Bitcoin Forum

With the current and projected rise in difficulty one clearly needs a higher hash rate at lower power consumption to stay profitable or reach a positive ROI at all. The train for starting small scale mining now may already have passed, most will tell it definitely has, especially if energy costs are an issue.

Despite the fact that the best price on energy I can get here is 0.11 EUR/kWh ~ 0.15 USD/kWh (excl. VAT, considering that I can let our company purchase and run the hardware) I am still considering to give it a try with a couple of the latest Antminer S3+ (https://bitcointalk.org/index.php?topic=671189.msg8329154#msg8329154) (my initial plan was to purchase 6 pieces and two EVGA1300 power supplies) or one SP30, both due to in September (20th September or earlier in case of the S3+, no date in September known for the SP30).

I know it seems ridiculous to pay for a SP30 in advance, and I previously did not consider this as a viable option, but now with Batch 7 of the S3 sold out and S3+ estimated at 20th September it's waiting either way. Although I have my concerns, Spondoolies-Tech is not considered to be a rip-off company, and hardware is actually being delivered (after all they got a rating of 85 by dogie (https://bitcointalk.org/index.php?topic=456691.0)).

You can play around with numbers forever and never get to a conclusion, I nevertheless gave it a try:

Device	#	S3+	#	S3+ (OC 250 MHz)	#	SP30
Power Efficiency	#	0.78 W/GH	#	0.80 W/GH	#	0.67 W/GH
Price Efficiency	#	0.85 $/GH	#	0.77 $/GH	#	0.93 W/GH

Profit Calculation S3+ (extrapolated to 4500 GH/s)

http://666kb.com/i/cqwoqr6jcifrowhdz.gif

Profit Calculation S3+ overclocked to 250 MHz (extrapolated to 4500 GH/s)

http://666kb.com/i/cqworl7b1d2bfjsw7.gif

Profit Calculation SP30

http://666kb.com/i/cqwos9lrc674w39qf.gif

However you turn the coin, the SP30 wins, especially as it's unlikely that you get all S3+ to overclock to 250 MHz, and there is no better hardware in sight anytime soon.

Thoughts?

Hosting costs for the SP30 and/or the decreased performance on 110v ?


The S1 upgrades announced today look even better if you already have S1 machines and power supplies, but IMO neither the SP30, the S3, the S3+ or the S1 upgraded will ROI meaningfully.

The only people making money on these machines is Spondoolies and Bitmain.

I have taken a look at hosted solutions, but according to my calculations I'm better off with self-hosting despite the relatively high energy costs (but being able to omit VAT is my secret weapon here). I can not speak for the decreased performance on 110V.

ANTMINER S3 -B7 SOLD OUT

ANTMINER S3+ -B8   Speed:    453 GH/s     Price:  0.58 BTC

Bitmain now announce the S3+, which is a revised version of S3 running stably at 453GH/s (frequency=225 MHz), while consumes 355 Watt power at the wall. Most of miners will be able to run up to 500GH/s but not guaranteed.

 AntMiner S3+ for sales shipping starts on September 20th or earlier

The SP30 is sub 4TH on 110.

My calculations above are already based on the new data for the S3+.

Since we have 240V here in Europe, calculating for 110V was not relevant for me.

Try running the numbers for the S1 upgrade for people who already have S1 machines and power supplies.
It beats all the other options.

The upgrade is .46BTC, performance the same as the S3+. OC potential to 500ghs.

This gives you a fixed cost of 4.6 BTC for  ~ 5TH

Of course it beats all the other options, but it does not help me due to the lack of previously purchased S1. If you are in the mood you can try at the Mining Dashboard | TradeBlock (https://tradeblock.com/mining/) yourself.  ;)

What am I missing? How does sp30 win any way you turn the coin?
Max profit for each of the 3 calculations is what?

Also, you are in fantasy land with 20% increase per month assumption.  You should use 35 as bare minimum and with next increases coming, it wouldn't surprise me to see a LARGE jump soon.

Why is this important? Because of salvage value.  How many people will want to buy your used s3 vs sp30?  You'll get a marginally higher resale price with s3, not sp30

The real question you might ask is why would anyone in their right mind spend over USD 3K in the hope of making a few hundred bucks after a year when there is so much uncertainty in the ROI equation.

The risk/reward ratio seriously discourages any mining by people who have to pay retail prices for equipment and electricity.

That and friedcat has released the tube

How is it you don't pay VAT ?

