Silk Road forums
Discussion => Philosophy, Economics and Justice => Topic started by: bynter on February 15, 2013, 06:53 am
-
Browsing Newegg for a potential mining rig earlier today got me wondering if I should wait a year or two for prices to fall before dropping a massive sum on a shiny video card, and then that led me to wonder about the long term. Specifically, how Moore's Law will factor into mining profitability, and ultimately the Bitcoin economy. Here are some issues I predict will affect Bitcoin notably, in the future:
-Moore's Law/hardware prices developing slower/faster than Difficulty.
--Moore's law slowing down and stopping within 15 years, as clock rates approach the speed of light and hardware can't be made more complex due to the size of the atom.
---Disruptive technologies, primarily quantum computers and 3dimensional architectures
Just something to think about......
-
Video cards will soon be useless, ASIC is probably the only way to go now
-
Unless you have cheap electricity, it seems that ASICs probably aren't as safe of an investment as just buying bitcoins with the money.
-
Unless you have cheap electricity, it seems that ASICs probably aren't as safe of an investment as just buying bitcoins with the money.
ASICs are way less energy consuming than VGAs. But indeed is not a safe investment I suppose difficulty will rise extremely fast in the next months. But If you manage to get one early you re in for some serious profit. The only thing that held me back was that butterflylabs seem very unreliable, I am glad I didn't preorder.
-
You're too late, the first ASICs are already being shipped. By the time you get one, it'll take quite awhile to earn back the investment and start profiting I should think -- though I haven't done the math to back that up, bear in mind.
Mining is no longer profitable, as far as I know. Every thing I read seems to say the same thing: people have been lucky to break even, if they did at all. A select few got in early enough or sunk in enough to begin with that they made a small profit, but nothing worth the time and effort and risk.
-
Hmmmm.... This will probably seem "small time", but when you mean "non-profitable", do you mean in the sense that it won't pay off the video card on top of electricity, or that it won't even pay off electricity? Would it be worth it to someone who was going to buy a high end card for gaming anyways?
and lets also not be afraid to get a little hypothetical here; whats going to happen to the Bitcoin if its still alive when consumer quantum computers take off?
-
What will happen? Not much. The network will agree on a different proof-of-work algorithm, one that quantum computing has little or no advantage in computing. Quantum computers are a lot less impressive than they've been made out to be, frankly. I mean like, a *lot* less impressive. Even when they're working, they won't be able to do anything regular computers can't. They'll just be able to solve *some* problems much faster. Not all, or even most.
If you're buying the graphics card anyway, sure, pop another $50 into it or something just for mining I guess. For the fun of it and all. I wouldn't expect to make that $50 back though. FYI, ATI cards are much better for mining than NVIDIA. The two use different methods in their hardware implementation, and it just so happens that NVIDIA's method fucking sucks compared to ATI's for bitcoin mining. Just one of those coincidences.
-
I've had an NVIDIA GTX 460 mining for 10 days straight now, just for fun, really. It's supposed to send me an email when it's earned .5BTC. Nothing yet. I've been thinking about getting an ASIC, but I'll probably just buy two Radeon cards and crossfire them together. It won't be as sporty as an ASIC, but I built the computer last year, so I wouldn't be investing a lot in new hardware, just the GPUs.
-
I'm sorry to give you the bad news, but... unless you're part of a pool with that card, you have about a 50% chance of solving a single block if you mine for half a year (180 days). When you solve a block, you get the entire reward (25 coins). But again... if you do it for 180 days, you have about a 50/50 chance of getting anything (very roughly). But either way, you'll be paying for that electricity... :)
-
Hahahaha.... yeah, I keep up with the mining subforum on bitcointalk. I know I'm just pissin in the wind. I'm really just doing it to get familiar with the software, the terms, the process, etc. I keep going back and forth about whether I want to get into this or not. I could buy those Radeon cards, or I could just buy $600 worth of BTC. Y'know? I mean, if I did get the GPUs, I'm already behind the curve, what with the ASICs coming out now. So, if I sprung for one of those, that might be profitable, but I dunno...
-
Yeah, I hear you. I did the same thing when I first found Bitcoin actually. I ended up just sitting there staring at the fucking thing try nonces and continuously failing. I even modified the mining program to output a message anytime it tried a value that came kind of close to working.
... I realized I had started spending up to a fucking hour sitting there staring at something that probably wasn't ever going to happen. I proceeded to shut it off before I became addicted to gambling as well :P
-
What will happen? Not much. The network will agree on a different proof-of-work algorithm, one that quantum computing has little or no advantage in computing. Quantum computers are a lot less impressive than they've been made out to be, frankly. I mean like, a *lot* less impressive. Even when they're working, they won't be able to do anything regular computers can't. They'll just be able to solve *some* problems much faster. Not all, or even most.
Maybe for most tasks, but the nature of their architecture lends them to being crazy good at prime factoring, and Bitcoin is all about encryption. and what about when first enter the consumer market, where a few rich enouh to affrd tem will get a few months to mine at a high margin before theyre rendered mundane(like what we just saw w/ ASICs)
Beyond at, there will be other disruptive technologies, like 3 dimensional crystal architecture.
If you're buying the graphics card anyway, sure, pop another $50 into it or something just for mining I guess. For the fun of it and all. I wouldn't expect to make that $50 back though. FYI, ATI cards are much better for mining than NVIDIA. The two use different methods in their hardware implementation, and it just so happens that NVIDIA's method fucking sucks compared to ATI's for bitcoin mining. Just one of those coincidences.
