Too Late to Join the Party?

[Wildy]

Hi guys,

Saw Bitcoin mentioned on a thread today, and after doing a little research the idea is fascinating! I've got a fairly basic rig at the moment, although I have two 8800 Ultras which are great for folding - though I noticed that Nvidia cards are severely outperformed by ATIs, why is this?

Looking at various charts the yields / difficulties seem to have reached the asymptote some time ago, so is it a little too late to be thinking about joining, even in a pool?

Would appreciate some feedback.

Thanks,
Mike

[1715612671]

1715612671

[russelljohnson]

if you don't pay for electricity, you can mine as much as you want.

[fasti]

If you thought about getting graphics card upgrade, get AMD one and you can get some of the money back from mining. But getting stuff just for mining.... your guess is as good as mine :)

[MoonShadow]

Hi guys,

Saw Bitcoin mentioned on a thread today, and after doing a little research the idea is fascinating! I've got a fairly basic rig at the moment, although I have two 8800 Ultras which are great for folding - though I noticed that Nvidia cards are severely outperformed by ATIs, why is this?

Ati's are better at raw integer calcs, while nvidia cards are better at floating point calcs.  The former is a better fit for bitcoin hashing, the latter a better fit for scientific simulation.

[[Tycho]]

Saw Bitcoin mentioned on a thread today, and after doing a little research the idea is fascinating! I've got a fairly basic rig at the moment, although I have two 8800 Ultras which are great for folding - though I noticed that Nvidia cards are severely outperformed by ATIs, why is this?

Ati's are better at raw integer calcs, while nvidia cards are better at floating point calcs.  The former is a better fit for bitcoin hashing, the latter a better fit for scientific simulation.

There is hadware bitwise shifts support in ATI cards, but not in the nVidia ones. Not to mention BFI_INT :)

[Wildy]

Damn, I didn't expect replies that fast!

Ati's are better at raw integer calcs, while nvidia cards are better at floating point calcs.  The former is a better fit for bitcoin hashing, the latter a better fit for scientific simulation.

Ahh, that would explain things a lot - guess ATI is the way to go then.

For the next three years or so power expenditure will be zero, so I guess it would be wise to make the most out of it! That said I'm not going to go and blow a load on an entire new rig, will probably just invest in a couple of new GPUs - it looks like the ATI 58XX models are the way to go?

Thanks for the quick replies everyone

[fnord123]

Damn, I didn't expect replies that fast!

Ati's are better at raw integer calcs, while nvidia cards are better at floating point calcs.  The former is a better fit for bitcoin hashing, the latter a better fit for scientific simulation.

Ahh, that would explain things a lot - guess ATI is the way to go then.

For the next three years or so power expenditure will be zero, so I guess it would be wise to make the most out of it! That said I'm not going to go and blow a load on an entire new rig, will probably just invest in a couple of new GPUs - it looks like the ATI 58XX models are the way to go?

Thanks for the quick replies everyone

Keep in mind that although your power expenditure is zero, your likely income stream will approach zero as well if difficulty keeps going up.  Mining for $0.1 a day isn't really that good at paying off buying two ATI 58xx cards.

[Wildy]

Damn, I didn't expect replies that fast!

Ati's are better at raw integer calcs, while nvidia cards are better at floating point calcs.  The former is a better fit for bitcoin hashing, the latter a better fit for scientific simulation.

Ahh, that would explain things a lot - guess ATI is the way to go then.

For the next three years or so power expenditure will be zero, so I guess it would be wise to make the most out of it! That said I'm not going to go and blow a load on an entire new rig, will probably just invest in a couple of new GPUs - it looks like the ATI 58XX models are the way to go?

Thanks for the quick replies everyone

Keep in mind that although your power expenditure is zero, your likely income stream will approach zero as well if difficulty keeps going up.  Mining for $0.1 a day isn't really that good at paying off buying two ATI 58xx cards.

Hmm, if the difficulty rises so sharply, how are people justifying the continuation of mining? Would this scenario not change if one were to join a pool?

[Genrobo]

Whether you're in a pool or not...
Difficulty increases affect everyone equally.

If it gets twice as hard for any individual to solve a block...
It gets twice as hard for any pool to solve a block...
Rewards are decreased accordingly, and nothing maintains the balance except for the BTC/USD ratio.

At the moment, profits are declining, because the value of BTC isn't increasing as fast as the network hash speed and difficulty.

Everything is speculation as far as that's concerned, but the certainties are that as more people join, the higher the difficulty goes, the lower your BTC rewards are.

[Wildy]

Whether you're in a pool or not...
Difficulty increases affect everyone equally.

If it gets twice as hard for any individual to solve a block...
It gets twice as hard for any pool to solve a block...
Rewards are decreased accordingly, and nothing maintains the balance except for the BTC/USD ratio.

At the moment, profits are declining, because the value of BTC isn't increasing as fast as the network hash speed and difficulty.

Everything is speculation as far as that's concerned, but the certainties are that as more people join, the higher the difficulty goes, the lower your BTC rewards are.

Thanks for that. I guess I have missed the boat then.

[Sukrim]

http://bitcoin.atspace.com/income.html

Imagine a horizontal line at your daily costs per 100 MH/s (Example: your rig does 300MH/s and costs 3 USD/day --> 1 USD).

If you would cash yout every BTC you mine at the same day, and this graph is above this line, mining is still worth it.

