Looks like the stress test is ending, hashrate has dropped to 1.871 TH/s as of right now
19 blocks wasn't a bad effort for a first test run.
Topic: ASICMINER: Entering the Future of ASIC Mining by Inventing It - page 1266. (Read 3917543 times)
February 14, 2013, 08:32:08 PM
19 blocks wasn't a bad effort for a first test run.
February 14, 2013, 08:28:53 PM
Looks like the stress test is ending, hashrate has dropped to 1.871 TH/s as of right now
February 14, 2013, 07:47:40 PM
A rudimentary script to help you guys cheer on ASICMINER:
Thanks for this, although having an issue with it (and no, I don't know what I'm doing)
made file asic.rb and copy/pasted your script..
installed ruby (sudo apt-get install ruby)
chmod +x asic.rb
then ./asic.rb ; it's clearly running and using ~25% of my cpu, but just sits there with a blinking cursor.. what am I doing wrong if you don't mind helping out a newb
Hmm, not too sure. I think it might be throwing an error every time before the "sleep 10" hits, so it just goes into an infinite loop. In that case, it's most likely a problem with the network connection to btcguild... I'd have to know a few more details to help more.
edit: perhaps replace the rescue part?
rescue Exception => e
puts e.message
puts "Backtrace:"
puts e.backtrace.inspect
next
end
February 14, 2013, 07:24:21 PM
Care to explain?
February 14, 2013, 07:22:16 PM
The probability distribution of finding a block is described by the Poisson distribution, not binomial distribution, right? How that changes these numbers I have no idea 
.
.Why do you think so? Every block competed for is a binary event with 1/6 chance of success and 5/6 chance of failure. Furthermore every race for a block is independent. That's classical binomial right? Like a runs of flipping a coin (but flipping a coin has success chance of 1/2 instead of 1/6, so more like running a dice
)Convince me I'm wrong and I'll rerun the numbers with Poisson distribution but off to bed now
Its Poisson.
February 14, 2013, 07:19:07 PM
The probability distribution of finding a block is described by the Poisson distribution, not binomial distribution, right? How that changes these numbers I have no idea .
.Why do you think so? Every block competed for is a binary event with 1/6 chance of success and 5/6 chance of failure. Furthermore every race for a block is independent. That's classical binomial right? Like a runs of flipping a coin (but flipping a coin has success chance of 1/2 instead of 1/6, so more like running a dice
)Convince me I'm wrong and I'll rerun the numbers with Poisson distribution but off to bed now
February 14, 2013, 07:13:03 PM
A rudimentary script to help you guys cheer on ASICMINER:
Thanks for this, although having an issue with it (and no, I don't know what I'm doing)
made file asic.rb and copy/pasted your script..
installed ruby (sudo apt-get install ruby)
chmod +x asic.rb
then ./asic.rb ; it's clearly running and using ~25% of my cpu, but just sits there with a blinking cursor.. what am I doing wrong if you don't mind helping out a newb
February 14, 2013, 07:11:46 PM
A rudimentary script to help you guys cheer on ASICMINER:
Code:
#!/usr/bin/ruby
require 'net/http'
require 'date'
current_hash_rate = 0.00
max_hash_rate = 0.00
loop do
begin
url = URI.parse('https://www.btcguild.com//halloffame.php')
http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true
request = Net::HTTP::Get.new(url.path)
response = http.start {|http| http.request(request)}
case response
when Net::HTTPSuccess then source = response.body
else
puts "Error in HTTP connection"
next
end
match_string = '67117 ' '
source = source[(source.index(match_string) + match_string.length)..-1]
match_string = ' GH/s
source = source[0...(source.index(match_string))]
source.gsub! ',', ''
hash_rate = source.to_f
if hash_rate != current_hash_rate
max_hash_rate = [max_hash_rate, hash_rate].max
current_hash_rate = hash_rate
puts "#{DateTime.now}:\t current #{"%.2f" % current_hash_rate}\t max #{"%.2f" % max_hash_rate}"
end
sleep 10.0
rescue
next
end
end
Awesome, thanks! Yes I have been cheering them on too
February 14, 2013, 07:06:24 PM
A rudimentary script to help you guys cheer on ASICMINER:
Code:
#!/usr/bin/ruby
require 'net/http'
require 'date'
current_hash_rate = 0.00
max_hash_rate = 0.00
loop do
begin
url = URI.parse('https://www.btcguild.com//halloffame.php')
http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true
request = Net::HTTP::Get.new(url.path)
response = http.start {|http| http.request(request)}
case response
when Net::HTTPSuccess then source = response.body
else
puts "Error in HTTP connection"
next
end
match_string = '67117 ' '
source = source[(source.index(match_string) + match_string.length)..-1]
match_string = ' GH/s
source = source[0...(source.index(match_string))]
source.gsub! ',', ''
hash_rate = source.to_f
if hash_rate != current_hash_rate
max_hash_rate = [max_hash_rate, hash_rate].max
current_hash_rate = hash_rate
puts "#{DateTime.now}:\t current #{"%.2f" % current_hash_rate}\t max #{"%.2f" % max_hash_rate}"
end
sleep 10.0
rescue
next
end
end
February 14, 2013, 07:02:35 PM
Doesnt ASICMINER have enough hashpower to solo mine for awhile ?
