Bitcoin Forum

Amazing how much dust builds up after a year...

I'm shutting down my machines, one by one, starting with the lowest MH/$. I might re-build one or two efficient machines, but that might not be possible given the hardware I have.
Chances are, I'll be spending the next 2 days with my air compressor and a damp cloth trying to make these GPUs/PSUs/etc. look as pretty as possible before I list them on Craigslist and/or eBay.

I did 3 machines already, which is only 6 of my GPUs.

I lived through the $1.50 BTC days and was barely breaking even -- but back then there was no ASIC mining on the horizon.

I'm really, really going to miss BTC mining. I've been doing it since May 2011. It's been a blast. The only time I hated it was when a storm came through and shut off my machines (requiring me to waste 20 minutes restarting them all) -- but even then, I bought a bunch of keyboards & mice so it would be easy to restart each machine. So even storms haven't been a big deal for the past several months.

I'm having a lot of fans die on me though -- GPU fans as well as $16 Holmes personal fans. I probably own about 12 of the latter, in various stages of worn out.

I'll be stuck with many of my video cards because the fans died. Most of the dead-fan-cards are Sapphire 5850s, but I also have (1) 5870 and (1) 5830 with dead fans as well.

For those that are curious, almost my entire GPU stock is 58XX series. The most efficient GPUs available for mining, in case you didn't know. My setups, however, leave a lot of room for improvement. My electricity rate is average, at 0.088 per kWh.

But using all this hardware (18 GPUs for a total of 5.5 GH/s) to make $1 a day on a good day didn't seem worth it. On a bad day (variance), I'd lose money. Plus there are fees when selling BTC on the market, etc. The margins are just too tight.

Anyhow, I'll probably hang around for a while, since old habits die hard. I'm curious how everything will turn out. I'm also a big fan of BTC for philosophical/financial reasons. I'm a small business owner, and I'd LOVE to see something like BTC take over. But I think my mining days are about done. *sigh*

There are tons of 58XX fan replacements, not all of them cost an arm and a leg either, that you can easily replace the stock fans with.

You're one of the few names I remember from my old mining days (I've been gone a long long time, $0.34/kWh is not friendly), and I didn't particularly like you, but it is a harkening to times I quite enjoyed, so I tip my hat in salute to end of an era.

I'm personally very interested to see where ASICs lead us, so I'll be around for a bit myself (and why I'm back after all this time).

AngelusWebDesign, you sound like an old timer now.  Best of luck!

I've got to ask, if you're selling 18 GPU's + other misc hardware, that should bring in at least $1500... why not buy an ASIC and keep the tradition going?

Well, I did spend a few days reading up on the pros/cons of ASIC pre-orders, thinking about it, and figuring out the best position on the issue.

My decision: ASIC is too risky right now. ESPECIALLY for pre-ordering non-existent hardware (yes, that includes hardware "in development")

1. No final hardware has been shown off; furthermore, there have already been many delays to the ship date.
2. One or more companies could be a scam. Bitcoin is all about making easy money, and it HAS attracted many scams and scammers in the past.
3. Recouping one's investment might already be impossible, unless you're in the top 5% of the wait list.
4. ASICs are only good for ONE thing. As gutsy as buying 18 GPUs was, it wasn't really all that risky. They can always be sold to gamers at any time. If you can't make back your $1300 Single due to skyrocketing difficulty, you're screwed. No one's going to shut off their ASIC. No one's going to repurpose a Single into a gaming rig. If someone quits and sells his Single, guess what? The buyer will fire it back up and it will be back on the Network. When the difficulty goes up, it WON'T be coming down...ever.
5. The whole ASIC thing gives me a bad feeling -- I mean about profitability. There is no barrier to entry; just plug & play. No more cooling setups, building clever rigs, finding space for them, watching the 15 amp circuit breaker limit, etc. My entire being says this is where things go "parabolic" -- where a dozen individuals or companies come in and BECOME the network. Or at least 95% of it.

Is your electricity is that expensive that on 5.5 GH/s you could only make $1 a day?

Also when you dusting off with air compressor, the best thing is to turn on your vacuum and suck in all the dust that comes out. I try to do it every week to all of my computers.  ;)

Interesting theories and thoughts Angelus, thanks for sharing.  :)

There's no way I'd have done the dusting indoors. It had to be done outside! :)
It's not too cold out, anyhow. High of 70 today here in south-central Texas.

Electricity here is 8.8 cents/kWh.

I think a lot of people aren't good at interpreting electric bills -- which is a whole other topic. Don't worry, I'll post that later :)

Best post I've seen yet on the current and near future of mining. Came to the same conclusion and dumped 8 GPUs last month on ebay and got over 75% of original cost back. Check out ebay now...some 6950s standing at 80 bucks! The entire mining scene has become sketchy and almost silly. The ASICs may or may not turn out to be dependable and outstanding in many ways...but as it stands now, it's just as likely that many will be scammed or stuck with unworkable/unreliable equipment by amateur industrialists that fold not long after making couple million bucks.

