Topic: [CLOSED] [Group Buy + Group Hosting] 3.5TH/s Avalon6 + Platinum PSUs! - page 2. (Read 7410 times)

Avalon sadly is not Spondoolies or Bitmain when it comes to demand in the used miner market but depending on how successful you're with timing your exit (the 3-month contract renewals might complicate that) you should be able to get 20-30% back after 6 to 9 months.

Unfortunately that part is even less predictable than difficulty/exchange rate/etc. Depends a lot on what other hardware is available at that time.

My example is more like buy btc OR buy hardware go get btc. At least thats what I was trying to do.

you are not putting btc into it, but $$. Don't put in bitcoin that you already have.
these calculations never account for btc price change, and therefore cannot be trusted.
Start with $$, end up with $$ and put in various bitcoin price scenarios.
Besides, where in your calculations is the RESIDUAL value for the miner?
It was ~55% in BTC and 35% in $$ for SP30 after 12 mo and 53% of SP20 in $$ and astonishing 78% in btc after 10.5 mo (even with a high price paid for SP20 before discount to $400-500).

This i my ROI point of view:

This calculation is based on the fact that you have the money to pay for the miner package + hosting for a year, on hand, so that the alternative is to buy coins instead of mining gear for that amount of money.

Calc inputs:
Difficulty increment: 4%   #I seriously doubt it will be below 4% on average.
Pool fee: 2.5%         #Based on Ant PPS. Even if you are mining on 0% fee PPLNS. You can't expect more than 98% luck on average IMO.
Hashrate: 3500         #Based on 3650TH/s -4%. They have stated +/- 5%, lets try to be realistic.      
Hardware price: 1436$ (Ex. hosting/PSU rental)
Hardware price in BTC: 4.373
Watts: 1000
Start date: 11-30
Maintain cost/month: ~58$ #Based on 0.0661$/KWH aka hosting, and 10$/month PSU rental



Seems like you will never earn back the amount of BTC you put into this.
Miner cost + 1 year hosting = 1436 + 696 = 2132$ total invested.
If you buy coins today for that amount of money, this will give you 6.41 BTC.
If you buy gear, you pay 4.37 BTC up front, which you won't mine back. Unless you get cheaper hosting or $/KWH.

PS: There are 3 months of hosting at 0.047$/KW, that is not considered in this calc.
PS: Calc does not take halving into consideration. If you can't ROI without halving, you probably can't with it either

Am I missing something?
Better to buy coins than gear, in terms of ROI?

Disclaimer: You can ofc ROI this in terms of $, if BTC price skyrocket. My point being, it's always better to buy coins.

I just do calculation up to halving, then a brand new calculation that assumes a ~9mos (or whatever it is exactly) lead time for 12.5btc/blocl. Then add the two together.

From what I see though, breakeven is around the point of halving, after which there may be another 10-20% return over an extra 6-8 months.

Biggest concern is that bitcoin price could move faster than difficulty, as I think it's realistic to assume $700/btc or better by 2017.

Not that I'm aware of.

spiccioli

Yes you are correct, sorry I should have said that.  Are there any calculators take halving into effect?

Finsky,

your model doesn't consider, it seems to me, that at the start of next July there will be next halving.

spiccioli

A couple problems with your calculator:

-Mining starting January 6th-  Miners are due to ship on November 20th, and arrive at the hosting facility to be operational by around Nov 27th if enough people get involved. I have it set as an early December start date in case of delays.
-Electricity Cost of $0.065 per KW & 1000W-  Electricity cost (hosting) is already included in the purchase price for the first 3 months.  Better off removing it from purchase price, and using $58.02 as a recurring monthly cost to cover electricity and PSU for the calculator.
-Hash rate and energy consumption- Firmware for 3.65TH/s is meant to be released shortly after miners ship, and the consumption figure of 1kW is based on the initial, higher figure.

Here is what I came up with:



Note, my model also has the lower exchange rate reflected.  If I was to run it at the same exchange of $333 from  yesterday, it would be a 323 day ROI.  Total return was BTC1.2, and higher exchanges would offset cost of hosting by more.

You can if you want, you can also renew the contract. Your choice!
As for ballpark shipping cost: It depends where you live. It would probably be best to use canada post's shipping calculator to get a rough estimate in CAD.

