log☇︎
884700+ entries in 0.638s
furuknap: Can't comply, sorry. Whether you acknowledge it or not, it still exists and affects pricing today.
ThickAsThieves: or at least wai til it's 1 yr away
furuknap: With the halving effect, AM is losing money for investors every week. With 30TH they wouldn't even beat NASDAQ composite.
furuknap: Well, to be nice to them, I used their purchased 62TH when calculating. With the 30TH, it would look very bad.
ThickAsThieves: but they are as big as any mining pool
ThickAsThieves: i do think they pulled some blades offline
furuknap: You're not unlucky just one day of the week, though.
furuknap: It's reverse calculated by looking at the blocks they solve, not the actual hashrate, so yes, to some extent. I wouldn't call it guess but it is based on luck.
Menoetius: isn't the publicly known AM hashrate just a guess based on luck ?
ThickAsThieves: random observation, AM farm hashing temporarily declines significantly on Monday for the past 3 weeks
ThickAsThieves: he probly needs to add each asset manually
parseval: I lost those logs you sent me, TAT
parseval: I need moar data though
ThickAsThieves: that's really the only thing that can stop the decline, that or massive resilient hoarding
ThickAsThieves: they have nothing better to do
ThickAsThieves: hard to stop them
ThickAsThieves: if it breaks that, then uhhhh
pankkake: so, how low bitcoin is going to fall? :)
ThickAsThieves: havent used them in mos
jurov: if one wants to consider mtgox at all
jurov: still, it's better to wait and see if they are really clearing the backlog
Chaang-Noi: that is good for them
ThickAsThieves: i know they re-enabled $ withdrawals
Chaang-Noi: gox quit trading?
Namworld: Didn't find the light color too annoying. It's not like most backgrounds on the web around light colored with dark text.
Namworld: Those headers are the darkest colors available, short of black, on Google Spreadsheets
Namworld: The tools are just to retrieve the data through the various APIs and insert it in Google Docs so eh =/
Namworld: You can choose whatever colors you want and what data/calcs to make if you make one =P
parseval: I'd be happy with getting early API access at this point :)
parseval: Namworld: That's pretty nice
ThickAsThieves: Nam, you should make the spreadsheets not inverted colors
ThickAsThieves: add things like Weex support or whatever
Namworld: Shameless insert of said tools: https://bitcointalk.org/index.php?topic=140427.0
ThickAsThieves: if you become the authority,
ThickAsThieves: then stick it to em
parseval: Oh, I like that
ThickAsThieves: become the authority
ThickAsThieves: get the userbase
parseval: I was thinking I'd leave it ad-free until I can start adding other stuff like escrow services or whatever
Namworld: I got a self-updating portfolio. But nothing too fancy.
parseval: I was thinking that too
ThickAsThieves: or if monetizing the service is the angle
parseval: The people over at Dragon's Tale might set up a PPC deal with me
parseval: Cool, thanks :)
ThickAsThieves: you know where to find me
parseval: Thanks. Oh, right... it was past the month, and I was redesigning so I left it out.. I'm still debating whether I want to try hitting you up for more btc ;)
parseval: Did you see, I have pie charts now, TAT
parseval: Such a nice spread last night too
ThickAsThieves: AM is paying everyone's bills these days
parseval: Over 8900 of that was AM
parseval: I think this year I'll be celebrating George Orwell's birthday
gribble: MtGox BTCUSD ticker | Best bid: 79.03200, Best ask: 79.03990, Bid-ask spread: 0.00790, Last trade: 79.03100, 24 hour volume: 109684.30615342, 24 hour low: 72.00000, 24 hour high: 86.26000, 24 hour vwap: 79.66696
Chaang-Noi: <thestringpuller> mpoe trades :P wow, i honestly assumed the bot had quit reporting it, or it closed. insane
thestringpuller: mpoe trades :P
Namworld: No, I guess people mostly trade Passthroughs and current holders are all keeping their stocks or something.
thestringpuller: don't think so
assbot: Last trade for S.MPOE on MPEX was at 0.00077573 BTC [-]
Chaang-Noi: iv not seen a ticker for mpex in forever, did it just close up?
