If current difficulty growth CAGR stays the same those investing today will never breakeven

1246712

Comments

  • ilia7777ilia7777 Member Posts: 113
    edited March 2016
    verjic said:

    And now I would ask, why everyone said that 2 GB card should not be bought? If mining will not be profitable in couple of months, means that DAG will never even approach 2 GB size :). For example in my country difference between rig with 5 GPU's of 2 GB vs 4 GB is 250 Euro. And if I read this thread it means that 2 GB cards are best investment for anyone! :) But of what I noticed, most of people which say mining will end in couple of months have 390 or other 4 GB cards minimum :)


    This is actually incorrect, I found that I could buy brand new R9 380 with 4GB for the price of R9 380 with 2Gb also brand new. They turned out to be 7% more economical than R9 390, you just need more space. But if I were you I wouldn't do anything until difficulty rise stops. Its a zero sum game at this point. When ethereum was in early stages short time inefficency existed where ethereum was priced very high relative to the cost of mining. There was a lag until this information spread to enough people which allowed to make money. We are at the final phase of this.

    http://libertystreeteconomics.newyorkfed.org/2015/08/entry-and-exit-leads-to-zero-profit-for-bitcoin-miners.html#.Vvzexdz0HIU

    From now on we will be in zero sum game phase where network hashing power will continue to rise to the point until profits will reach electricity cost exactly as with bitcoin. The method of mining actually doesn't matter ASIC or GPU. The only difference is with powerful ASIC you are wiped out right away as soon as its released while with GPU you just face exponential decline of your profits. The end result is the same. To make it crystal clear lets use simple example. Suppose you invested 3000 dollars and the rig makes 1000$ a month. You are jumping from joy thinking that you will get your money back in 3 months. The difficulty rises 50%, which means that in a month you get 500$. Next month another 50%, you get 250$. Next month again 50% you get 125$. Then 62.5, 31.25, 15.62, 7.81. Lets add it all up, the total is 992.18 USD. Incredible isn't it? This is the power of negative compounding. Warren Buffet became the richest person in the world because his capital was compounding at just 20% ANNUAL rate. We are facing 40% negative MONTHLY compounding situation.
  • adasebadaseb Member Posts: 1,043 ✭✭✭
    You are wrong in saying $1000->$500->$250, etc ,etc

    Since to go from $1000->$500 the hashrate would need to double, but to go from $1000->$250, the hashrate would need to quadruple.

    As the difficulty grows, it becomes harder and harder for it to keep growing at the same pace.


    Even if we get 3 good months out of this... its still alot of money.
  • syaoran99syaoran99 Member Posts: 204
    adaseb said:

    You are wrong in saying $1000->$500->$250, etc ,etc

    Since to go from $1000->$500 the hashrate would need to double, but to go from $1000->$250, the hashrate would need to quadruple.

    As the difficulty grows, it becomes harder and harder for it to keep growing at the same pace.


    Even if we get 3 good months out of this... its still alot of money.

    Figures say you're wrong. The difficulty IS keeping pace. You're just in denial. Look at the difficulty increment overnight. It's now 24 Tera Hash as compared to 20.5 just a few days ago. 12 Tera Hash early February.
  • adasebadaseb Member Posts: 1,043 ✭✭✭
    syaoran99 said:

    adaseb said:

    You are wrong in saying $1000->$500->$250, etc ,etc

    Since to go from $1000->$500 the hashrate would need to double, but to go from $1000->$250, the hashrate would need to quadruple.

    As the difficulty grows, it becomes harder and harder for it to keep growing at the same pace.


    Even if we get 3 good months out of this... its still alot of money.

    Figures say you're wrong. The difficulty IS keeping pace. You're just in denial. Look at the difficulty increment overnight. It's now 24 Tera Hash as compared to 20.5 just a few days ago. 12 Tera Hash early February.
    Then go and sell all your GPUs on eBay then if you are worried.
  • ilia7777ilia7777 Member Posts: 113
    syaoran99 said:

    adaseb said:

    You are wrong in saying $1000->$500->$250, etc ,etc

    Since to go from $1000->$500 the hashrate would need to double, but to go from $1000->$250, the hashrate would need to quadruple.

    As the difficulty grows, it becomes harder and harder for it to keep growing at the same pace.


    Even if we get 3 good months out of this... its still alot of money.

    Figures say you're wrong. The difficulty IS keeping pace. You're just in denial. Look at the difficulty increment overnight. It's now 24 Tera Hash as compared to 20.5 just a few days ago. 12 Tera Hash early February.
    My friend you need to go back and study your math. I don't know how much more clear I can explain what daily/monthly compouding is.
  • ilia7777ilia7777 Member Posts: 113
    adaseb said:

    syaoran99 said:

    adaseb said:

    You are wrong in saying $1000->$500->$250, etc ,etc

    Since to go from $1000->$500 the hashrate would need to double, but to go from $1000->$250, the hashrate would need to quadruple.

