Hardware / Software technical advice please (new to mining)

[Shubin]

Hello,
Please excuse the ignorance factor here as I am new to mining but will be quick to learn.
I would like some advice and suggestions regarding mining using small distributed ASICs in PCs

As a business we are fortunate as we rent out several hundred Dell Optiplex PCs. Running windows The OS running is XP Pro and Windows 2003 server.
These PCs are on 24/7 and connected to the internet.
Our clients would be more than pleased to have us charge them a lower price by using the PCs to mine.

We have ordered 3 Jalapenos’ for evaluation.
We would be hoping to install these physically in the PCs once we have evaluated the PSU requirements.
We would also like to be running the software unattended under the system account.
Any advice as to which software we should be installing on the PCs would be appreciated.

Our existing software has a sentinel that we have developed as a watchdog.
Our intention would be to monitor the mining software and the miners and restart them if there was a problem.
Any advice in this area would also be appreciated.

With respect to the business model I have been told by BFL that we could anticipate a return on investment in approx. 8 months.
Is that possible and if so what would forum members consider the expected the useful life span of an ASIC device to be?
Considering alternative Crypto currencies, is it possible to flash an Asic or use it for other protocols?
A big thanks to everyone for creating a very informative and mostly positive forum.

S.

[psilan]

Care to quote the Bfl 8 month roi that you were given?

[ice_chill]

As far as I understand you can run even 100 ASICs off a single laptop (just use 100 port USB multiplier), you don't need PCs.
PCs used to be good for mining when it was done on the CPU, but this no longer brings any money.

[Shubin]

Care to quote the Bfl 8 month roi that you were given?

Sorry I dont understand the question?

[psilan]

" I have been told by BFL that we could anticipate a return on investment in approx. 8 months. "

Who from butterfly labs said that about Asic and can you quote then? I find it hard to believe.

[Shubin]

As far as I understand you can run even 100 ASICs off a single laptop (just use 100 port USB multiplier), you don't need PCs.
PCs used to be good for mining when it was done on the CPU, but this no longer brings any money.

The only reason to instal in individual machines is.

1. Power is not paid for by us (I realise this is probably a trivial consideration with Asics)
2. All being in one place is a single point of failure for PSU or PC or network.
3. A Dos attack is less likely to happen on a distributed network of devices.

Remember these PCs are already being used and powered in a commercial environment.
They are willing hosts to the mines and offered without any costs.

Your point is however noted and we will be installing a mini rig or similar in our server room.
That is if the ROI figures quoted are found to be realistic.

Regards
S.

[Shubin]

" I have been told by BFL that we could anticipate a return on investment in approx. 8 months. "

Who from butterfly labs said that about Asic and can you quote then? I find it hard to believe.

I think it was Josh at the conference.
He was sat with the Minirig on Saturday 15th Sept approx 1.30pm.

Why are your doubtful?
Please tell me more.

[psilan]

So, that is the mini rig then, not Asic.

Sceptical not of the claim, but the fact they said that about Asic. It would be shooting themselves in the foot.

[Shubin]

Sorry maybe I am being a bit dumb here.
Are you saying that ASIC ROI would be better than that?

[ice_chill]

There is ASIC MiniRig doing 1000Ghash and is yet to come out, and there is an FPGA MiniRig doing 25Ghash and is already out.

[psilan]

Sorry maybe I am being a bit dumb here.
Are you saying that ASIC ROI would be better than that?

Just quote it

[Shubin]

Hello Psilan,
Sorry but I am new to these forums.
I am just trying to get some advice.
Your answers are far too cryptic for me to understand.

Can anyome tell me typically the BTC return per Ghash of processing?
Hopefully I will be able to do my own sums from there.

[bobitza]

Imho, the decision to invest on that should be based on the following concept (taken from Wikipedia - Bitcoin)

Difficulty is intended as an automatic stabilizer allowing mining for Bitcoins to remain profitable in the long run for the most efficient miners, independently of the fluctuations in demand of Bitcoin in relation to other currencies.

That means in the long run, the profit per Mh/Gh will be just a few pennies above the miners' average cost of electricity (perhaps a few pennies more for depreciation costs).

The keyword there is "average". So, to be profitable, you need to run your rigs somewhere where electricity costs are below the average cost paid by all the miners in the network.

The time needed to recoup your initial investment depends on your $/Ghash/s aka how much did you pay per Gh. The longer the time, the more risks are involved (new technology coming along, current technology getting cheaper). The bigger the initial investment however, the more money are you making right now ... so you need to find your sweet spot that balances these two variables.

[Shubin]

Its still all a foreign language to me.
If a Jalipino from BFL costs IRO $150.00 it is 3.5 GHs
In the current environment how long will it take to earn 15 coins?

Regards
S.

[pieppiep]

Its still all a foreign language to me.
If a Jalipino from BFL costs IRO $150.00 it is 3.5 GHs
In the current environment how long will it take to earn 15 coins?

Regards
S.

At the moment the difficulty is 2,864,141.
This means you need to hash at average 2,864,141 blocks of 2^32 values to find a block of 50 BTC.
The average time it takes to find a block is 10 minutes.
This gives 2^32 * 2864141 / 10 / 60 = 20.5023 Tera hashes / second.
If you add 3.5 GHash/s to it the total speed will be 20.5058 THash/s.
Your part will be at that moment 20.5058 / 0.0035 = one 5858.8th of the network.
With mining on a pool you get on average 50 / 5858.8 = 0.00853417 BTC each 10 minutes, so 15 BTC will take about 1758 times 10 minutes or 12.2 days.

But when the BFL miners will be available you won't be the only one who got one of them.
If they sell 3000 jalapeno's, 1000 single's and 100 mini rigs (I just made up this numbers) that would be 3000 * 3.5 + 1000 * 40 + 100 * 1000 GHash/s = 150.5 THash/s.
Total speed would be 171 THash/s of which you own 3.5 GHash, a 48857th of the total.
This will make the time to earn 15 BTC 48857 / 5858.8 = 8.33 times longer.
The 12.2 days becomes 101,6 days.

If they deliver after the reward half it takes 2 times longer, 24.4 and 203,2 days.
If they even sell more hashing power than my wild guess.... and then some competitors come with their ASIC's....

[bobitza]

@pei
Thanks for crunching the numbers for us. Very good example.

@shubin
There is no "current environment" once the ASICs hit the market. The difficulty will sky-rocket and it will take a while before it stabilizes, so any numbers thrown are just guesstimates. (Un)fortunately I think pei's numbers are actually pretty conservative; I expect the difficulty to raise even more.

As more and more miners bring ASICs online, those 203.2 days will become 303, 503, 603 days, etc. At some point in time, miners will decide that the time to recoup the investment is too long and will no longer buy new ASICs. This is when the difficulty will stabilize somehow.

So, with a stable or slow growing dificulty, most of the miners will now need to make a decision on keeping or not the ASICs online. Since they already paid for them and they are not good at anything else, that decision will be made by comparing the electricity costs with the benefits obtained from running the machines. Since we have to pay the electricity in USD, EUR, etc. the echange rate will play an important factor here. Basically if you mine and lose money doing it, the logical decision is to shut the whole thing down.

Which brings us to my previous post: if you mine with electricity costs below the average cost of all miners, you will be profitable (aka check your local rates with the average rate paid by the mining network). Will you recoup your initial investment? Well, it depends on where we at with the difficulty. At 203 days or 603 days?

Influencing factors:
- exchange rate BTC-USD fluctuations
- some miners being illogical
- technology leaps (ASIC 2?)
- and a gazillion of other things, lol.