That's no joke. I was trying like crazy to get in on one of the deals over "Cyber Week", and the deals that would pop up would literally be sold out within 2 minutes of appearing. I read a few threads in FatWallet and Slickdeals in which people said things to the effect of "Damn Litecoin miners! Some of us want to play games with these!"
Shortly thereafter I came to the realization that, if this many people are jumping into the Litecoin pool, mining Litecoins is going to get a lot harder very quickly, so I stopped bothering to look.
Smart cryptominers don't mine Litecoins. They mine the most profitable coin at the moment, then sell it and convert it to LTC (assuming they believe in LTC).
Truth! I have a co-worker who had to jump on getting 2 of them from Newegg. They're nearly impossible to buy now unless you catch a 40% markup on ebay or something. I'm very sorry for all the gamers out there. This isn't fair to them. :( . I just purchased my LTC with good ol' USD. Besides Bitcoin, we both are betting that Litecoin will "go up to da moon!". Specifically, I expect LTC to reach $900 by end of 2014. I have no data whatsoever to back that up with other than the Litecoin-related announcements I sometimes see on various forums of small businesses accepting LTC along with BTC. I'm already holding some[1] and will probably buy a few more if I see another price drop.
Yup, that's exactly how I did it. coinbase.com->BTC-E.com->OfflineWallet. And I'll add extra emphasis on making sure you transfer to an offline wallet(one that only you have the private-key for). I used https://liteaddress.org/ to generate a unique pub/priv pair and as soon as the trade on BTC-E.com from BTC->LTC completed, I sent them to my newly created address. So if BTC-e.com gets hacked(it has happended before), nobody can steal my coins since the private-key is only on my machine.
Also note that both these pages work completely offline. You can git clone the repos. Copy the files to a trusted-offline computer and generate the pub/priv there for extra confidence that the private-key never left your PC.
ExtraNote: Do _NOT_ lose the private key. If you start to obtain lots of value in your coins, make backups of your wallet.dat or even better make sure you have the private-key on a piece of physical paper and properly secured in your home, in case of wallet.dat corruption/misplacement/deleted/etc. Don't end up like this: http://i.imgur.com/oJWtJjl.png
ExtraNote2: Ignore the chatbox(commonly called "trollbox") on btc-e.com. It's nonsense advice that'll lose your money, but it's funny to read. ;)
There are even mining pools such as middlecoin that shift from one scrypt-based alt coin to another depending on what is easiest to mine and sell on exchanges for BTC. As a miner, you don't have to lift a finger, just point cgminer at the pool's address and supply a BTC address for payouts, the pool operator takes care of the rest.
Some smaller altcoins (such as stablecoin) have had to make changes to their difficulty calculations because these pools, with relatively massive amounts of hashing power, come in, push up a coin's difficulty and then leave again once it's no longer profitable. Sitting on the sidelines, it's been really interesting to watch.
This is a cool side effect of GPU mining, this is all commodity hardware. In a year or less miners will need to upgrade their GPUs and will be selling their old GPUs. Gamers can then buy them up for cheap. With ASIC Bitcoin miners the hardware is useless once it becomes inefficient to mine on it.
Hopefully nobody figures out how to ASIC Litecoin. If it stays on general purpose hardware for a few years, we will have a huge network of processing power ready to be harnessed.
If someone makes Litecoin ASICs, we will have a huge network of space heaters.
GPU mining is still not as decentralized, as basically only high-end PC gamers (with their own default cards) and the people specifically buying GPUs for mining will have a shot at it.
CPU-only mining would be a lot more decentralized, if such a thing is even possible.
ProtoShares (and a few other coins, I believe) focus as much on being memory-hard as CPU-hard. While CPU/GPU can be "cheated" by designing ASICs, there's no such thing as application-specific memory, making it a better candidate for an honest signal. [1]
Won't that just lead to people making machines with giant racks of RAM, causing the exact same difficulty race as we're seeing with the giant racks of ASCIs in Bitcoin?
And even if that's not the optimization (I'm not an expert on ProtoShares or memory-constrained algorithms), I am highly skeptical of anyone claiming a computation problem is optimally solved by a laptop, and cannot be optimized with custom engineering.