You ship
Ship to a colocation facility outside the Eurozone; for example we have several European customers hosting S3's with us at $65/mo.

Right, so include hosting costs in the calculations for the SP30.

However, the OP suggested he was self hosting and not paying VAT.

Next to nothing, I know.  :(

Difficult to tell what's going to happen, the recent spike made a lot of people pretty nervous. 20% increase per month and 0.1% on BTC value definitely don't match the reality, calculating with 35% will nothing ever reach a positive ROI at all.

Agreed, the salvage value is not to be underestimated, and the SP30 will clearly loose here. Another matter to consider: It's way easier to put a bunch of S3+ on different power circuits than a single SP30 (our regular 240V circuits only have 13A breakers).

Good question, why would anyone want to invest in mining hardware from now on forward? OK, some are just stupid, others may think it's fun (is it?), and the kiddies think it does not matter with something like "free engery" (THERE IS NO FREE ENERGY) in mind.

Letting the company purchase and run the hardware (indeed works if you own the company ;) ).

While I agree there is no "free" energy, there are a number of apartment complexes (not dormitories) where electricity is included in the rent. So while I consider the energy "free" from an ROI perspective, I still count it as part of my mining expenses from a tax perspective.

and in most of these apartments, are you able to run the SP30?  I doubt there's a proper circuit and a sound-proof dedicated room :)

but, winter is coming... heating included!!

I have not done the calculations but you seem to be missing the fact that if you order an S3+ you get 453GHS more for the .58 BTC with the Upgrade you pay .46btc but only get 253 GHS since your current S1 can already produce ~200GHS. There is no way that upgrade makes any sense at those prices.

Good point, and clearly a game changer for people considering that upgrade offer!

Not relevant for me though, I'm still sitting here without any miner at all (and the more time passes the more I'm disinclined to begin with; but well, for some reason baring any logic I'm still intrigued!).

good point !

IMHO, just S3s. You need a 30 amp circuit for an SP30 - right.

In your calculations you are assuming you will get a full month of mining with each difficulty. How do
you think that will happen? Each increase is on average every 14 or so days.

UPDATE: Think average of -12% every 14 days. The power of compounding-works both ways. Your ROI prognosis now
looks very different...

its amazing how the mining game has changed. it used to be bet on the company that will deliver 6 mo ahead of time. if right you win 2-3x.if youre wrong lose it all. now you know these companies will deliver (low risk) but we are now talking ~20% max roi over a year (low reward).

how much is an upset gf worth???

I cranked upset gf into my BTC calculator. Using modified assumptions of $500/BTC exchange rate, 12% per bump every 12 days, $0.10/KWR power, $0.95/GHS investment in hardware and a 85W/GHS efficiency rating, it turns out that today, an upset gf is worth precisely 43% of an upset wife.

Up Front Costs
10 S3 at 4.5TH/s is 3550 watts.  5.8BTC + PSUs.
1 SP30 at 4.5TH/s is 3000 watts.  $3895 + $300 shipping

Advantage: S3

Hosting
10 S3s, plus PSUs, lots of cabling, non-standard design
1 SP30, 1 network cable, 1 power cord, standard rack design

Advantage: SP30

Maintenance
10 S3s, individually configured, or creation of generic deployment scripts, individually managed firmware updates, or scripted, have to hunt on Bitmain's website to see if new versions are available
1 SP30, individually configured, individually managed firmware updates with clearly visible new version notifications

Advantage: SP30

Support
10 S3s, support handled through web-based tickets or email.  Long turnaround times and RMA through China.
1 SP30, direct support from Spondoolies-Tech staff including using TeamViewer to diagnose and troubleshoot problems directly on your machine.

Advantage: SP30

Time to Market
10 S3s shipped on or before 9/20
1 SP30 shipped end of September

Advantage: S3 (depending on how much before 9/20 they ship and what "End of September" means for SP30)

Disaster Recovery
10 S3s, losing a unit only costs 10% of your overall hashing speed
1 SP30, losing a unit costs 100% of your overall hashing speed

Advantage: S3

Conclusion
While the SP30 wins out in ease of deployment, support and maintenance, it loses on cost and potentially on time to market.  In favor of the S3 is precisely their ability to be spread around.  Use them as space heaters in the cold months, add a few more here and there to compensate for difficulty increases.  They can be reliably run at rated hashing speeds on 110V power.  The SP30 loses speed on 110V power due to PSU and firmware limitations.  You can test your luck and try to override those settings, but you'll void your warranty.