But if ASICs do ending working as promised, it's even worse. Like stated above, this will become a plutocracy where a few well-heeled miners will drive the difficulty stake into the heart of all the small miners like myself. I'm out of the game now and kind of relieved. It was a lot of fun and very challenging but time to move on to a new hobby. There was never really any money in it for small timers like me but it was fun fooling myself :)

I doubt very much that this will happen.  Any fluctuation in BTC price would directly correlate with profitability of any miners, and if such a plutocracy was to form, it would have to be due to electrical expenses.  I don't see any point in the near future where the BTC price is stable enough to allow big miners to out the small miners just based on electric costs alone.  Any fluctuation could make or break a "big miner" mining on that thin of margins.

Everything is a matter of perspective. One neat thing about ASICs is that if they work as intended, even the small fry will be profitable for a long time. To whit:

http://tpbitcalc.appspot.com/?difficulty=993438908.9602&hashrate=60000&exchangerate=12.42&bitcoinsperblock=25.00&rigcost=1300&powerconsumption=60&powercost=0.10&investmentperiod=355

A BFL Single SC @ 1Billion Difficulty, still makes money. Not much, but not nothing. At a Billion difficulty (300x now). Something current day GPU miners could never say.

Exactly  ;D

With the lower power consumption,I could have 500GH & use almost 1/4 the power of my current 3GH mining setup  8)

Probably not the best example...  Did you notice the Net Profit line?  Or the 15 year payoff?  ;)

It's not ideal compared to what we are used to, but a 15 year payoff is 6.7% annually. This is considered to be a good investment in the non-bitcoin world. If you found such a good investment, you could ride it infinitely with a 4% Safe withdrawl rate (http://www.mrmoneymustache.com/2012/05/29/how-much-do-i-need-for-retirement/).

Bitcoin people kind of have their heads up their asses when it comes to what a normal return is.

I think I might be one of those people with my head up my ass when it comes to normal return because I don't know what you are saying. How in the world is it a good thing that it takes 15 years to pay off a piece of technology before a profit is seen? Not to mention it's technology that can, within months, become inferior to subsequent ASICs or whatever new technology that will generate 100x the best ASICs being made now. These things could be obsolete in 6 months for all we know. If anyone thinks the bitcoin technology and circumstances are going to be stagnant for 15 years, they're kind of crazy.

I think at this point the honest person has to say to themselves with regard to ASICs, "I don't know". If you're a person of means in a good financial situation and can afford to throw around a few thousand dollars for fun, I say have at it and you are lucky, appreciate it and enjoy it. But if you live on a tighter budget like most of the world where a $1300 loss means something, it's probably in your best interest to be agnostic about ASICs and the bitcion ecosystem over the next 6 months and admit there isn't enough solid data to say anything with confidence...admit "I don't know".

Let's use your post as an example:
Paying off something you plunked money down for-- is not profit.

Hey bud. I need a 5870 reference design with a dead fan .


Do you have one?

BitBlitz, I was not suggesting anyone BUY in to asics at 1billion difficulty. The point was that even a small miner who already had one (hopefully someone bought it waaaaay earlier) could leave it hashing away even if other people jacked the difficulty up to 1 billion later on, which presumably would take quite a long time being a Three-Thousand percent increase in difficulty/hashpower and still make monies.

Would I recommend buying an ASIC of the current gen specs if difficulty were at 1 billion? Of course not.

Agree.  Paid for devices can be profitable for a long time.  It will be interesting to see how fast ASICS sell after the difficulty rises from the first few batches.  It does seem to be a "get in early, or watch from the sidelines" game.

Keeping it on-topic (Or my first on-topis post in this thread :)):  Nice write-up, Angelus..

This statement about "most efficient" is not actually true, but it's interesting to hear from people who are shutting down and throwing in the towel, and why, and so on.

FWIW, it is still possible for long-term mining individuals to rule the roost, so to speak, in comparison to people with rent/property tax/staffing costs. I would say the people who are using the profits from prior mining efforts have a significant advantage to the newbs who are going into the red to buy up ASIC.

For people who are still in the black after the pre-order frenzies? Everything the ASICs make is technically a net profit from The Mining Operation, counting from day-0. "Still ahead, gonna stay there until it can't pay for its own expansion."

My advice for newbs trying to do it semi-seriously, or for midsized profits by climbing into the pre-order lists this late? For competing?

Too late. You now have two significant disadvantages and you haven't even turned on a miner yet.

Here's the only reason these people should be trying to get ahold of ASIC: They want pristine coins with no history, and in small amounts they can use to buy small things, or want to get ahold of a handful of Bitcoins for long-term long-haul storage. It is a deflationary currency after all.

Ok, would you care to enlighten us all on what the most efficient GPU is?

Its actually an underclocked 7970. You can run them at about 0.95v at 550m/h. They use about 120 watts each. 5970 are awesome as well but your best case looking at 600m/h and about 190 watts

also check out the thread on re-oiling gpu fans... it works about 90% of the time. Drill a hole about 3 mm off center and drip some 3 in 1 lube in there

Wow, this is impressive.
I can send that money away to a bank for 5.5% annual interest with only 1 year lockdown of the cash.