BUT:

If you want to buy hardware for mining, keep in mind that you will have to earn for it as well! A 180 USD GPU with 300MH/s to be paid off in 60 days would need to have this graph (at 0 electricity costs) on average above 1 USD (please note the logarithmic scale!) for the following 2 months.

[Dyaheon]

If you only need to buy 2 couple of GPU's, I'd say it's not too late. If it was a whole rig it'd be harder to say, but 2 GPU's aren't that risky of an investment.

Sure profits have diminished with exponential difficulty growth, but the growth seems to have stalled for now (http://bitcoin.sipa.be/speed-lin-10k.png). Who knows what the future holds, but if the hashrate stays around where it's now, the next difficulty jump won't be very big. Current estimate is something like 512k, only a 18% jump whereas the last one was 80%.

With 2x 58xx being able to do some 700-800+ MH/s, you wouldn't need many weeks to get your investment back, if difficulty and price stay around where they are. I'd say go for it.

[Nythain]

You only missed the boat if you are looking for an insta rich solution here. Bitcoins are still young in their existence, minting wont be exhausted for decades to come if the concept lasts that long. The boat is still leaving harbor if you care about more than just making money. Every day more and more services and companies are accepting bitcoins. Every day the word gets out more and more. Every day i make a little bit more bitcoins to spend as an alternate currency for things like pizza and amazon products and cheesy gambling.

[mrb]

Ati's are better at raw integer calcs, while nvidia cards are better at floating point calcs.  The former is a better fit for bitcoin hashing, the latter a better fit for scientific simulation.

This "floating point Nvidia myth" keeps being repeated, but it is not true.

I am the author of: https://en.bitcoin.it/wiki/Why_a_GPU_mines_faster_than_a_CPU#Why_are_AMD_GPUs_faster_than_Nvidia_GPUs? The same 2x-3x performance advantage that AMD has applies to both integer and floating point GPU instructions. It used to be that Nvidia had a (slight) fp advantage, but not anymore. Even when comparing against Nvidia's professional Tesla range which has a fully unlocked double precision unit:

* HD 6970 = 5100 single precision GFLOPS and 1275 double precision GFLOPS
* Tesla 20xx = 1030 single precision GFLOPS and 515 double precision GFLOPS

[Atroxes]

Regardless of Bitcoin price, those who have ways of not paying for electricity, will certainly have a huge advantage over those who have to pay.

Also, GPU development is entirely graphics focused at the moment. Folding and Bitcoin is extremely niche.

In 2 year, yes, the difficulty will be ALOT higher, however, GPU's are faster and need less power to run. Software has gotten better. Bitcoin price has risen (maybe?). More stores online (and offline?) will be taking Bitcoin.


All in all... If you have the money, it's a fun hobby. One that could potentially pay off. The worst case scenario? You spend $1.000~ on a hobby that doesn't pay off

It's a gamble, but it's a gamble you have a degree of control over and it's a gamble that you can somewhat tilt in your own favor by making the right choices at the right times.

Have fun  

[V2-V3]

Even if it doesn't pan out in the next few months because of difficulty increases Im still having fun and that's well worth it  in my case.

[Veldy]

You only missed the boat if you are looking for an insta rich solution here. Bitcoins are still young in their existence, minting wont be exhausted for decades to come if the concept lasts that long. The boat is still leaving harbor if you care about more than just making money. Every day more and more services and companies are accepting bitcoins. Every day the word gets out more and more. Every day i make a little bit more bitcoins to spend as an alternate currency for things like pizza and amazon products and cheesy gambling.

Mining stops at 21 million BTC.  At best, there is 5.6 years left.  However, that is theory based on 6 blocks generation per hour which is the premise behind the setting of difficulty levels.  If the BTC/USD rises significantly and confidence in it is strong, then there will be a flood of new hardware once again and we may see 13-15 blocks / hour.  Also, if new hardware becomes available [i.e. the next greatest mining hardware, GPU or not] and it is cost effective, then there will be another flood of hardware causing another spike until difficulty catches up and brings it back down to 6 blocks / hour.

So, 5.6 years is almost certainly not how much time we have left; I am guessing 1 to 2 years at most and it is highly speculative since it needs to have a real world application when coin mining ends [there are still transaction fees ... but that isn't mining] which means a real market of some type has to grow around this, not just trading coins for cash (that would just be a bubble ready to pop).  Fortunately, there are signs that some smaller vendors are getting into this, but we need to see much more growth around the bitcoin economy to support it when mining comes to an end.

[Nythain]

did your calculation include the diminishing returns on solved blocks? last i heard the estimated time frame for the end of minting was around 2036 or some jazz?

[grue]

Damn, I didn't expect replies that fast!

Ati's are better at raw integer calcs, while nvidia cards are better at floating point calcs.  The former is a better fit for bitcoin hashing, the latter a better fit for scientific simulation.

Ahh, that would explain things a lot - guess ATI is the way to go then.

For the next three years or so power expenditure will be zero, so I guess it would be wise to make the most out of it! That said I'm not going to go and blow a load on an entire new rig, will probably just invest in a couple of new GPUs - it looks like the ATI 58XX models are the way to go?

Thanks for the quick replies everyone

5850 is the most cost effective card, in terms of performance/$, but they're rare, even more rare than 5970, or 6990. 5830/5870 are the next best cards. use 5830 if you don't have much money to invest with, or 5870 if you have lots of cash.

[fnord123]

One thing to remember in reading all the pro-mining posts is that most posters have a vested interest, that is, they will be better off if mining continues to be profitable.  This isn't to say they are being deceptive - just that nobody likes to believe the gravy train they are riding will stop.