How to use low performance mining devices

[mberg2007]

Hello all,

I would like to use some low cost hardware for mining. It's based on the Cortex A9 processor (probably like the one in your smartphone) and by rough estimates it will clock in at around 1.5 mh/s.

I understand that with this sort of hashing power I will not likely be quitting my day job anytime soon. But that's ok, this is just a proof of concept for me at the moment.

Question is can such a device be used for mining *at all* ? What I'm worried about is the time it will take to find a share on just about any pool. Won't every share returned be stale? Would it help if I ran a stratum proxy and then told a bunch of these Cortex based devices to proxy through that?

- Michael

[BitcoinBarrel]

With that hashrate you will not make hardly anything using a Pool and most likely will not be to acquire a big enough balance to even cash out. If it were me I'd leave it solo mining in hopes of getting lucky and solving a block to earn 25 Bitcoins. Only problem is it could take over 1000 years.

[LogicalUnit]

Hello all,

I would like to use some low cost hardware for mining. It's based on the Cortex A9 processor (probably like the one in your smartphone) and by rough estimates it will clock in at around 1.5 mh/s.

I understand that with this sort of hashing power I will not likely be quitting my day job anytime soon. But that's ok, this is just a proof of concept for me at the moment.

Question is can such a device be used for mining *at all* ? What I'm worried about is the time it will take to find a share on just about any pool. Won't every share returned be stale? Would it help if I ran a stratum proxy and then told a bunch of these Cortex based devices to proxy through that?

- Michael

I had a similar idea based on a Spartan 3E FPGA, but I determined that it's just not worth it. If you want to experiment with mining, do it on your GPU.

[grue]

Even with a cheap gpu (less than $200), you can get more than 200x hash power. decide for yourself.

[mitty]

Regarding it only finding stale shares; the pool server will provide your miner with the current block header whenever a new block is found so your miner should stop working on the old block and start the new one; Your stale percentage shouldn't be any higher on slow hardware than fast hardware.  (unless you're using something like a BFL single that won't stop working until it finishes a whole nonce range)

Mining with 1.5 mh/s isn't worth it though, even if you had both free electricity and hardware... unless of course you find enjoyment in getting nonstandard hardware to mine.  e.g.: http://hackaday.com/2013/03/25/mining-bitcoins-on-a-nintendo/

[dben428]

http://hackaday.com/2013/03/25/mining-bitcoins-on-a-nintendo/

Pure genius.

[flipper]

With that hashrate you will not make hardly anything using a Pool and most likely will not be to acquire a big enough balance to even cash out. If it were me I'd leave it solo mining in hopes of getting lucky and solving a block to earn 25 Bitcoins. Only problem is it could take over 1000 years.

Not a bad idea going solo. It would be like playing loto for (almost) free.

[bcpokey]

With that hashrate you will not make hardly anything using a Pool and most likely will not be to acquire a big enough balance to even cash out. If it were me I'd leave it solo mining in hopes of getting lucky and solving a block to earn 25 Bitcoins. Only problem is it could take over 1000 years.

Not a bad idea going solo. It would be like playing loto for (almost) free.

It'd be more like playing a lotto where the odds are 1,000,000 times more strongly stacked against you, and the reward is  millions of times lower.

[mberg2007]

With that hashrate you will not make hardly anything using a Pool and most likely will not be to acquire a big enough balance to even cash out. If it were me I'd leave it solo mining in hopes of getting lucky and solving a block to earn 25 Bitcoins. Only problem is it could take over 1000 years.

Well, with a single device this would be true. But what if I have 100.000 of these things?

The issue here is not really how profitable a single device is, but whether or not they can be used as a matter of principle. If the miner is so slow that it can't finish one share before a new block is found, then having 100.000 of them won't matter because they will each be facing the same problem.

What I want to know is could 100.000 very slow devices be pooled together to make a viable mining platform? Just as a technology proof of concept, never mind the cost or the technical hassles involved. I understand these will be massive at best.. :-)

-Michael

[mitty]

With that hashrate you will not make hardly anything using a Pool and most likely will not be to acquire a big enough balance to even cash out. If it were me I'd leave it solo mining in hopes of getting lucky and solving a block to earn 25 Bitcoins. Only problem is it could take over 1000 years.

Well, with a single device this would be true. But what if I have 100.000 of these things?

The issue here is not really how profitable a single device is, but whether or not they can be used as a matter of principle. If the miner is so slow that it can't finish one share before a new block is found, then having 100.000 of them won't matter because they will each be facing the same problem.