For data center deployment, the SP30 is the flat out winner.  The standard rack design works very well with high-density cabinets.  There's only a single machine with a power cable and a network cable.  The data center only needs to worry about managing a single IP address and providing remote access to the web-based configuration interface and direct SSH access can be controlled easily.

This is wrong. SP30 is built using a modular design. The whole unit can't fail so that you loose 100% hashrate. Each chip is independently and also each PSU is independently. So you can fail either 1 or more chips or 1 or 2 PSUs (highly unlikely). Most frequent fails are in a chip or series of chips (SP10). One SP30 chip is around 5% of the hashrate so the advantage would be SP30 here.

I agree that the SP30 is modular and can certainly have single chips fail.  So can the S3.  I'm talking about failure.  Disaster recovery.  You lose that SP30, you're out 4.5TH/s.  You lose the S3, you're out 450GH/s.  It doesn't really have to be very dramatic.  Rip the network cable out of the unit.  Controller board fails.  It can happen, and when it does, losing that SP30 costs you everything.  Hence I gave the advantage to the S3.

but that is an avg gf against an avg wife

you need 3 scales of gf against 3 scales of wife.

poor -normal - good    to poor-normal-good

a good gf  being upset  is  worth almost 95% of a poor wife being upset.

would be higher but a fully upset poor wife has a huge downside ---- alimony / child care.

(tongue firmly planted in my cheek)

Don't underestimate the hosting advantage, you will be charged for space as well as power and 10 S3's take up a lot of space which can more than make up the initial difference in hardware cost.

Be carefuls with ninjas!

He trieds to do a S3 buy that was trickys and was a lease!  And made it sound like Bitmain backed SLA - https://bitcointalk.org/index.php?topic=715299

Ship to somewhere safe with your miner! Watch out with dealing with ninjas!!

Hardly. Almost every hosting contract for miners is rated by power. Even for S3s.

I prefer sp30, cause the power supply is included and they looks more neat. Antminer with power supply is a mess..

Most regular DCs will charge you for the space too. DCs that know how bitcoin works won't charge you for space. Only for power(hint: Advania).

I've been watching the exchange rate go down over the last couple of days, that must have a huge impact on large mining farms not holding the BTC. :o

Playing around with mining calculators does not make any sense (well, it never really did) with unpredictable difficulty increase and exchange rate, but what the heck, I'm still considering investing in 6 or 9 S3+ (multiple of 3 to have each 3 on one EVGA1300) and buying some BTC (10 should do, this is a gamble, no serious investment, I've had worse luck with some stocks).

Although the SP30 (despite the higher purchase price ) would make more sense in the long run due to the lower energy consumption it pretty much is out of the picture now, because one has to think of selling the hardware off once not profitable any more, and a bunch of S3s is definitely easier to sell than one SP30. With the ROI calculations not considering the selling price of the hardware it should nevertheless be possible to reach a positiv ROI because used Antminers just sell.

What I have not found out so far:

If you place an order at Bitmaintech, when do you have to pay?

Orders placed with Bitmain must be paid in full within an hour of placing the order.  They only accept BTC, and when you place the order, you're given a BTC address to which you will send payment.

Tell that to psahx.....

Oh


Ok. A single type of problem that was solved in no time with a hard reboot. For me the winner is already in my possession hashing nice and steady.

Congrats on your S3 purchase ;)

With the current price of Bitcoin SP30s hosting requirement may now be considered a disadvantage. In less than 90 days it wont even be able to pay the hosting fee

That's relevant to both the SP30 and the S3 equally.  The advantage still goes to the SP30 regardless of what the price of BTC does.  The SP30 requires less space and less power in the hosted environment.

No, because Bitmain prices the S3 in BTC. Currently the S3 is significantly cheaper.

I agree that the 0.58BTC per unit price of the S3+ currently is cheaper, but it also works in reverse.  BTC shoots up to $800 a coin and it's not.  In any case, that's not relevant to hosting.  Hosting advantage goes to the SP30 - it's less rack space and less power consumption.

No, You are incorrect. Sp30 needs to be hosted for most people, s3 does not

I am comparing the cost of hosting the SP30 to the cost of hosting the S3.  The SP30 is going to win because it is a single rack-mountable unit that needs less power than the S3s for the same hashing power.  I already gave the advantage to the S3 for home usage, precisely because I can spread the load around to different circuits, whereas the SP30 can only be put on 2 circuits at most.

Professional hosting is about $100/kw.

