 sols per sec per rig

https://www.zcashmining.tech/rig-builder/

With that said, I have two (2) brand new (running about 5 days now) eight (8) GPU 1080ti rigs that do between 5800-6300 h/s for 20k. Course I'm in the US on the east coast.. :sunny:
Warranty for everything is transferable

What does your profit per day look like?

You running those on 30 Amp lines? I stopped a 7 1080 Ti's since that is a 78% load on a 20 amp line. The 8th GPU broke the bank at an 88% load (not safe for a 20 amp line). The 30 amp lines are a pain, as you have to find adapters to plug in the PSU's, and there is the issue of hard remote shutdown only rated for 15 amps unless you want to go industrial. Works on the 20 amp lines since I just connect it to the master PSU and it will shutdown both. Not ideal but better that trying to synchronize two remote switches.

You will need a case. It's on sale :slight_smile:. Amazon: BlockRigMiner. Sorry if I'm not supposed to post.

1 Like

Just on those two rigs it is about $56.00 a day as of 07/11/17, 4 months ago is was about half of that, and 4 weeks ago it was twice that

@ZC93 z93 Both PSU's on each rig has a dedicated 30 amp breaker

@Zrig if I sold it I would sell it in tact, with rig, ready to plug and play

Get 6 Bitmain Antminer L3+ LTC Miners and mine LTC. You’ll burn five grand in power the first year and pay fifteen grand for the miners and PSU’s (dunno the cost of shipping).

If you are serious about mining ZEC, then you’ll need to build your own rigs if you have the ability. My calculations tell me that you’d need eleven rigs, each with 6 GTX1060s. You can get them for 250 each:

https://www.newegg.com/Product/Product.aspx?Item=N82E16814137037&cm_re=gtx1060--14-137-037--Product

I believe they get about 300S/s (H/s, approx), IIRC. Your base rigs, the motherboard, PSU’s, HDD and RAM will run about 300 each. That uses all of your money and hits your target of 20000H/s by giving you 19800H/s in power.

I’d estimate you’d pull approx 8.5Kw in power. Expect another six hundred in power each month. You’ll need to house them, power them, secure them and vent the abundance of hot air they’re going to generate. You’ll likely need four, 20Amp lines for power. Build your “Woody” cases and don’t piss your money away on anything that doesn’t get you hash power. You will need misc cabling, splitters, risers (there’s 350 bucks), dual PSU connectors, etc.

Here are 4 of my 6 Woody miners. I have a few cards out of service waiting on new PSU’s, but the whole enchilada is good for about 5kH/s.

Note that the mining calculator tells me to expect about 195 Zec annually. If you do that AND the price rebounds to the high last month, that’s about 40 grand USD. You need to consider the expected difficulty and all that other business stuff. Good luck.

1 Like

I figure you'll make close to that, or even more, if you just buy coin and hold it.

Four, 7 GPU GTX 1080 Ti rigs with a price of about $5K each will get you your 20KSol/s rate. However, At ZEC current price and difficulty, that has a 265 day ROI (assuming difficulty does not go up and that never happens), and comes with a $600/month power bill at 0.11/kw/h, draws 7500 Watts and 63 amps. That is almost 9 months before you make a dime. Then if its summer time for you have the added cost of cooling.

Just buy the ZEC instead, its cheap now, you will make a bundle. At least wait awhile to see what the price is going to do as it could change significantly in a month. But then if you had just bought ZEC, and it goes up, you would be in great shape

1 Like

I see it that at the end of the year term, the equipment can be sold to recover some of the money spent on it, something you can't do with a buy and hold plan, which is also good, especially now since there's bleeding in the streets :frowning:

1 Like

If you cannot figure this out without needing to be spoon fed all of this, then buy the coins. If you are willing to dive in and figure it out, then build the computers. But stop asking for everyone to feed it to you. Building your own computers will likely be more profitable, but it will require WORK. Work to figure out what to buy and then to figure out everything on the software side. So be prepared for that. If you are looking for someone to do it for you, prepare to be taken to the cleaners.

1 Like

Okay smart ass. Lets be very clear about your personal dumb assery. Even full on programmers have problems understanding equihash. I have already spent money only to find out that mush equipment cant do equihash and that hashes are not relevant equipment is. It's rocket science. Don't be an asshole if you can't help/. Here is why:

Part I, the overall mining scheme since megacrypto has asked for it.

To mine you get a “work” which is first part of final block, consisting of version (4 bytes), prevhash (32), merkle root (32), reserved (32), time (4), bits (4) and nonce (32) total 140 bytes. When pool mining, you will get the first part of nonce from the pool, and only the second part you will choose by yourself. Otherwise if you sole mine then all nonce is yours.

Then you make 2**20 = 1048576 hashes, each is 50 bytes of BLAKE2b function of 140 bytes of “work” concatenated with 4 bytes of hash index. Then you split each hash in the middle to two strings of equal size, getting 2**21 = 2097152 strings of 25 bytes (200 bits) each. This requires at least 50 megabytes of memory, effectively kicking out modern FPGA and ASIC approaches to high performing mining for other coins.

Note, number 21 is equal to N / (K + 1) + 1 and number 200 is equal to N, where N = 200 and K = 9 are Equihash parameters chosen for Zcash.

Next you solve Wagner's algorithm. What is it? You must peek 2**K = 512 different strings from those 2097152, such as binary XOR of them is zero. Moreover, some other properties of your selection must be assured. First, on each split by two of your indexes the left ones must be ordered, and second, on each split by two of your indexes resulting XOR of strings must have 20*height zero bits from the left. More on algorithm is in the part II.

When you find some solution (typical nonce provide around of two solutions on average), you fill the rest of block with solution length (3 bytes in special encoding always the same value 1344) and solution itself, which is encoded by 21 bits per index, thus 21*512 bits that is exactly 1344 bytes.

There is a target, which means the difficulty barrier. You make sha256 hash of sha256 hash of your total block of 140+3+1344=1487 bytes, and then your compare your double sha256 with the target. Below the target is good, and then you can propagate your block if sole or submit your job results if you do pool mining.

Caveat of bit ordering and start of indexing everywhere.

Part II, the Wagner’s algorithm, which is the core of Zcash mining.

You will make K+1=10 steps, and in the beginning you have 2097152 original strings of 25 bytes each (200 bits). On each step you will be interested in next 20 bits from those 200 bits strings, i.e. bits from 0 to 19 on first step, 0 to 39 on second etc (counting from zero from left). On each step you try to combine pairs of strings from previous step and keep only the pairs that have common interesting bits (in other words, XOR of which is zero in interesting bits). So, on each step you will get growing binary trees of indexes in the original strings: one index on step 1, two indexes on step 2, four indexes on step 3 etc. Note that all indexes in one tree must be unique. At the final step you will have 512 unique indexes combined into desired properties of the algorithm.

There are some optimizations implemented in reference miner and more optimizations in winners of miner contest. Basically it’s about peeking correct pairs. Note that you don’t need to try each index to each, you can first sort them and then try to combine only while interested bits are the same (reference miner), or you can put strings in the buckets indexed by interesting bits (other miners), and keep those buckets compact to fit into processor caches (that’s the key of speed). Also, you can do things in parallel to use multicore CPU and GPU advantages.

who is megacrypto?

.........