Namworld: Beside, you'd want a dedicated processor if you wanted to do hashing exclusively, hence ASICs...
Namworld: AMD just happens to be somewhat good at hashing, but it's not something people would normally look for in their GPU cluster.
Namworld: and hence why they're used in clusters.
Namworld: I think it's due to more tools/support and better stability or something like that for Nvidia.
cads: Namworld: blind leading the blind, lol
cads: (disclaimer - take all this with a grain of salt - I'm speaking from very deep knowledge)
Namworld: I thought they had a pretty similar market share.
cads: not sure why AMD has been an underdog for so long - the situation's been similar between them and intel, with intel being more popular despite AMD offering, in my opinion, a better product
cads: while AMDs have tended to give more honest performance for the buck
Namworld: gamers seems to prefer the high end Nvidia cards too for some reason.
Namworld: No, it's just that Nvidia is slightly better for traditional GPU usage. AMD is more of consumer grade GPUs for playing games. The way they're made, they're also good for some other calculations like hashing.
cads: Namworld: after a simple search I don't find any companies that offer AMD GPU clusters, which stinks.
cads: I know for certain things like engineering simulations, consultants can provide huge markups on their work (which consists of, say, a Finite Element Analysis of a product), and it makes great sense to run in a High Performance Computing cloud.
Namworld: They don't rent a Nvidia at 1/10th of the cost it would be to run the AMD equivalent at home =P
Namworld: Economies of scale is all fine, but if the GPU you rent are 5-10 times slower than AMDs but cost the same to buy...
Namworld: If they offered the AMD equivalent for the same price they do their Nvidia cluster, might be worth it to rent.
cads: running in the cloud, you have to pay all of that, amortized by the service provider, and you also have to pay a profit for the provider
cads: but I also have doubts - running at home, they just have to pay capital costs, electricity, minimal maintenance, and depreciation
cads: I have a feeling that people running GPUs like mad might actually save time and perhaps even cash by running in the cloud
Namworld: Never bothered to get a better PSU to mine.
cads: I have to find out how competitive cloud based GPU clusters are
Namworld: It runs fine for gaming tho, which is all that matters.
cads: in any case I agree - even 102 core-year cost would not be too great.
Namworld: and I can't get vanityminer to run on my 6990 for some reason to try it.
Namworld: Yeah, I'm not sure at all... I remembered the difference between CPU/GPU being similar to hashing... but it's been a long while since I tested.
cads: has that changed?
cads: and not something that GPUs can do very well
cads: Namworld: I might be wrong by now, but at one point I learned that creating elliptic curve keys is altogether separate from the Sha hashing that GPUs excel at.
Namworld: Costs aren't too prohibitive to do that =P
Namworld: A high-end GPU can probably find 1Microsoft pattern in far less than 1 month.
Namworld: Just like CPU would get a few hundred thousand keys/hash per seconds, a high-end GPU can go a over a thousand time faster.
Namworld: I got a six core CPU and it goes on at 102 years estimated to find 1Microsoft pattern.
cads: so a company like Microsoft could do it without batting an eyelash, if they thought having a custom "1Microsoft31uEbMgunupShBVTewXjtqbBb" key
Namworld: You can generate private keys all you want, finding one which match just a short part of a public address takes so many.
cads: But hey, 15K core-years costs only about $6M on the Amazon Ec2 cloud computing network.
Namworld: There's a reason why it's considered hard/impossible to have a full collision while generating a private key for an existing full Bitcoin address.
cads: 15 thousand cpu-core-years to do it, lol.
cads: oh, I made a mistake -- multiply that number by 56.
cads: rather, it would take around that time for it to get a 50% chance of success
cads: it would take my poor laptop computer around 256 years to compute the vanity address starting with "Microsoft"
Namworld: It's not like generating more than 8 characters for a vanity address is easy anyway...
cads: That's a good practice - I do that for passwords where having a longer password beats doing a typo or transcription error