More likely someone using Nicehash, or a GPU farm that didn't bother with individual names for it's workers.

The pattern of work looks more like a large Nicehash buyer though.

As I can't redo most of the wiring in my current location, I deliberately build my rigs to draw appx. 700 watts (which is a hair under 6 amps at the voltage I see at my outlets) then put 2 of them on each 15 amp outlet (but see below).

The weird part is that the WIRING in this place is sized for 20 amps per circuit (except for the one 30 amp 234VAC outlet) and most of the breakers are 20 amp, but almost all of the circuits feed a SINGLE duplex 15-amp outlet - if I could find a NEMA 5-20 duplex outlet that didn't cost an arm and a leg I could upgrade the rigs quite a bit (though PS tend to get a lot more expensive per watt past around the 860 watt range, and are starting to climb at THAT level).

If you're not a trained electrician though, I'd not recommend doing it yourself (I was a union-trained Journeyman and licensed electrician for a while in my youth, and I've kept up on the relevant Code changes).

GTX 1070 and 1070 ti are GDDR 5, NOT GDDR 5x
1080 1080 ti and the Pascal Titans were the only GDDR 5x cards.

They can change to fix their ASIC - but it takes MONTHS once they have access to the new algorithm - redesign the chip, validate, do a tapeout, test, then *if* the tapeout chips are right order production to go into the QUE when space comes open.

I can see Bitmain trying to do this perhaps ONCE, I doubt they'll try after the first time due to the COST.

 

There are only 10 coins or so currently on the Cryptonight algorithm that I can find.
Monero has already stated "changing on March 28" - that removes more than half of the total network hashrate of ALL of these coins combined.
ETN has sorta-announced they'll be changing algorithm - they HAVE to for their "mobile mining" project to be viable - there goes 3/4'ths or a bit more of the REST of the combined total network hashrate.

If Bitmain puts their typical "about 3000" miners into each of their ANNOUNCED X3 batches, that's going to be ballpark 30 TIMES the total current combined hashrate for all the remaining Cryptonote coins arriving in the next few months.

It's going to be a worse bloodbath than SIA or X11 by far if so.

As of the last figures I saw, XMR was pushing sorta close to 1 Gigahash, ETN 300some Megahash (close to 400), all the rest COMBINED were less than 100 Mhash and at least one of them was less than ONE Megahash.


AMD card prices have already DROPPED a bit (vega in particular) and availability has improved a LOT since January - but ETH is still profitable enough that some new rigs are getting added, just at a much slower pace the last couple weeks.
ETH seems to be the ONLY major coin that has lost less than 50% of it's "peak value" at this point.

Final point - most AMD cards are mining ETH. NOT XMR or other CN coins.
Vega is the only major exception since it has so much higher performance on CN than on anything else.

At MSRP on each, I'd go 1070 ti - but it's a very close choice, you won't go wrong either way.

Don't get a 1070 ti to mine ETH with, get a 1070 - even lower cost for the SAME hashrate.

I was working while mining - and reinvested every penny I could spare after bills into more mining gear.
The job I worked was seasonal - I'd have good income from about the end of January to late June - and 2 years ago I realized that if I just saved the last couple months worth of work income it would give me enough to move to the "land of low cost electric" and that my mining income was enough to live on with some left over for more expansion, while DROPPING my overall cost or living some.

Technically though, I didn't have "20 miners" at the time, more like 12 - but I also don't live fancy which keeps my cost of living down.

El Salvador - TROPICAL climate.
Wouldn't drop the actual miner consumption much - but might drop the AIR CONDITIONING requirement a ton - which is what I think they were claiming.

The problem with that concept is that there is NO OVERCAPACITY in Bitcoin mining gear - total network hashrate STILL has not caught up with the price rise from the lows back when it was pushing $200 AND WAS PROFITABLE AT THE TIME if you had low-enough cost electric.

They're still selling every miner they can get out the door, as they can't get enough CHIPS due to lack of FOUNDRY CAPACITY available.

I suspect that's the primary reason eBang moved to Samsung (and it's why Nvidia put the 1050/1050ti on Samsung) - TSMC and GF can't keep up with demand while Samsung had SOME slack (which has largely gotten eaten now, though not entirely).

Terminus shipping = finally get to put some serious time in on 14/16nm based project?

High-backpressure-tolerant many-blade fan designs are notorious for their high-pitched "whine" when running at high RPMs.

It's PROBABLY normal.

I don't - if they weren't sensible enough to perform due diligence, THEIR mistake.

It was mentioned in a Q&A video linked from their facebook page, but details were pretty ambiguous.
They HAVE to change the algorithm though for their "mobile mining" model to survive at all.

Not gone yet if you don't pay GOUGE pricing and don't have high electric cost.
The days of 2-3 MONTH ROI however appear to be gone, barring another major price JUMP like a year ago or the introduction of a new coin like ZEC that gets crazy-high priced in it's first week or two.

I think some clarification is needed - 10 machines with 1 GPU each or 10 machines with 19 GPUs each would be a HUGE difference.

My ongoing project of "convert my old Folding rigs to riser rigs" has REDUCED the number of "computers" I have actively mining, but I have more GPUs mining today than ever because I have more GPUs PER computer.

X15 with one algorithm added, not that big of a deal for an ASIC.
It would get a LITTLE more complicated with the "order switch" stuff but not all THAT much so.

Not even close to being any more ASIC resistant than X15 is (note that the ORIGINAL Baikal did X15).

3 GB is a lot more RAM than any FPLA has.
Might still be able to get away with 2GB on some of the "spinoff" coins though.

Monero has announced fork on March 28, and several of the common miners for CN (stak-xmr and xmrig to my specific knowlage) have already updated their code for the new algorithm.

Electronium (ETN) also announced they would be forking, but I've not seen a specific timeframe for them yet.

Barnum is outdated.
It's not a fool born every minute any more, it's more like one born every SECOND.

There is also the question of how BIG batch 1 was. If it was only a few hundred, instead of Bitmain's tendancy to release batches of a few thousand, it wouldn't TAKE all that many folks that DON'T know about the Monero fork.

XMR and ETN have already announced they are doing so - that's 90% (give or take a percent) of all Cryptonight total network hashrate right there.

Less than 100 Mhash total between ALL THE REST of the CN coins combined....