Bitcoin Forum

So, I've got the BC client running on my PC for about 2 weeks now.... with ZERO bitcoins to show for it.

The FAQ says that, after I get all of the blocks from the network, "Your average coin creation rate will be (6 * 50 coins / hour) * (your CPU speed / the total CPU speed in the Bitcoin network). At the moment (June 2010) it might be something like 50 coins / day, for example."

Well, I checked the wiki and my clock count matches the count they have there (69649). And the 50 coins / day is a figure that was given for just a month ago.

Right now, I'm doing about 400 khash/s, and I've yet to see a single coin.

Ideas?

The difficulty of generating coins has increased dramatically in the last month; keep waiting.

The difficulty for generating a block has gone up, it's currently at 181 when it was as low as 23 a few weeks ago. Check out this page for help in calculating your probability of generating a block:
http://www.alloscomp.com/bitcoin/calculator.php (http://www.alloscomp.com/bitcoin/calculator.php)

Keep in mind that block generation is a lottery, each hash you compute is a "ticket", so the faster you generate hashes the more chances you have to win the lottery. For instance, I have a desktop at home that can do about 1600 khash, I ran bitcoin for 3 hours and generated a block, however my desktop at work has been running bitcoin for days with no blocks generated.

There's some douche who's got 1000 cores crunching through hashes, so all of us are kind of left out to dry for now. =P

More incentive for us to create wealth generating services so we can get bitcoins from others.  ;D

Kiba: Indeed.  *Cracks whip*

Do we know who this is?  Academic?  Botnet?  NSA?  Chinese hackers?

What are they doing, other than generating most of the BTC?

Edit:  OK, I get that it's nenolod (http://nenolod.net/thoughts-on-bitcoin).  So, who's nenolod?

Edit2:  OK, I see that he's the technical director of SystemInPlace (http://www.systeminplace.net).

Unfortunately, they're a US-based company, or I'd consider using them.

http://bitcointalk.org/index.php?topic=431.0

I wouldn't.  The guy uses servers he's rented to other people to generate bitcoins.  If I have paid for a server I want all of the cores doing what I want.

Hmmm.  If you're right, he's taking a risk to disclose that here.  Not that I'm vindictive, or anything  ;)

He claims he isn't using his employers resources, but *shrug*.

Yeah, I saw that after I posted.  I was under the impression he owned the company, which would mean they are his resources.  Just having 1000 spare cores at home is kinda crazy, but I guess someone realy dedicated to computers might.

Or he might have been using Amazon Cloud services (Or microsofts, or IBM's for that matter).

Or, his employer might give him a bunch of virtual machines to work/experiment with (thus, 1000 cores isn't that big of a claim, considering you can have a bunch of low-quality cores).

*shrug* Who says he isn't some rich snob willing to waste money? =P

He claims (http://bitcointalk.org/index.php?topic=431.msg5141#msg5141) that it cost just $130 in expenses.  Let's see -- paying $130 to use 1,000 CPUs for 168 hours implies $0.0008 per CPU per hour.  That's a very good price, I believe  ;)

I'm either facing the losing side of the statistics, or doing it wrong.
It's been four days since I've seen fresh coin, and I'm running 7200 khash/sec (aggregate).

That puts me over 95% percentile.
I'll stay away from Vegas this weekend.

That's silly logic, it'll even out.

Yea, we still haven't gotten easier since nenold withdrew his 1000 cores.

That's because he never had them to begin with. Mainly because there are some members on this forum that have triple his claimed resources and they know how much he should be really generating if he did have that setup going for that long.

No, he had something.  Otherwise the difficulty wouldn't have increased so significantly.

Can any of the following time series be determined, or estimated, from block data?

(1) number of nodes
(2) total Khash/s for all nodes
(3) Khash/s per node

I still don't believe it though, mainly from a logistics standpoint. I won't argue if he has 85K BTC but you can go back in time on all the blocks here http://nullvoid.org/bitcoin/statistix.php and with a good spreadsheet and knowledge of how fast they are created (creation times are all there), he would have to created every *single* block (meaning not a single person was getting one) for 11 days straight 24/7

Further evidence is since he claims to have turned it off, the difficulty still has not changed. So either the interest in Bit Coin has grown to a level that the difficulty will naturally remain that high or someone else has filled in his place for block generation with another large server farm setup. 85K BTC is like sitting on over $4,200 USD which unless he had free server time, would have cost him nearly $33,000 USD for the 11 day sprint (with no competition from anyone else mind you) if he was using the Amazon Cloud Computing like he claimed.

And to top it off, say he did spend the $$$ to place a 1,000 core server farm at his fingertips, that's still small in comparison to the total amount of CPU out there generating Bit Coins, so he would be lucky to get 1 in every 10 blocks that way, further making it take over 110 days to generate that much BTC at a cost of nearly a half-million dollars of CPU time.

The numbers don't add up for his feat in my opinion unless he had a 100,000 core server farm setup for free. I would place my bets on that he was just blowing smoke.

We've still got 1060 something blocks to go before it readjusts difficulty.

Also, while I don't believe his claim on how MUCH BTC he got, I still believe he had the 1000 cores crunching on it.  Also, there's MAYBE 1000 cores generating BTC =/ The user base isn't THAT big yet.

So, did he generate ungodly amounts of BTC?  No.  Did he put enough CPU into the system to disrupt the difficulty process, which has yet to adjust back downward? Yes.

That's a third option I hadn't considered, thanks for pointing that out.

According to those statistics the difficulty should either increase or stay the same, but not decrease.

You've got something mixed up. If more than 6 blocks are completed per hour on average then difficulty increases and if less than 6 blocks are completed per hour on average then the difficulty decreases.

Oh, right, hadn't realized it was at 10/min.  Still, it's a whole lot slower than when I started generating, so yea, it's supposed to be this hard. =P

OK, we're at 9 blocks per hour now, and the target is 6.  Therefore, difficulty is about to increase by ~50% -- and NOT decrease  :(

FWIW, I did just get a block on a machine doing ~2,000 Khash/s, which implies ~0.1 block per day on average.  That's about right, I believe (not that I can conclude anything with N=1).

It's obvious that the swarm's aggregate CPU resources are still growing rapidly.  Perhaps nenolod, if s/he ever had 1,000 CPUs in the swarm, is just renting them to someone else.  Indeed, perhaps s/he's renting N-CPU packages  ;)  Or  ???

Anyway, there ought to be a way to know what's happening in the swarm.  We need to know if we're being attacked, right?  And we need mechanisms to defend the swarm, right?  I'm not talking about breaking anonymity.

swarm metrics (network metrics) would be quite useful.  Anything is better than watching an IRC forum full of bots.

I don't do IRC, and have never looked.

Is there a way to log the IRC channel (potential n00b question, I know)?  Do all nodes access it?  I recall seeing something about a flag for not doing so.