SP30: $300/mo
S3: $350/mo

So you need to run the SP30 for 20 months to make up for the price advantage of the S3. Good luck with that.

10 s3's are now more than $1500 cheaper than an sp30 so it would be 30+ months.

I've learned the hard way: With pre-orders you have to factor in the possibility that your miner a) won't come at all b) won't arrive on time c) won't be on spec.   That has to be factored into your profitability calculations.  Also the farther out the delivery date, the more your having to estimate the difficulty rate, and the less accurate your results.   

If we're talking about right now, this very moment, then the S3 is infinitely cheaper than the SP30, since you can't even order an SP30.  Obviously at this point the advantage is to the S3 for everything.

The SP30 is a standard rack-mountable design with included PSUs.  Throw it in a rack, plug the two power cables into the PDU and the network cable into the switch.  Done.  You get charged standard $100/kw rates.  The S3 requires external PSU.  It won't fit into a standard rack, so requires shelves of some sort.

If I walk into a data center with an SP30 and say I need space for this, it's going to use about 3kw, they're going to say no problem.  I walk into that same data center with a bunch of S3s and say I need space, they're going to look at me funny.  Then they're going to charge me for custom space and setup fees and whatever else they want to handle my non-standard servers.

My point in all of this is that from a purely hosting standpoint the SP30 is going to win.  It's a single unit.  It's a standard rack size.  It's easy for the data center to deal with.  I already gave the S3 the price advantage for up front costs.  I did not amortize savings across the board, as that wasn't my intent with the comparison.  I considered each bolded section individually, and gave the advantage to whichever unit deserved it for that section.

Full disclosure: I own S3s.  I also owned an SP10 that I sold recently and replaced with S3s.  I've owned an S2 and S1s and U2s.  I do not own an SP30 because the unit came in far under the promised specs at a higher power usage than what the initial specs claimed.

Why would you go to a non-bitcoin-mining oriented and ask for space? The very few datacenters that charge you for space also charge $0.2/kwh or more.

And even if they do charge you for space, those ancient DCs are only capable of cooling 10-20kw per cabinet so you won't be charged 2u per sp30.


Yes, I think we all agree that sp30 is more efficient and more dense. It is also one of the best looking asics and comes with possibly the best customer support available.

But all that is pointless if it takes 30 months to pay off.

For me the question never was about self-hosting vs. data center hosting. I want to buy the mining hardware and have it here at my location.

The S3+ is cheaper then the SP30 (currently a lot cheaper, calculated to equal hashing power), the SP30 has a lower power consumption, but that only matters in the very long run. Right now nothing speaks for the SP30, especially when considering the resale value.

With the SP31 already being announced destined to be shipped in the second half of October some more current calculations:

Device	#	S3+	#	S3+ (OC 250 MHz)	#	SP31
Power Efficiency	#	0.78 W/GH	#	0.80 W/GH	#	0.55 W/GH
Price Efficiency	#	0.83 $/GH	#	0.75 $/GH	#	0.77 $/GH

Profit Calculation SP30

http://666kb.com/i/cr5fiwjsocb5zyr7g.jpg

No idea where this is leading to.  :(

The game is over unless you like loosing money.

You can mine and have profit only 1)if you have free electricity or 2) the BTC price will raise more than $1000

 in hand I would take  1 sp30  4.5 th over 10 s-3's in hand.

my conditions are :


 walk it into my house set it up for me and I fork over enough 100 dollar bills to pay you.

I live close enough to lakewood nj to do the deal.

but for now I do not want either I am not buying much right now.

Or you can just buy btc and use it either as currency or store of value

I'm tired of all the dumbasses got suck into bitcoin (usually they heard about it on mainstream media) and think,- " i want to PRINT MONEY!!!!!"

Most of these fcktards are all about "making money" without ever knowing BTC IS MONEY.

For example,

It seems that you are only noob miners with no experience at all in the altcoin world. If you can only think in Bitcoin terms then you better forget to buy ANY mining equipment, this game is experts only!

Mining bitcoin is fruitless, that ship is saled around a year ago! Forever, it will never come back!

If you are thinking of buying a low cost SHA-256 mining gear, then you are an expert altcoin miner or a total lunatic.

I am mining altcoins for  more then 8 months now, so I can tell you that if you want to ROI your gear you have to be an expert in altcoin mining, you have to have enough time (6-8 hours a day) to check new altcoins and mine them from the begining. If you are not capable of doing that, then just buy bitcoin from an ATM or from any exchange around. That is much much more profitable for you.