If a bank promised me 6.7% if i invested my money with them for 15 years i would laugh them in the face.

There are a few places, in the real world, where you get 7+% for less than 15 years of lockdown :)

Taking a chance and buying an asic that might be worth 0  and generate 0 USD, 7 or 9 or 14 years into the future is just not an option.

/GoK

Well, I wasn't talking about underclocking.

But even so, I have my serious doubts that an underclocked 7970 is more efficient than an underclocked 5970.
Your claim doesn't even make sense. The 7970 has more "extras" for gaming that would certainly draw a bit of power, regardless of how fast you're running the card (via the Voltage setting).

Your claim would have to be independently verified before I would believe it.

those numbers were verified with a killawatt now it has been months since I did it so my memory may be off +/- 10 watts but I remember that 7970 was much more efficient. It is more efficient because it is only 28nm process which needs less power. If you are looking for efficiency your not going to leave it running at stock clocks or volts

What bank offers 5.5% interest rate for 1 year lockdown? I'm honestly curious, as I have money that needs investing.

Actually, it's not a 1-year lockdown. It's a 15-year lockdown.

5.5% APR would be impressive for a 1-year CD in 2012.

However...

CDs are backed by the FDIC. CDs are considered "cash" in the realm of investments, and the lowest risk one can have in a portfolio (except maybe physical gold bullion!)

Bitcoin mining would be more like Penny Stocks -- extremely volatile, and in the highest risk category in one's portfolio.

So to get a reward comparable to CASH, while taking a risk comparable to HIGH RISK STOCKS -- it just doesn't add up for me and many others.

You are not the only one, been doing the same with some of me less efficient rigs. Decided to keep my 2 best ones running as long as they generate some kind of profit (free heating and it's like -15C here now :p). Luckily since I've been running my cards undervolted for almost a year now, I've not had a single fan die out of 20 cards since they barely spin up when each card pulls only 75-80W. 

I am going to have to back up the 7970 claims. 

clock a 5x7970 rig at  930/780/0.937v  and it pulls about 840w at the wall with 2 SeaSonic 750's. 
--probably less since I have them all on 220 now. also the mobo has all defaults, could probably squeeze another 50 from it if one wanted to.
 
still getting 3.3Mh / watt   5970 and 5870 cant compare.


Back on topic.
I have not shut any rigs down, although i did downvolt everything cause I pay about 0.125c for power.
after elec profit about 30bucks a day with 47.5G.

I have 2 months or so worth of loyalty points at coinlab, will be running them very soon.  Hoping coinlab gets the HPC work going so I can keep the farm running.

Preordered some asics, bummer they didnt come in before the reward halving.

I shut one of my machines down the other day and was shocked at all the dust that had built up over two months.  Insane...

I like reading these posts from the folks like myself that were around buying up 5850s before the June 2011 "hoarding". I've actually been considering shutting down my 8 GPUs and parting them off into gaming computers for friends and family (and their kids). They've more than paid themselves off, that's for sure.

How much is a 7970, and when could you possibly have bought them? They haven't been out that long.

Do you hope to recoup this very recent investment?  A $500 card depreciates much more in 6 months than a $200 or $150 card.

I personally haven't added to my GPU collection since August 2011.

Keeping a mining operation going is one thing. Going in for new cards? For some time now, that's been a bad idea -- especially with the Great Halving and FPGA/ASICs on the horizon.

I have been building my 7970 rigs all year, the rig I showed has been running 24/7 since ... ithink january. downloaded the beta driver in mid january

whats your power setup with that? you have to be pulling close to 100 amps right? I assume this isn't residential? Most homes only have 100 or 150amp mains

I can honestly say that I have kept my main breaker 'warm' throughout the year.

Most homes in our area have 200 amp service.

1 year CDs are at about 1.1%, 5 year CDs are at best 1.8% (http://www.bankrate.com/cd.aspx). I know that there are temporarily good interest rates to be had in Australia, but here in America I don't know where you think you're getting such a good rate.

6.7% on an item that has an lifetime warrantee like the BFL SC is all profit because of the capital that is stored in the machine itself, which can be sold at a future date. Saying that it's not profit is like saying a landlord isn't making a profit on the rental property they purchased. It might sound right to a layperson but is makes no sense in the financial world.

Basically everyone who responded to me proved my point: they don't know anything about the realities of finance and what a person can reasonably expect from an investment. Let's just assume that ROI is as low as 50% per year. That means you could open an account with $1000 in it and in 19 years have $2.2 million. But people don't retire on $1000 tucked away once when they are 46. It takes years of saving $1000 or more every month to build up a retirement account you can live on.

But this thinking is an epidemic in our culture and it has lead to people pushing tons of money into the hands of scammers. The only reason we have had this return so far is because we've been operating in an environment with 25% or more monetary inflation but it's been balanced by the exponentially expanding network of people being introduced to bitcoin.

Why not buy an FPGA or pre-order an ASIC? Yeah, they're expensive, but they're much more efficient than any GPU, and if you pre-order an ASIC you should be able to rack up some cash before the difficulty increases a shit ton.

You probably would want to read this thread first then:


then following that: