Topic: Howto: Mining on Amazon EC2 Cluster GPU instance (Read 103660 times)

what about now

You need to create it in your .bitcoin directory.

I have started the miner and it is working fine with the production,
now I am looking how to switch it to the test network and can't
found the bitcoin.conf file. Will appreciate any help on that ?

Thanks in advance.

I hope you are right , cool I'll check it.

I imagine so.

btw , is that address http://deepbit.net:8332/  stays the same ?

No problem.

Thanks

Set testnet=1 in the bitcoin.conf file.

Hi,

I am curious: How can I launch that miner to mine on the test network ?

Thank in advance.

watching

Not really now, but you can mine CPU coins (Litecoin, Tenebrix) for free:

https://bitcointalksearch.org/topic/how-to-mine-cpu-coins-for-free-50708

What about mining tenebrix with the cpu in this baby ?

Maybe that makes it profitable ?

Watching

I was interested to see how well the Amazon EC2 GPU Clusters perform at GPGPU tasks such as Bitcoin Minings and I came across this thread. Very well defined procedure for how to go about using Amazon EC2 for Bitcoin Mining.

This is ofcourse for fun and to learn how to use Amazon EC2 GPU Clusters. Probably, if GPU Clusters with a huge stack of ATI HD 6990's were online as a Virtual Server (with root access), would be interesting to see how much it would cost per hour

yes sure will 

Indeed the EC2 compute boxes aren't that powerful for Bitcoins.
But the figures quoted are a little lowballed, i got 70-80Mhashes/s on each Tesla GPU and about 1.6Mhashes/s per 'CPU' core ~170-190Mhashes/s total.

Though it was still a fun idea you had there Kefir 

I used Amazons 'spot pricing' to bring the cost down to around $0.70-0.80/hour, though that still doesn't seem to offset the cost enough, even when i had the miners autostart with the instance to run automatically during the off-peak hours 

Esp. when my HD5850 easily does >370Mhashes/sec at a cost of about $0.10/hour.

OK, it's approximately valid for more than a week, I guess. Sorry, I'm just a pedant sometimes.

But watching the FPGA thread something might happen to difficulty soon, and on a much longer perspective I guess there'll be an overnight doubling in 2013 when bitcoins-per-block falls from 50 to 25, and again in 2017, etc.

As an indicator of how much you definitely should NOT be paying for a bitcoin, it could be a good number. But it's actually very inflated, because while the Amazon EC2 Cluster GPU instance is (probably) good for scientific stuff, it has a bad price/performance ratio for generating bitcoins, because the hardware is very suboptimal for this task. If Amazon's GPU instances had two ATI Radeon 5970 mining at maybe 700Mhash/s each (instead of 70Mhash/s for each Tesla card), then ErMurazor should have had approx 3.8BTC instead of 0.38BTC for his $20. Unless I made a mistake already, this would make the calculation looks like:

20/3.8 = 5.26 USD/BTC

Of course, my assumption does assume that someone working at the same efficiency as Amazon EC2 actually starts setting up servers suitable for mining at similar prices. That's not happening for a while... And it probably means that people selling bitcoins on Mt Gox even as high as $8.50 aren't making a fortune, unless they've generated those bitcoins when it was a lot easier.

Never at current difficulty, so never this week?

It will be interesting if the exchange rate goes high enough to make this profitable.