It's about cost versus yield. Custom-designed ASIC chipsets are orders of magnitude more efficient at their intended task compared to general-purpose CPUs; whereas it's impossible to buy or make "special" RAM, and cheap RAM is something everybody wants for everything anyway.
That may be true, but there isn't any convincing evidence that botnets are targeting (for example Primecoin and Protoshares), at least I am not aware of any, and I follow this area fairly closely.
Except banks hire employees, pay taxes, provide a service, build ATMs, offer lines of credit, provide consumer protections, etc. Thats 100% different than neckbeards going through garbage dumps looking for lost coins and kids buying GPUs with daddy's credit card for bragging points.
Shame this isn't just arbitrary math instead of contributing to something like folding@home.
Creating/sustaining an inefficient financial industry [1] is a bad use of society's resources. Just like breaking a mirror for the sake of giving work to the repair man is a net loss to society.
[1] I claim "inefficiency" compared to crypto-currencies. Example: transferring $150 million via Bitcoin uses less resources than doing it via the legacy financial system: https://news.ycombinator.com/item?id=6782706
Providing services is not "broken windows." Bitcoin isn't going to get me a mortgage or a business loan. Or have a real live person to yell at when I have an issue. Or provide a level of consumer protection for me. Or give me an entity to sue when I feel wronged.
>Creating/sustaining an inefficient financial industry [1] is a bad use of society's resources.
Not only are you being reductionist and naive, that statement is 100% opinion. Its not "inefficient" to have banks and the services they provide. That's just your pie-in-the-sky take here.
I am not calling all financial services inefficient/useless. Obviously there is value in providing loans, customer support, etc.
But at least some activities of the traditional financial sector can be made more efficient by crypto-currencies, hence the value they provide to society. (Again I refer you to this one example: https://news.ycombinator.com/item?id=6782706)
The broken window fallacy is itself fallacious in that it is often used with an implicit assumption that all economic choices are zero-sum. Some times you can fix your windows and make shoes.
That isn't to say you should break your windows on purpose. That's just stupid.
>Except banks hire employees, pay taxes, provide a service, build ATMs, offer lines of credit, provide consumer protections, etc.
And all of those things are consuming real resources, far more than bitcoin ever will consume in electricity or "neckbeards going through garbage dumps". Building a bank is probably way worse for the environment, man-hours/wages spent for employees, armored cars transporting physical cash, etc.
Not to mention the other criticisms of banks which are way more significant than this.
GPU vendors also hire employees. The interesting thing about GPUs as opposed to financial system (and ASICs) is that they can be used for something other than mining.
I like to think in terms of the energy wasted by all the bank employees driving to work every day. If even 10% of those made redundant by a distributed banking system were able to start working from home doing other things, I'd guess it was a net win.
Actually, what I said (or implied) is: fire most bank employees, all of whom will end up doing something more useful, and 10% of whom will no longer need to drive to work every day. :)
Keep in mind that the mining won't last forever. Bitcoin mining has already become all but unreachable[1] for normal household resources. Litecoin's blockchain is still new enough that it hasn't reached that point yet. Just like BTC, LTC was first just CPU, now it's GPU, then ASIC will show up then finally it'll just be pools for 0.00003 LTC a week or you'll need a freakin' Pixar+Dreamworks+ILM SuperPowered-RenderFarm or something that no ordinary person can reasonably afford access to.
Okay then someone starts up some other crypto coin. It doesn't ever have to end. Boom, bust, and hopefully cash out before the bust. GPU shortages and breathing in extra coal emissions continues.
Doesnt the bitcoin network need mining to work? As far as I understand it, mining needs to be done forever, even if its impossible to collect new coins.
I don't think it does. Bitcoin can be split to the eighth decimal place, IIRC. If bitcoins become hard to get, the price should go up, then needing one whole btc to buy something should get less common. Or maybe you're right and years from now it'll all collapse and some other improved crypto-coin will have to take its place.
I guess we'll find out, because there is a hard-limit on how many btcs can be obtained(21 million to be exact) by mining, so when that runs out...