Have you been living under a rock? People have been able to turn a profit mining bitcoins with hardware bought only a few months ago.


Or you could just not waste your time gambling on shitcoins and just mine bitcoins like everyone else.

In the past few months the number of shitcoins has more than quadrupled, meanwhile the market cap (in usd) of all altcoins combined has more than halved.

If you want to gamble, just go to a casino where you will have better odds and higher rewards without having to sift through the cesspool which is the altcoin forums.

Well, as I've written in the S3 support thread (https://bitcointalk.org/index.php?topic=671189.msg8566384#msg8566384), I have come to the conclusion that I really have to write off the idea to start mining with either a bunch of S3+ or two SP31, as it's just not worth it.

Even when calculating energy costs with only 0.0921 $/kWh (which is an unrealistic value for me here, that I could only achieve with a little trickery) and a difficulty increase of just 20% per month they are not going to reach a positive ROI or turn negative just when reaching. And yes, it's not only about reaching a positive ROI, but to make a profit as well.

We have to face it, with a total network speed of more than 200 PH/s at the current difficulty, increasing with such rates the 28nm ASIC technology has to be considered as obsolete, as it's just not possible to squeeze it at less then 0.5455 W/GH (at the wall; that value is for the SP31). Energy is not going to get cheaper neither, so we really need new hardware.

The hardware manufacturers currently are not able to keep up with the increasing difficulty, and without a miracle it pretty much means the end of private mining. This may seem pessimistic and there may be new opportunities with tech <20nm but it really depends on the manufacturers' strategies, because if it stays the norm to keep customers waiting while mining the hell out of new units themselves till difficulty kills the opportunity for the customer to reach a positive ROI again, nothing is going to change.

I really was that close to hit the order button on two SP31, but now I'm just going to watch the BTC value, buy a few BTC for hodling when it feels right, and wait for new hardware to be announced (which most likely is not going to happen this year).

Good luck to you all, especially to those who are as late to the game as I am and didn't do the math.

As I said I am mining altcoins since january, so I know that without investning 10-100k USD into BTC mining it is impossible to get a decent profit from it! It is not for home miners any more!
I just ordered a few Antminer S3+ and I will be ROI with them around 1-1,5 months, then I will gain profit! I never mine BTC only altcoins and my mining style and hard work give me excelent rewards!

Mining altcoins is not a gamble it is only if you do not know the market or what to look for when you check the new coins! There is a way to avoid scams! I have the knowledge and the experience to do that successfully, for the rest, you should buy BTC directly before the next boom in price! It is nearer then you think!

Master One, I really appreciate your thinking and your attitude of detachment. It is rare to see people as objective in their strategical choices and not blinded by the greed.

I am looking too for a strategy to live from mining. For now, unfortunately I came to same conclusion as you.

However, here are some things I am trying to explore to improve the profitability of an investment in mining:

• Go through a company to benefit from favorable prices for electricity (0.04 to 0.09 EUR / kWh) and for the deduction of certain expenses.
• The antminer and PSU(EVGA) can be resold after 2-3 months of use which can recover a portion of the capital invested.
• Start large (60 to 120 antminers) so you can mine solo and so you can negotiate prices with bitmain (I do not know if that's possible !)
• stay in the starting block for the next generation of antminers.
  

What do you think?

@manu92, same valid thoughts, but quite tricky. There is no way to get a cheaper price on energy in the country I'm living in, and I have no interest in any cloud mining scum, so it's miners in my possession or not at all.

It's true that ROI calculation does not honor the fact that the miners and PSUs can still be solded at a price after having to give up on them, so it may not be that black and white after all.

Starting a large scale mining operation is out of question for me, because with what I have/had in mind I'd not have access to a power plant providing that much energy, and I did not want to put in that much money as well.

I have some hopes now for the shortly to be announced S4 and L1, so it may not be completely over after all. If only Spondoolies-Tech would speed up with the next generation of miners (= SP50?) instead of trying to optimize current tech (= SP31), but we'll see. I'm nevertheless quite impressed by Spondoolies-Tech company, so I'll keep a close eye on what they are doing.

Right now I can see more sense in just trading BTC. ;)

The next milestone is the final deadline (21st October) for the extended comment period for the BitLicense proposal  ;)

http://www.coindesk.com/new-york-extends-comment-period-bitlicense-proposals/ (http://www.coindesk.com/new-york-extends-comment-period-bitlicense-proposals/)