What I want to know is could 100.000 very slow devices be pooled together to make a viable mining platform? Just as a technology proof of concept, never mind the cost or the technical hassles involved. I understand these will be massive at best.. :-)

-Michael

100000 1.5mh/s devices pooled together would work (they would mine bitcoins at 150 GH/s), but "viable" with respect to mining implies your income exceeding your expenses.  If you can't make one device profitable then you can't make 100000 profitable.

You would probably need to set up a local pool server (/solo mine) or use a Stratum proxy at the very least though.  No pool would allow you to directly connect 100000 slower-than-cpu/x86 miners. :p

[gourmet]

The issue here is not really how profitable a single device is, but whether or not they can be used as a matter of principle. If the miner is so slow that it can't finish one share before a new block is found, then having 100.000 of them won't matter because they will each be facing the same problem.

I think that even when trying to finish any one share, luck is always important part. So it is quite possible that some of the slow miners will be lucky enough to find a share before the block is completed. In fact, 100 000 very slow workers equals to one "less slow" one. Together they find shares 100 000 times more frequently than any one of them.

[edit:] It seems like mitty posted more important aspects of the thing in between... :-)

[PCMiner]

Yea, as a proof of concept it's a neat idea.  Even if you collected every phone and tablet you could get your hands on you are still looking at a slow mining setup combined. BUT, and BUT is the key here, 1,000,000 phones crunching away is a huge boost, even if it's not all directed at one person.  A million phones also is only a tiny drop in the bucket of the possible phones available out there to mine....  It's feasible I guess.  While it would strengthen the network, no one would really win if this happened.  Difficulty would just go up even more hurting us normal miners.

I have some old iPhones crunching litecoin using "MobileMiner".  In 2 1/2 months I've made.......1 LTC total.  They sit on my desk at work crunching away out of sight, out of mind.

[progolferJP]

What is the best way for me to mine? Which software should I use?

[PCMiner]

What is the best way for me to mine? Which software should I use?

If you want just plain simple no hassle go with Bitminter although they've been having some DDoS attacks the last few days.   If you are a little more technically minded, check out CGminer.  Just be aware it requires some tinkering skills.  For absolute beginners though Bitminter is the way to go.  It's where I started before learning more about how all this works.  I've since graduated to CGminer on my machines.

[jeannie]

I think better use for mining alt coins which are much more easy.  I use mine to mine Porncoins getting more than 10k of the coins for fun

[SirMintALot]

I was running AndLTC on my HTC Desire HD for a while. It is submitting a share every 3 hours and sucks the battery empty in a short time. IMHO its not worth the hassle

[lucas.sev]

Are you trying to develop malware for smartphones that will mine BTC for you? 

[jamesc760]

Are you trying to develop malware for smartphones that will mine BTC for you? 

I had the same thought as soon as I read op's question. Nobody in right mind would think of trying to mine on cpu's these days, let alone on weak cpu's like the arm, and where do you get 100000 of these cpus?

[mberg2007]

Are you trying to develop malware for smartphones that will mine BTC for you?  

I had the same thought as soon as I read op's question. Nobody in right mind would think of trying to mine on cpu's these days, let alone on weak cpu's like the arm, and where do you get 100000 of these cpus?

Placing a trojan in an Android phone wasn't my idea. No, I was thinking more along the lines of buying a bunch of MK802/808/908/whatever SoStick devices that I could plugin in an leave running at work or at home, since they don't draw a lot of power. You'd be surprised at how powerful these things have become these days - we're talking quad core 1.7 ghz A9's here. Definitely some potential for CPU friendly altcoins there.

My calculations however show that this is not really feasible. You can get these things for around $40 a piece, and with a mining speed of around 4-8 khz on a good day it would take years for them to break even. That's assuming they don't burn out and that the difficulty rate/exchange rate coupling doesn't deteriorate below today's level.

Actually my original idea was to take one of these USB systems and throw a 10w solar panel on it, and then just leave it out on the lawn. Scale up as cash and lawn space permits.. :-) But it's just not feasible because of relatively high cost of the solar panel and voltage controller.

If we could plug a Block Eruptor into an Mk802 and use a solar panel to power the whole thing (with wifi being supplied by a central access point), it might be a fun and worthwhile experiment. But again, with eruptors costing hundreds of dollars each, the ROI is non existent. So it would be a mostly academic exercise.

(off topic: Who's willing to pay hundreds of dollars for these things? They will never, ever break even..)

- Michael

[runam0k]

(off topic: Who's willing to pay hundreds of dollars for these things? They will never, ever break even..)

Novelty factor 5.

Or buy a bunch more than you need, sell them on ebay and keep the rest for < 1BTC each.