A ball park figure on cost for that to be done would also be nice to know. (again into the math of all this)

Just confirming that we ship the miners from the data center to ourselves after the hosting contract is done?

hmmmm....what I get on tradeblock.com with these new numbers. Using 1-15 numbers ..if we got to over 50 units the savings would be about  50 bucks FYI

Also this is NOT including any pool fees which MAY be included in hosting. If I am incorrect run the same calcs with 2.5% pool fees as others have sugested (or 1% your choice unclear on this myself)




used default 7% diff rise on the calculator


anyway not too rosy on above would not roi in 1 year

Alright everyone! I had the pleasure of speaking with Kevin @ GND this evening to get some details about their facility and power situation.  The way their 3-phase power & PDU's are arranged require that each leg inside each PDU be balanced so as not to cause them problems.  Each PDU can handle 7500W maximum, or 2500W per leg, balanced.  The only way to safely arrange this w/ Avalon6's is for each PSU to power 2x units after the firmware upgrade to 3.65 TH/s & ~1000W DC, which would give ~2100W after PSU losses (6300W per PDU).  The only way we would be able to get 2.5 miners per PSU as originally planned would be to keep them at 3.5 TH/s as opposed to the planned firmware upgrade to 3.65TH/s, which I'm quite confident nobody would want to do.

As for lease rates, I am offering a rate of $20 per PSU package per month, including shipping to the data centre, replacing any faulty units, and covering lost mining profit due to PSU problems beyond 99% uptime. This will work out to $10 per miner per month for the PSU for this Group Buy. Considering the issues revolving around joint ownership of the PSU's, responsibility of who would replace faulty PSU's in the case of a failure, and having to hastily liquidate the hardware at a discount and ship PSU's to the new owners at the end of the hosting term, I think this will make a lot of sense to most of you. If anyone is willing to sign on to a 6-month or longer hosting contract with GND, I would be willing to drop the rates for them to offset shipping there and back. Hosting contract is a fixed 3-month.  We can talk about discounts for those looking to renew w/ GND after their 3 months is up.

If you have any questions about this, please feel free to ask!

Looking forward to it!

Good idea, I've done that.

That's exactly what I've been thinking, although I'm stubborn--and I'm planning on running these miners all the way up until they don't turn a profit anymore. Great North Data is actually saying that they will have even lower rates in a few months, which we will get. So that's a nice thought.

this is quite a good deal, its a shame the hardware cost is just a *smidge* high. If BTC could jump back over $350 I would be interested in 1-2 units

realistically the hardware would stay at GND for a year or more because of the costs to ship it elsewhere, and the fact that its cheap enough electricity that difficulty could pretty much quadruple at $320/BTC and still make a few dollars a month.

Can you please specify "noon" or "midnight" (12:00 PM is surprisingly confusing for a lot of people) and perhaps add UTC times too.

I just need to speak with GND about some of the details for PSU configuring to get a firm price, but I think this could be very beneficial to the GB members. Leasing will eliminate the issues of people owning portions of a PSU, and having to liquidate them at a hefty discount on short notice when some people opt out of the hosting. Not to mention the issue of who would be responsible for replacing defective PSU's if they do fail, down-time due to PSU issues, as well as additional costs for having heavy PSU's shipped cross-border to return them to the potential owner(s) if opting out of hosting. I would assume responsibility for all this.

I will update as soon as I hear back from GND.

They definitely won't knock the retail costs of their pi kits off the total cost of the miners if we source from someone else.
And please remember that the listed prices of hardware are from a recent quote and are likely to change based on where the BTC price goes and qty purchased. We will only know the full cost after we have a certain number of members ready to buy.

Of course, if they jack their price up then I can refund you guys the reservation fee and we can abort the purchase--it's only fair.

From what I've heard from blockC, the pi's come configured.
Great North Data is not cool at all with giving anyone remote access. Sorry.   The onsite tech will do all those operations himself--I have their contact info.

Another update: Finksy got back to me with an excellent idea on how to deal with 2.5 PSUs a miner while minimizing the cost. He's putting together an option for a 3 month (renewable) lease on all the required PSUs. This of course cuts down our upfront costs by a fair margin and makes withdrawing from the buy much less of a hassle (we simply send excess PSUs back and we renew a smaller qty in the next lease).

If the price is reasonable I would be very excited for this option!

Loving it so far!

Are the Pi's pre-installed with some kind of software for running the avalons? Or do we need to do that ourselves? If so, who will do it?
How will you have remote access? Would it be possible to get remote access to the miners web gui for the participants in the group buy? I understand if we can't, just a question.
A easy solution would be to have ssh access to the Pi running the avalon, on some random high-port. Running a SOCKS5 proxy you could just tunnel the web traffic to your miners LAN address.
Tho this is probably a unnecessary security risk to take, and probably would require some management/infrastructure work if you wanted to split the miners into seperate sub-nets.

Again, totally understand if we can't have access. I think I just answered my own question.