sure does but i can see why they don't want to take the variance risk in the short term.. we have maybe 2-3 weeks here unless BFL blows up
When you control 1/6-1/3 of the network there is no such thing as variance (well technically there is but you get what I mean).
there sure is when we are talking about a couple of weeks, no such thing in the long term i agree, but we need guaranteed revenue here... though 2 weeks of great luck on solo would be awesome, bad luck would be very unfortunate.
That's utter nonsense.
Say "a couple of weeks" means 2 and ASICMINER only achieves 1/6th of the network (the two pessimistic assumptions in your range). Then:
In 2 weeks approximately: 14*24*6=2016 bocks will be mined and for each block we will have a chance of 1/6th, so we are expected to mine 336 blocks on average.
The chance to mine at least 310 blocks is: 1- binomcdf(2016, 1/6, 310) = 93.7 %
The chance to mine at least 300 blocks is: 1- binomcdf(2016, 1/6, 300) = 98.4 %
The chance to mine at least 286 blocks is: 1- binomcdf(2016, 1/6, 286) = 99.9 %
If the chance of success is large enough variance is hardly the bitch we all know her to be
. February 14, 2013, 06:58:45 PM
Am I the only one refreshing the BTC Guild page and thinking "go little ASIC, go"? I'm literally cheering on a machine. There is no hope for me.
February 14, 2013, 06:38:52 PM
Hm, when i see the stats at the pool i see that the rate goes down some time and goes back higher some time. It looks to me like they put a board in, it doesnt work fully and they take it out to repair it or something. It would be interesting to know the real hashrate, not the average only.
I wonder if they really work on one board only at a time. Because if they do i wonder if it wouldnt be better to set up all boards first and put it into because most of the chips will work. Then start taking out single boards to repair some errors there.
But maybe i only misunderstand the stats. Maybe they dont take out a board and its only a variance in the hashrate that cant be changed. So that they work on one board at a time to put it in later.
At the end i think they will have the best plan to do this beforehand. Its only a bit strange to watch the stats.
I wonder if they really work on one board only at a time. Because if they do i wonder if it wouldnt be better to set up all boards first and put it into because most of the chips will work. Then start taking out single boards to repair some errors there.
But maybe i only misunderstand the stats. Maybe they dont take out a board and its only a variance in the hashrate that cant be changed. So that they work on one board at a time to put it in later.
At the end i think they will have the best plan to do this beforehand. Its only a bit strange to watch the stats.
While I do believe they are tweaking things on their end, you're also going to see hash rate variance. Since they are mining at diff=1024 currently, you see a few % migration + or - over the course of an hour. The variable difficulty target rate at BTC Guild normalizes to a less than 1% variance when measured for a 24 hour period for any miner of 1.2 GH/s or higher.
February 14, 2013, 06:35:40 PM
Im surprised no one has setup a passthrough yet. I know there is a trading platform coming but in the meantime a pure passthrough might be the way to go.
Do their rules allow it? They seem pretty intent on doing everything possible to keep it as a legitimate investment and avoid the kinds of cluster-fucks we've seen in the past due to sketchy intermediaries. You can almost guarantee that if they allow anyone and everyone to operate a pass-through, someone will eventually scam people who invest in the pass-throughs.