From what I understand, the work that the miners do currently needs to continue for transactions to be added to the P2P ledger. This is why whoever wants to process a transaction will need to pay a fee that goes to the miner once Bitcoins stop getting generated to reward mining in 2016 or something. Otherwise, people who are in it for the profit will stop doing the "mining".
So do the banks' and payment processors' datacenters and hundreds of thousands of employees.
In addition, when the market becomes more efficient we'll see all commercial mining done in places like Iceland that have dirt-cheap renewable energy sources and cheap cooling.
1) The total electricity used in Bitcoin mining is pegged to the bitcoin supply, which is decreasing towards zero. So the amount of electricity used should be decreasing. It may temporarily increase while prices go up, but the long term trend will be towards zero. And transaction fees shouldn't change this too much. Competition amongst miners should eventually drive those towards zero too.
2) It's not just bookkeeping that's happening, it's dispute resolution. Currently your financial disputes are solved by calling your credit card company and having them deal with it. There's a whole network of dispute resolution services going all the way up to the SCOTUS. All of which uses resources. Bitcoin does all that through mining.
3) It's not actually that much power. The hashrate is 7m GH/s. Let's assume 2 watts per GH/s, that's 14 megawatts. You could power that with three big wind turbines.
> The total electricity used in Bitcoin mining is pegged to the bitcoin supply
No, its not. "Mining" is a misleading term based on the early-phase reward. A better name for the function would be "providing resources for transaction processing". In the early phase, the reward is new bitcoins (hence, "mining" as the common name for the process), in the late phase it becomes transaction fees.
> And transaction fees shouldn't change this too much. Competition amongst miners should eventually drive those towards zero too.
Competition among miners only drives transaction fees to zero if there are a large number of miners compared to the number of transactions -- i.e., if either usage isn't high or lots of people are dedicating resources to mining (and thus, electricity usage isn't going to zero.)
> It's not just bookkeeping that's happening, it's dispute resolution.
No, its just bookkeeping.
> Currently your financial disputes are solved by calling your credit card company and having them deal with it.
True for certain classes of financial disputes when credit cards are used, but not generally the case. But, whatever.
> There's a whole network of dispute resolution services going all the way up to the SCOTUS. All of which uses resources. Bitcoin does all that through mining.
No, it does not do the same function as the entire court system in terms of financial dispute resolution. Disputes involving Bitcoin are still resolved through the legal process. In fact, the absence of chargeback like mechanisms means that any disputes require escalation to more expensive mechanisms of dispute resolution sooner.
(Now, if transactions are done in a manner which is anonymous, dispute resolution may be practically impossible in the same way that that can be become the case if you are doing anonymous drops to exchange cash and goods -- but that doesn't mean that mining is fulfilling the dispute resolution function, just that you've adopted a protocol which excludes dispute resolution.)
If there is zero electricity being used to mine bitcoin, it has failed and will be worth nothing.
Describing the relationship as 'pegged' is also sort of aggressive. A better statement would be that most people will only spend electricity if they are able to recover that cost in bitcoin (and probably a small profit). So there is some upper limit of mining activity to be expected given the current bitcoin price, size of reward and typical efficiency of mining hardware (cost of electricity will also factor in, but that should be relatively stable).
Except the lower limit is pretty close to the cost of electricity too, since there are lots of opportunistic miners interested in bringing more capacity online if there's a profit to be made.
So, if the upper bound and lower bound are pretty close, what's wrong with saying "pegged"?
Things haven't been stable enough for that to be true.
For example, you've put the current electricity consumption at something like $100,000 a day (14,000 kw * 24 hours * $0.20 = $67,200). The mining reward is closer to $3 million per day.
In the winter the waste heat can offset your heating bill. Now we just need someone to use mining chips as the heating elements in heat pumps and we could have energy-neutral mining year-round.
That's exactly what I'm currently doing. Six 7950's are running and keeping the 70 sq m first floor rather warm when I'm not here. Outside temp is around 0 degC. Even if the coin prices drop so much that it breaks even in electricity, it's still good "free" heating. Hopefully the rigs pay themselves before the price drops that much.