    As the difficulty grows, it becomes harder and harder for it to keep growing at the same pace.


    Even if we get 3 good months out of this... its still alot of money.

    Figures say you're wrong. The difficulty IS keeping pace. You're just in denial. Look at the difficulty increment overnight. It's now 24 Tera Hash as compared to 20.5 just a few days ago. 12 Tera Hash early February.
    Then go and sell all your GPUs on eBay then if you are worried.
    I'm not worried cause the amount of money I invested is insignificant to me, but I just want everybody to understand the rules of the game. People tend to make crazy bets in desperation to make money.
  • blueboxbluebox Member Posts: 181 ✭✭
    The point of @ilia7777's thread is to suggest that it's not wise to be throwing more money at gpu systems today with the probability of not recovering the investment at a future date if you are basing it on current rates of return. His analogy of the problem using a compound annual growth rate based on diff or global hash rate for returns isn't too far off the mark.

    Diff/GHR is a parabolic rise, though may not fit traditional compounding models, because price is not a constant. I believe too many "new miners" may make the mistake of using current returns in their calculations without factoring this in. Unlike ilia7777's two-year scale, there's the one year (or less) looming brick wall called PoS, so investments should not be made that can't be recovered before then assuming price stays at current rates. If you're willing (or naive enough) to believe price will continue to rise and "keep parity" with falling ETH returns, well that's a risk I don't think many would take. FWIW
  • benvickers_benvickers_ Member Posts: 5
    I'm not great with math but have been trying to make some calculations based on current numbers and projections being thrown around in this thread.

    I wondered if anyone had a spreadsheet they could share here where they've mapped out possible variables based on different factors described? Best case and worst case scenarios etc

    I just did a very basic one for my setup, where I assumed a 40% monthly rise on Difficulty and a 10% monthly drop on my MH/s from GPU's - seems like it's okay in terms of returns over next 3 months but then gets very dire very quickly (regardless of whether Ether is $10 or $100).

  • Marvell9Marvell9 Member Posts: 593 ✭✭✭
    @benvickers_ you have the right idea always assume the worst when speculating on these things
  • adasebadaseb Member Posts: 1,043 ✭✭✭
    Honestly, I am finally starting to get worried.

    Looking at some of these posts today in this forum, there are too many people who want to mine ETH and have no idea even how to build computers. Pretty soon my Grandmother will start mining ETH.

  • blueboxbluebox Member Posts: 181 ✭✭
    @adaseb An indicator that ETH mining has jumped the shark: look at the bitcointalk altcoin mining forum, more than every other post is about how to mine ether...

    $12/coin attracts a hell of a lot more flies than $2...
  • retherrether Member Posts: 258 ✭✭
    adaseb said:

    Honestly, I am finally starting to get worried.

    Looking at some of these posts today in this forum, there are too many people who want to mine ETH and have no idea even how to build computers. Pretty soon my Grandmother will start mining ETH.

    Yes, and I'm also seeing posts in various totally unrelated places about mining. For example, on the comments of an article about gaming like "dude just upgrade your graphics card, if you can't afford just use it to mine some ETH".

    We are at the edge of the cliff IMO.
  • adasebadaseb Member Posts: 1,043 ✭✭✭
    Our only hope is GPU shortages but I don't see that happening yet.
  • ilia7777ilia7777 Member Posts: 113
    We do experience shortage over here in Russia, but US manufacturing machine can satisfy any demand even if aliens will order one billion GPUs in few months the order will be filled.
  • adasebadaseb Member Posts: 1,043 ✭✭✭
    Its good that there are shortages there at least. In North America, you can buy GPUS right now in unlimited quantities.
  • hasherhasher Member Posts: 642 ✭✭✭
    In South-East Asia, theres plenty cards around for sale. I could pick up the phone and have 500 new cards delivered to me tomorrow if I wanted to :)
  • ilia7777ilia7777 Member Posts: 113
    I will give you another example from the world of economics before I go to sleep. In US about 90% of all cargo is transported by small trucking companies, sometimes consisting of just one truck driver that owns the truck. Incredibly most of them barely survive and in many cases even loose money. The problem is there no barrier to enter this market. If you want to be in the business of morgage insurance for example than good luck to you. Not possible without very significant amount of capital. Yes mining is a bit technical , but not technical enough and not difficult enough to figure it out in few weeks or may be a month for not very talented. 8 million dollars to increase todays capacity 40% is a tiny drop in the ocean.
  • LagniappeLagniappe Dirty SouthMember Posts: 136 ✭✭
    But you're comparing hot shot trucking to ethereum miners. assembling a rig. Running it. Dealing with how to wire up your own electricity like I did...or you could pay an electrician. Of course then you must navigate ubuntu or windows. Something people with computer science degrees on this forum have even failed to do. (Handling ubuntu, that is.)
  • adasebadaseb Member Posts: 1,043 ✭✭✭
    I think what he meant to say is that it doesn't take a genius or even a smart person to start mining ETH. Which means anyone can do it and pretty soon everyone will start doing it.

