Bitcoin mining operation. Possible start?

[Yoshu]

So I've been lurking through the forums for a week or so now trying to gather up as much information as I can. I've formulated somewhat of a plan to get started I was just curious as to if this sounds logical/profitable in the long run. I understand the risk involved, and that my investment might not be returned to me ever but that's why I'm thinking these steps:
I've got a few PC's that I have access to, none of them really have any Mhash/s potential currently, but I'm hoping with some minor upgrades to be able to change that.
All the PC's have the required capability of running the 5850 HD radeon card which from what I've gathered is one of the cheapest/most powerful for the investment. My goal would be to get all 3 of the PCs I have equipped with just one card each, which I've seen for as low as 60 bucks sold locally. From there I would just allow the 3 machines to mine and return my initial investment. The next step would be to buy an appropriate motherboard/processor (I'm thinking http://www.newegg.com/Product/Product.aspx?Item=N82E16813130274&nm_mc=TEMC-RMA-Approvel&cm_mmc=TEMC-RMA-Approvel-_-Content-_-text-_-) as it has space to expand into 5 cards, and equipping it with a tiny HD I have laying around, then adding the 3 cards I have currently to the newly built rig and continue mining while sourcing 2 more cards to bring the rig to its completion. Then doing the same thing again (adding cards to the 3 PC's I have currently) until I can further expand into each subsequent rig. Just curious as to if anyone has any thoughts on whether or not this is a good approach? I'm trying not to spend any/as little money out of pocket here.

[Yoshu]

Yeah I've read about them somewhat, the cheapest one I've seen was one for 150 I believe that did 7-800 mhash/sec? I could be wrong.. What I propose will be much cheaper for me as I have most of the hardware already I would just need the actual graphics cards which would be a cheaper initial investment returned fairly quickly (~2 months according to my calculations). I apologize if I'm wrong, I'm still learning alot. If you could point me in the direction of some additional resources on it I'd be more than happy to read it thank you for taking time to read my post.

[jwzguy]

Yeah I've read about them somewhat, the cheapest one I've seen was one for 150 I believe that did 7-800 mhash/sec? I could be wrong.. What I propose will be much cheaper for me as I have most of the hardware already I would just need the actual graphics cards which would be a cheaper initial investment returned fairly quickly (~2 months according to my calculations). I apologize if I'm wrong, I'm still learning alot. If you could point me in the direction of some additional resources on it I'd be more than happy to read it thank you for taking time to read my post.

Yoshu, when you did your calculations, did you decrease the amount your machines were making every ~12 days by ~10% ? Are you aware that if ASICs are delivered in the next month, as promised by BFL, that difficulty will start increasing by 400% every few days until your GPUs are making less than electricity costs? It's not invalid to bet against them, I just hope you're aware that that's what you're doing.

[Yoshu]

From what I understand I am aware that the BFL systems are reportedly coming out next month, and that as such the difficulty will go down and directly decrease the amount I'd be generating with these cards, but are you saying that it's completely ineffective to set up a rig with these cards when these systems come out? I'm not doubting you at all I'm just trying to grasp a better understanding of how these new machines could make all the old rigs completely obsolete

[techstrordinary]

Just a quick one, regarding the maths.

Current consumption of electricity for a 5850 rig would be approx $1 a day per machine (300watt load running this card, mayhaps a bit less), assuming your paying the current average US price of 16c a kw/h and running it 24/7. (if your unlucky it might be more, but probably not less)
Approx PPS from one of the big sites would mean you'd be pushing about $1 day of bitcoins from 1 machine, but consuming the same in power.

I suppose if your not paying for power, then you'll eventually come out evens in 2 months or so for the initial investment in the cards, and then start turning a profit.

If your paying the power though...well.....your just throwing money out the window.


PS, you wont be mining from your CPU, so there's no need to upgrade it. I've got an old P4 3Gii running a 7950 and it pushes the same an i5 for just running the GPU.

[dooferorg]

From what I understand I am aware that the BFL systems are reportedly coming out next month, and that as such the difficulty will go down and directly decrease the amount I'd be generating with these cards, but are you saying that it's completely ineffective to set up a rig with these cards when these systems come out? I'm not doubting you at all I'm just trying to grasp a better understanding of how these new machines could make all the old rigs completely obsolete

Difficulty will go UP .. It is your pay-per-share amount will go down.

Don't forget too that the block reward will be halved soon too. All in all it's a crappy time to 'get started' in mining bitcoins.

[freefal]

I agree with ASICs and reward halving upcoming, mining is going to get a lot less profitable. (3 or 5 posts, woot!)

[420session]

I wouldn't invest in any GPU mining hardware unless they can pay themselves off in about 3 months.

Even if they are delayed, ASICs are coming. The cheapest is the BFL Jalapeno which should provide approx 3.5 GH/s on a fraction of the electricity a GPU uses. Combine that with the reward-halving that will occur in December and I don't see how you can pay off new GPU hardware as difficulty skyrockets, unless you have access to free electricity.

I'm in pretty much the same boat having just gotten interested in bitcoins. For now I pre-ordered a Jalapeno and decided to just mine what I can using my gaming rig (~280 MH/s) until ASICs are released and more info is available.

[jwzguy]

I wouldn't invest in any GPU mining hardware unless they can pay themselves off in about 3 months.

Even if they are delayed, ASICs are coming. The cheapest is the BFL Jalapeno which should provide approx 3.5 GH/s on a fraction of the electricity a GPU uses. Combine that with the reward-halving that will occur in December and I don't see how you can pay off new GPU hardware as difficulty skyrockets, unless you have access to free electricity.

I'm in pretty much the same boat having just gotten interested in bitcoins. For now I pre-ordered a Jalapeno and decided to just mine what I can using my gaming rig (~280 MH/s) until ASICs are released and more info is available.

Yes, they are. And if BFL delivers on time we're looking at 7 weeks max before the network power is drastically raised.
I don't think there's any GPU hardware available that pays off in 3 months in the best case (IE barely any difficulty hikes.) What boggles my mind is all these posts from people "just looking into mining" who haven't looked at the difficulty charts. We've had HUGE increases in difficulty every single adjustment period for a while now. Even without adding ASICs into the mix, it's unlikely new purchased GPUs will pay off unless the price goes up drastically. If you're counting on the price going up drastically - just buy bitcoins. Make a profit instead of breaking even.