Coincidence? I thought it was due to Intel's philosophy emphasizing clock speeds over throughput.
-
Maybe you didn't get my meaning. If that appears to be an imminent event, I believe that the majority of Bitcoin-implementing program maintainers will come to a consensus on a new proof of work algorithm if absolutely necessary.
Otherwise, whoever owns a quantum computer will have a period of 2 weeks where they'll make more than other people. And that's about it; the network self-adjusts every two weeks so that a block is solved on average every 10 minutes.
I'm also not sure if they really would/will be that great at factoring... but it's been a really long time since I bothered thinking about how they work on that detailed a level... and I'm too lazy to look it up right now. Anyway, what do you mean by 3D crystal architecture? Are you talking about Intel's 3D CPU design?
What about it? It's a way around the problem of... fuck, I don't remember what they call it. The more transistors you squeeze onto a little chip, the closer together they are and the more electricity is going through the chip. There's resistance in transistors, and electricity traveling through a resisting medium radiates energy in the form of heat (among other things). Eventually that heat becomes so intense the chip is no longer viable. Moving from a 2D architecture to a 3D architecture will allow them to keep stacking transistors tightly packed together while letting more heat dissipate faster. It's just a way to avoid melting the CPU as soon as you turn the computer on. It's not a magic trick or anything.
-
Maybe you didn't get my meaning. If that appears to be an imminent event, I believe that the majority of Bitcoin-implementing program maintainers will come to a consensus on a new proof of work algorithm if absolutely necessary.
Otherwise, whoever owns a quantum computer will have a period of 2 weeks where they'll make more than other people. And that's about it; the network self-adjusts every two weeks so that a block is solved on average every 10 minutes.
I'm also not sure if they really would/will be that great at factoring... but it's been a really long time since I bothered thinking about how they work on that detailed a level... and I'm too lazy to look it up right now. Anyway, what do you mean by 3D crystal architecture? Are you talking about Intel's 3D CPU design?
What about it? It's a way around the problem of... fuck, I don't remember what they call it. The more transistors you squeeze onto a little chip, the closer together they are and the more electricity is going through the chip. There's resistance in transistors, and electricity traveling through a resisting medium radiates energy in the form of heat (among other things). Eventually that heat becomes so intense the chip is no longer viable. Moving from a 2D architecture to a 3D architecture will allow them to keep stacking transistors tightly packed together while letting more heat dissipate faster. It's just a way to avoid melting the CPU as soon as you turn the computer on. It's not a magic trick or anything.
Heat isn't the prblem were facing.Within ten years, well be making CPUs so complex and tightly packed that making than simply will be impossible, so it we wont be able to make them smaller. Additionally, clock rates will be going the speed of light, so they cant go faster. Thus, technological progress has reached it's limit. However, 3D architecture would raise this limit by a factor of ^2
-
... the speed of light? Ummm... Energy, in a vacuum, travels at the speed of light. Your CPU clock is measured in Hz. 1 Hz is one "cycle" per second; what that "cycle" refers to can be any repeating process. How is it possible for that to reach the speed of light? Especially since electricity conducted through a resisting material can never reach or exceed the speed of light... nothing can, unless it's pure energy in a pure vacuum.
I'm sorry, friend, but... I'm really beginning to wonder if you have any clue what you're talking about, of if you're just so tweaked that you're having fun discussing things without thinking about them logically. I used to do that sometimes. A long time ago; like any drug, you get used to it and end up basically "normal" on it eventually.
But seriously man... the speed of light? Where on Earth are you getting that?
-
If you like, you can read a mildly technical explanation here http://www.reddit.com/r/askscience/comments/ngv50/why_have_cpus_been_limited_in_frequency_to_around/
-
I'm sorry, friend, but... I'm really beginning to wonder if you have any clue what you're talking about, of if you're just so tweaked that you're having fun discussing things without thinking about them logically. I used to do that sometimes. A long time ago; like any drug, you get used to it and end up basically "normal" on it eventually.
But seriously man... the speed of light? Where on Earth are you getting that?
Well I certainly dont know exactly what Im talking, but I do remember my qualified-as-fuck CS professor saying somethaing along the lines of the speed of light being too slow
-
Well, since just about nothing in my computer can be measured in terms of the speed of light (I wear glasses, so, even the light coming in from the monitor, coming in through the polycarb lenses and then through my actual lenses, travels at a bit less than c), I have decided to fuck BTC mining. It's too much work for not enough reward. At least, not enough of a guarantee of reward, not enough for me. That, and the fact that my income sucks right now due to war injuries.
bynter, calm down, man. It's okay. We're all on drugs here. You are right about one thing, though. IF clock rates were to approach the speed of light, they could (theoretically) go no faster than almost as fast as the speed of light. 3-D architecture wouldn't save you, either - only one thing can go faster than light (+1 karma for the right answer - don't look it up). But, we are nowhere near that. The Large Hadron Collider is 27 kilometers long, and it takes that distance and very, very high energies to collide particles at close to the speed of light. If physicists could do that inside a processor, believe me, they wouldn't have built the LHC. Your quad-core CPU is a turtle next to this thing (technically, it's comparing apples to horseshoes, but I think you get the picture). "the speed of light" and "slow" don't really go together in the same sentence, when the entire discipline of physics can only come up with one thing that goes faster.
Physics in a drug forum. Who knew?