  • Zealotki11erZealotki11er Member Posts: 10
    At this rate those that pay for electricity will make $0 mining. I think the problem here is that most people though price would have gone up by now. It has been 10-12 for the past month. It needs to be ~ $20 to make the same as we did before.
  • verjicverjic Member Posts: 143
    Even if I want that price of ETH to go up, at the moment I don't. If price will rise to 20$, that for real will be a problem. At this point, a lot of people a still afraid to get in, if price rise, they will be motivated. But I observed that big miners become bigger day by day. So minimum half of the problem that difficulty rise are big miners!
  • dittoditto Member Posts: 139 ✭✭
    verjic said:

    Even if I want that price of ETH to go up, at the moment I don't. If price will rise to 20$, that for real will be a problem. At this point, a lot of people a still afraid to get in, if price rise, they will be motivated. But I observed that big miners become bigger day by day. So minimum half of the problem that difficulty rise are big miners!

    You can choose to reinvest profit, or stand back I guess.
  • verjicverjic Member Posts: 143
    Problem is that if all big miners reinvest their profit, they are killing mining faster.... Probably if they will not add so fast more rigs, they would make more profits with less investments.
  • Marvell9Marvell9 Member Posts: 593 ✭✭✭
    hasher said:

    In South-East Asia, theres plenty cards around for sale. I could pick up the phone and have 500 new cards delivered to me tomorrow if I wanted to :)

    I can belive it , it will take at least a year for shortages to happen imo since there was no reason to buy alot of these glut of GPUs sitting around anyways when litecoin minining was ended by asiscs.
  • Marvell9Marvell9 Member Posts: 593 ✭✭✭

    At this rate those that pay for electricity will make $0 mining. I think the problem here is that most people though price would have gone up by now. It has been 10-12 for the past month. It needs to be ~ $20 to make the same as we did before.

    This is why i was advocating for the lower power use Maxwell cards, ive seen this movie before right now a 7990 or a 390 and 280x and 290s are still decently profitable, what happens when hash rate doubles ? then you are paying 220 watts per card for 50% of the coins you made before.

    In the meanwhile the maxwells are trucking along at 150 watts per card or lower 130 ish an will be profitable 3-6 months longer.

    Same thing happend when the 750s came performance/watt always wins in the long run because if an inifficent card is profitable those will be kicked out of the farms first.
  • o0ragman0oo0ragman0o Member, Moderator Posts: 1,291 mod
    Marvell9 said:

    Same thing happend when the 750s came performance/watt always wins in the long run because if an inifficent card is profitable those will be kicked out of the farms first.

    @Marvell9, the question here being what is 'the long run'. The 750's even if they weren't crippled and you could hash at 9MH/s were still going to take >18 months to win out over power guzzling huge hashing 7990's. With profits as they are (even though dwindling) massive hash power far outweighs the considerations of energy costs for the blocks they get now, not for the dwindling future. Whereas the maxwells can only be at an advantage when difficulty is too high really both with.
  • adasebadaseb Member Posts: 1,043 ✭✭✭
    If it ever reaches the point where you are paying the same for electricity as mining profit generated, it means its time to sell your GPUs.
  • Marvell9Marvell9 Member Posts: 593 ✭✭✭

    Marvell9 said:

    Same thing happend when the 750s came performance/watt always wins in the long run because if an inifficent card is profitable those will be kicked out of the farms first.

    @Marvell9, the question here being what is 'the long run'. The 750's even if they weren't crippled and you could hash at 9MH/s were still going to take >18 months to win out over power guzzling huge hashing 7990's. With profits as they are (even though dwindling) massive hash power far outweighs the considerations of energy costs for the blocks they get now, not for the dwindling future. Whereas the maxwells can only be at an advantage when difficulty is too high really both with.
    nah the gtx 970s hash for 22mhs non OC for around 150 watts .14 efficency vs a 390 or 390x which is 28.5/280-300 watts around .09 effciency thats a huge difference
  • workwork Member Posts: 2,084 ✭✭✭✭
    edited April 2016
    @Marvwll9 My R9 290s put out up to 31.5MH/s at ~180 watts.... I have a hard time believing 390s are crazy less efficient at ethash since it is the same.chip.
  • verjicverjic Member Posts: 143
    edited April 2016
    @work, No way you get 31,5 mhs at 180 watts! R9 390 have this hashrate at ~ 270-280 watts and you want to say that 290 can use only 180 watts?
Sign In or Register to comment.