Essay: A Guide to Mining in 2013

[faceman]

Essay: A Guide to Mining in 2013
By: Tom Peckelpent

This is the fourth essay. An essay for all seasons.

Welcome to bitcointalk.org, the go-to place for all things bitcoin. In this essay I will cover the best ways you can mine bitcoins. Why would you want to mine bitcoins? The idea is of course to make money, no pun intended. While there are many reasons to mine including promoting a robust bitcoin network, the best one is the money. And there is still a lot of money left to be made mining. Here's what you need to know.

First, you need to know what mining really is. Mining is a promise to buy bitcoins in the future. Think of mining hardware as a bond. Something which you buy, and which in turn pays back a certain amount of money to you at a fixed rate. This rate is variable. Some bonds pay out each year, some pay out each quarter, and some pay out each month. Still some bonds pay more or less often than that. So mining hardware is bought at a fixed price just like a bond, and mining hardware pays out at a variable rate as you mine coins with it. This is not the same as buying coins because you need to think about stuff like the cost of power, or mining difficulty. The takeaway is, you need to know how to keep long term costs down.

Power (electrity) costs are controlled by class of hardware. This is because class of hardware affects all of the other variables. When you mine with a GPU (a video card) your electricity cost is very expensive because it must support the rest of the computer too. Plus, the GPU itself is not as good for mining compared to a FPGA. FPGA units used for bitcoin mining use one tenth the power of a GPU. It's true that they cost more, but with power costs up these days it is good to know. If you're planning on serious mining and you leave a computer turned on 24 hours a day, 7 days a week it will cost more than $100 per month. At the end of the day, running a couple video cards versus a FPGA unit is likely to cost more than the computer (and forget the video cards)! Mining with a GPU is great if you're just starting out, but we have now passed the point where buying a rig just to mine is going to be a waste of money.

ASIC units are even better than FPGA. If you're looking out six months or a year you will save even more money. So much so that it may not be worth getting into FPGA mining "now". If you are serious about making money mining, do not buy FPGA units "now". You will end up making less money than waiting and buying ASIC units when they ship in a couple of months.

The only long term cost you will face is power. Now that you know about that, you should think about cost per mHash (or cost per gHash). Here is what you need to know:

1. BTCFPGA's bASIC-72 $1069.99/72 = $14.86 /gHash
2. BTCFPGA's bASIC-36 $599.99/36 = $16.66 /gHash
3. Avalon ASIC $1299.99/66 = $19.70 /gHash
4. BFL MiniRig 'SC' = $29899/1500 = $19.93 /gHash
5. BFL Jalapeno $149/4.5 = $33.11 /gHash
6. BFL "little" Single SC = $649/30 = $21.63 /gHash
7. BFL Single 'SC' $1299/60ghash = $21.65 /gHash
(Note: No ASIC units are shipping yet. Buyer beware of fraudulent ASIC order takers. For maximum safety do not pre order any ASIC product.)

This is just a quick look at three companies. There are many more. This does not mean you should run out and only buy #1 on this list. Competition is strong and no one is shipping yet. But you need to understand what is going on compared to the previous generation FPGA designs.

1. BFL FPGA Single: $599/0.832 = $720 /gHash
2. Enterpoint's Cairnsmore1 "Quad XC6SLX150" Board - $640/0.880 = $727.27 /gHash
3. BTCFPGA ModMiner Quad $1069.99/0.840 = $1273.80 /gHash

See here, even the cheapest FPGA boards are 33 times more expensive than ASIC! And ASIC units use just a fraction of the electricity that FPGA does. It's easy to see where FPGA units will be going once ASICs hit the market: into the garbage! In short, there is no other option if you are interested in profitable mining other than to buy an ASIC product.

Then what should you buy? We once again return to the comparison between a bond and mining hardware. If you buy hardware, you will have to buy the whole thing yourself. On the other hand, you may be interested in buying a mining bond. But if you do, be careful that you are not buying into someone who is trying to dump their FPGA (or worse, GPU) units on you. Make sure that you only invest in high quality ASIC mining bonds that do not mine with GPUs or any FPGA units. The best place to invest in mining is on BTC-TC (http://btct.co). It is a bitcoin exchange that allows you to buy a mining bond in a mining company. You don't have to spend $29,899 for a 1,500 tera hash ASIC mining rig. You can spend one bitcoin or less on a mining bond that represents one share in that hardware and get about the same return.

In conclusion, it does not matter if you choose to buy mining hardware or pool your resources with others in a mining bond. The only thing to worry about is buying ASIC products. Steer clear of anything related to FPGA or GPU. ASIC is the only way to fly in 2013.