As long as the work is hard to compute but quick to verify, sure. But you're probably going to run into the problem of having one central entity providing work units.
Hash the folded protein with the previous POW. Problem solved, no?
(I think the important thing is that you have to prove you didn't start the current pow until the previous work unit is done, right? Fortunately, there are lots of proteins to be folded. You can use the previous pow to select the parameters for the next pow. The hashes are uniformly distributed, so that will eventually range over the set of proteins.)
Not at all. What's stopping me from hashing the result with something else? The base of the work must be in changeable. That's the whole point of POW in this case. If you want to rewrite history you need to do the POW again, in your system that doesn't need to happen.
Fascinating. Slighty related: I've stumbled over Primecoin¹ while researching Litcoin mining. Primecoin uses prime chains as proof of work. Existing GPU implementations are slower than modern processors and I'm not sure about the feasability of FPGAs and ASICS. It's an interesting idea.
That's the same in EU - amazon.co.uk shows 1-3 months delivery for R9 280X and in Poland the arrival date of new batch was already moved twice in the last 3 weeks.
3 weeks ago I placed an order fo R9 280X, the LTC difficulty is now 2x of what it was then. I think I will soon be cancelling the order as LTC price falls and the break even point is ca. 2 months instead of 1 month.
One reason is that AMD has a single-instruction bit rotate, which is used a lot in the hash function. Nvidia needs three instructions.
Another is that AMD uses a large number of simple cores, and Nvidia has a smaller number of more complicated cores. The simple cores are all the hash function needs, so AMD can run more operations at once.
The gist of it is that AMD GPUs have more, simpler ALUs, while Nvidia GPUS have fewer, more complex ones. For simple tasks such as sha-256 and scrypt algorithms, the former approach is several time faster than the latter.
It has to do with the amount of power usage and the return rate or mine rate of LTC/BTC. AMD seems to have a better ratio of coin per wattage. So basically just better bang for the buck.
Bang for the buck = efficiency, so you just answered the question "Why is it more efficient?" with, "Because it is more efficient."
The reason, more or less, is that AMD GPUs tend to have way more cores than Nvidia GPUs, and their cores can, in hardware, do the integer operation necessary for the hashing algorithm. Nvidia can only emulate the integer operation, so it takes three instructions on Nvidia vs one on AMD.
IIRC, AMD GPUs support and are optimized for newer versions of OpenCL, while Nvidias Cuda solution is much (much) less effective for mining. At the very least a Google search with that in mind will get you the info you need.
Can confirm, working in hardware marketing in Europe, many retailers and AIB manufacturers are talking about Litecoin causing stock issues (and/or sales peaks, depending on which side they're looking at it from). Have even seen a major systems manufacturer I work with change up the GPUs in their SKUs based on availability issues.
This AMD GPU shortage isn't quite so terrible. I have a pair of AMD 5850s that I've been offered $250 each for. Never imagined those kinds of returns on 4 years old graphics hardware.
I wonder if the demand for altcoin mining cards will end up becoming so great that Nvidia will modify their architecture to better facilitate it.
Alternatively, you could try mining yourself if you're curious and want to learn about cryptocurrencies. I wrote a guide on how to get started with mining in <10 minutes (mostly for friends with gaming PCs), for some of my friends who were interested.
From selling my 7950s earlier in the week on eBay UK, the going rate appears to be around 35% more than normal (Amazon) retail price. It's quite amusing, although I do feel for those caught in the CrossFire™ (sorry), who simply want a card to play video games.
Clearly, this is in no way related to the busiest shopping period of the year.
I'm sure there is some increased demand from cryptocoin mining, but really its pretty myopic if you think even half of the Nov/Dec demand for GPUs is going that way.
Took me about a week to find 2 of them on newegg. If you check newegg daily, you can probably find a couple here and there. You can possibly even find a name brand one.
Shortly thereafter I came to the realization that, if this many people are jumping into the Litecoin pool, mining Litecoins is going to get a lot harder very quickly, so I stopped bothering to look.