Bitcoin Forum

Hello Guys,

Can someone Tell me what is the exactly mathematical problem behind Bitcoin mining?
Is it possible to build a perfectly fitting Hardware to solve that problem?

best regards, XuS

Computing a SHA-256 hash that is numerically below a set threshold value (i.e. the first bits must be all zero). It's like randomly drawing 256 bits and hoping that the first N bits are all zero.

With each additional zero bit, the chances go down by 50%.

Christian

Imagine you have a hat with 100 pieces of paper in it, numbered 1 to 100. You pull out a piece of paper every minute and look at what you got (then put it back and shake up the hat). If it is lower than 20, you win, and you would win on average every five minutes. If you started checking numbers faster than every minute, I could slow down how often you win by making the highest winning number 15 instead of 20.

Bitcoin mining is kind of like that, but instead of 1 to 100 numbers, there are 1 to 1.1579E+77 possible numbers that you get when you take the hash of some data, and Bitcoin awards you 50 BTC if you find a hash of the current transaction block that is 1.7248E+61 or smaller.

A SHA hash is a complex mathematical formula that original data is put through, and the formula creates a number on the other side, like a 'signature' of the original data. Other hashes you might be familiar with in computers are MD5 or CRC. Since hashing the same transaction block over and over would always give you the same SHA hash, your computer adds some more random data to the end of a transaction block (called a nonce), to change the hash that comes out. SHA is cryptographically secure, in that it is impossible to tell what the hash will be from the nonce you add, so there is no shortcut around just trying billions of different nonces and checking the hash that is generated.

so if someone would implement a SHA-Chip in Harware he woul habe incrediable performance to mine bitcoins?

best regards, XuS

Yes, exactly. There's a handful of people already trying to do exactly this. ArtForz is rumoured to have an ASIC design, and there's some people in the Mining forums working roughly in that direction by making an FPGA-based mining device.

The future of bitcoin's going to be real interesting with things like this!

Yes, I think that will be the future. But don't forget that network adapts itself. So if everyone has specific hardware, no one will have outstanding profits because the difficulty will increase.

Guys, let me get it straight. For mining ( allocating ur resources ), u get bitcoins. The resources are basically used to check the timestamping of the transactions so as to avoid double transactions. The number of coins are limited to 21 mil. Once the entire amount of coins are released, how is the system going to monitor transactions?

What am i missing? What did i wrongly comprehend?  ??? ??? ???

Was it really necessary to bump a thread from over two years ago?

As to what you are missing.  You are missing the fact that the "miners" are compensated in two ways.  They receive a "block subsidy" (currently 25 BTC) of newly minted bitcoins for every block they solve, and they also receive the transaction fees of every transaction that they include in their block.

Over time, the block subsidy will shrink (it gets cut in half approximately every 4 years), and as bitcoin gains popularity, the value of the transaction fees will increase.  Eventually (long before the subsidy ends), miners will receive a larger portion of their compensation from fees than from the subsidy.  Perhaps then we'll stop calling them miners, and just call them "transaction processors" instead.  It doesn't really matter what we call them, they'll still be performing the exact same service and they'll continue to be compensated for that.

i want get out from the newbie section...

First up, thanks mate for ur response. I wdnt have bothered if it wasnt necessary. Much appreciated.

Then you shouldn't have bothered.  This question has been answered thousands of times.  You should have been able to find the answer by searching.  If you couldn't then you could have created a new post.  It was not "necessary" to bump a 2 year old thread that was unrelated to your question.