If you enjoyed this essay please donate some bitcoins to me, Mr. Tom Peckelpent, at 18dsWSHbTZFE7cWKtJPFEgTLhA3BFbdVeS. You may check this address to see how much money this story has generated by visiting http://blockchain.info/address/18dsWSHbTZFE7cWKtJPFEgTLhA3BFbdVeS. Please mine responsibly.

[b!z]

A Guide to Mining in 2013:
Don't.

[erik555]

Even with all the hashing power an ASIC offers.... it will still take months to recoup any initial investment on hardware... this is because the network hashrate is predicted to skyrocket when the ASICs finally reach the market. I've heard predictions that range from a 200TH/s network by March... to a 1PH/s network hashrate by the end of the year... so your 4.5GH/s BFL jalapeno is just gonna be a decoration piece unfortunately (i preordered one too... but a bit sorry i did so now after working out a few calculations).

What do you think the network hashrate will be by April 1st 2013? Put your BitCoins where your mouth is here: http://betsofbitco.in/item?id=1139

The only possible way of getting a quick return on investment with an ASIC is if you're one of the first one to receive it... and you can make a LOT of coins in the first 2 weeks before the difficulty skyrockets

[rudrigorc2]

A Guide to Mining in 2013:
Don't.

great advice, but, you do realize that mining is needed for bitcoin to work, right?

[jamesg]

You might want to remove BTC FPGA from the list as it seems they are imploding under their own weight.

https://www.btcfpga.com/forum/index.php?topic=1041.0

[fresca]

The chargebacks were initiated when the owner Tom disappeared during a weekend when a potential US investor was conducting due diligence for a buy-out.  BuzzDave was/is in the dark about a lot of the actual design problems as well.

What I think happened and might possibly happened to BFL is that some of the, inexperience, if you will of designing a first-time ASIC ran into problems with the highly competitive and rigid process of the conductor companies.  We read about some design problems and to me, that translates into a problem with a batch, however small, of chips that basically couldn't hash btc.  And after a back and forth the ASIC manufacturers built exactly what they were told to.  This error by the mining companies may have happened a second, or even third time.  This could explain the delays and lack of transparency.  To add additional problems to this each batch of designs sent to manufacturers eats into the overall profitability of the project, possibly making it a losing business.  At which point there are discussions of "selling" it to someone who can invest more capital and complete the project.  And in a worst case scenario, money is owed to both the ASIC manufacturers as well as all the customers who are now requesting refunds.  Myself included.  It's my speculation though and until we get something concrete about what happened we won't know.  The btcfpga.com/forum has been locked down for several days now.

[odolvlobo]

First, you need to know what mining really is. Mining is a promise to buy bitcoins in the future. Think of mining hardware as a bond. Something which you buy, and which in turn pays back a certain amount of money to you at a fixed rate. This rate is variable. Some bonds pay out each year, some pay out each quarter, and some pay out each month. Still some bonds pay more or less often than that. So mining hardware is bought at a fixed price just like a bond, and mining hardware pays out at a variable rate as you mine coins with it. This is not the same as buying coins because you need to think about stuff like the cost of power, or mining difficulty. The takeaway is, you need to know how to keep long term costs down.

This is a very poor analogy. Mining is not a promise to buy bitcoins in the future, first because there is no promise, and second because you get bitcoins as you mine, and not in the future. Mining really is like mining.

Also, saying that mining equipment is like a bond is saying that any capital equipment is like a bond and that is so wrong. I guess you could come up with a complicated scenario where bonds and capital equipment work similarly, but there is no need. Mining equipment is capital equipment.

[Moshi]

While I can say this looks like a functional guide on buying mining hardware, it's not exactly a "mining guide". Although it may just be an issue of wording, I would have expected to see less on price/performance of new mining hardware coming out, and more on how to mine with existing hardware: many people (myself included, when I started) have capable GPUs for mining as-is.

On the other hand, if this -is- a buying guide, it functions rather nicely. I think, however, that you should have much more thoroughly explained a theoretical income/cost/net profit of using current hardware such as GPUs or FPGAs, and, instead of dismissing them, simply state that they're likely unprofitable in most scenarios.

Lastly, your small chart of different ASIC units will pretty much be gibberish to anyone who doesn't already know all of these companies, their product names, and the units you're using. I think it would be a pretty good idea to first explain what all of it means to laymen, and then organize the lot of it a bit better in a nice and pretty chart rather than a list.

A good essay, though; cheers