February 14, 2013, 06:30:23 PM
Hm, when i see the stats at the pool i see that the rate goes down some time and goes back higher some time. It looks to me like they put a board in, it doesnt work fully and they take it out to repair it or something. It would be interesting to know the real hashrate, not the average only.
I wonder if they really work on one board only at a time. Because if they do i wonder if it wouldnt be better to set up all boards first and put it into because most of the chips will work. Then start taking out single boards to repair some errors there.
But maybe i only misunderstand the stats. Maybe they dont take out a board and its only a variance in the hashrate that cant be changed. So that they work on one board at a time to put it in later.
At the end i think they will have the best plan to do this beforehand. Its only a bit strange to watch the stats.
I wonder if they really work on one board only at a time. Because if they do i wonder if it wouldnt be better to set up all boards first and put it into because most of the chips will work. Then start taking out single boards to repair some errors there.
But maybe i only misunderstand the stats. Maybe they dont take out a board and its only a variance in the hashrate that cant be changed. So that they work on one board at a time to put it in later.
At the end i think they will have the best plan to do this beforehand. Its only a bit strange to watch the stats.
February 14, 2013, 06:29:08 PM
Im surprised no one has setup a passthrough yet. I know there is a trading platform coming but in the meantime a pure passthrough might be the way to go.
February 14, 2013, 06:21:58 PM
sadly one of the few glbse assets i did not hold at least a few shares 
You can still get exposure to ASICMINER via mining bonds/funds holding their shares.
I'm buying JAH for this reason.
https://bitfunder.com/asset/JAH
Anybody know about others, besides the [closing] GBF?
DMC
Yeah, each DMC share has a bit under 1/10th of an ASICMINER backing it. There's some other assets owned by DMC as well but the rest aren't worth much (BTC-Mining which gave its funds away to the scammer AmazingRando and BTCMC which has some assets but is being operated in a very dodgy manner and now claims a lot of its cash was in Bitcoinica despite never having mentioned that before). Picked up my DMC (for my fund) at about .02 each (meaning profit if ASICMINER worth over .25 - even if all other DMC assets are totally worthless which isn't quite the case) - think cheapest now were about .03 or .035 last time I looked (so you'd need to value ASICMINER over .4 or assign some value to the other DMC holdings to make them worth buying).
February 14, 2013, 06:21:01 PM
[FUD] I hope they distribute mining clusters elsewhere so that its not all located in one place where the US gov can perform a drone strike
I don't think even the US is ballsy enough to go all drone striking places in China.
Thats why they hire mercenaries within china. Or send in some criminals to do it. The Dirty Dozen is my favourite movie
Or 007 from MI5
February 14, 2013, 06:15:25 PM
[FUD] I hope they distribute mining clusters elsewhere so that its not all located in one place where the US gov can perform a drone strike
I don't think even the US is ballsy enough to go all drone striking places in China.
February 14, 2013, 06:12:08 PM
[FUD] I hope they distribute mining clusters elsewhere so that its not all located in one place where the US gov can perform a drone strike
February 14, 2013, 06:05:52 PM
Doesnt ASICMINER have enough hashpower to solo mine for awhile ?
sure does but i can see why they don't want to take the variance risk in the short term.. we have maybe 2-3 weeks here unless BFL blows up
When you control 1/6-1/3 of the network there is no such thing as variance (well technically there is but you get what I mean).
there sure is when we are talking about a couple of weeks, no such thing in the long term i agree, but we need guaranteed revenue here... though 2 weeks of great luck on solo would be awesome, bad luck would be very unfortunate.
That's utter nonsense.
Say "a couple of weeks" means 2 and ASICMINER only achieves 1/6th of the network (the two pessimistic assumptions in your range). Then:
In 2 weeks approximately: 14*24*6=2016 bocks will be mined and for each block we will have a chance of 1/6th, so we are expected to mine 336 blocks on average.
The chance to mine at least 310 blocks is: 1- binomcdf(2016, 1/6, 310) = 93.7 %
The chance to mine at least 300 blocks is: 1- binomcdf(2016, 1/6, 300) = 98.4 %
The chance to mine at least 286 blocks is: 1- binomcdf(2016, 1/6, 286) = 99.9 %
If the chance of success is large enough variance is hardly the bitch we all know her to be
Jump to: