dreamspark
|
|
May 16, 2014, 09:50:35 AM |
|
Yes, I can't prove that, I didn't see any proofs. But I don't think that he lied us. There is no reasons.
There are a few reasons I can think of; Casting doubt over whether the devs have released the best miner they can. Discouraging others from mining as there's no point if you have a 4 times faster miner. I could go on, Im not saying he was lying but saying there are no reasons isn't true.
|
|
|
|
lasybear
Member
Offline
Activity: 113
Merit: 14
|
|
May 16, 2014, 09:56:18 AM |
|
Evidence?
I posted it earlier: Yes, I remember that. Some person on the Internet saying that some other unnamed person said he did something hardly constitutes evidence. I'm not even doubting that optimized asm code could make a big difference. Just not sure how to know whether this is real or not. Rumors and FUD are rampant, so it is just hard to tell. You don't see a lot of evidence for it in the hash rate. Currently a 20 H/sec computer mines about 2 coins a day, worth about $1, and uses very roughly 30c of electricity (15c/kwh rate). So it is still quite profitable to mine this coin using a regular computer and the standard miner. It does require extreme patience though, with no pools available yet. Yes, I can't prove that, I didn't see any proofs. But I don't think that he lied us. There is no reasons. i5 [2014-05-16 13:54:57] accepted: 48/48 (100.00%), 137.72 hashes/s (yay!!!) [2014-05-16 13:54:57] thread 2: 7 hashes, 21.35 hashes/s [2014-05-16 13:54:58] accepted: 49/49 (100.00%), 121.27 hashes/s (yay!!!) [2014-05-16 13:54:58] thread 2: 7 hashes, 21.49 hashes/s [2014-05-16 13:54:58] thread 0: 40 hashes, 26.60 hashes/s [2014-05-16 13:54:59] thread 1: 88 hashes, 30.90 hashes/s [2014-05-16 13:54:59] accepted: 50/50 (100.00%), 110.51 hashes/s (yay!!!) [2014-05-16 13:54:59] thread 2: 7 hashes, 36.26 hashes/s [2014-05-16 13:55:00] accepted: 51/51 (100.00%), 125.27 hashes/s (yay!!!) [2014-05-16 13:55:00] accepted: 52/52 (100.00%), 125.27 hashes/s (yay!!!) [2014-05-16 13:55:01] accepted: 53/53 (100.00%), 125.27 hashes/s (yay!!!) [2014-05-16 13:55:01] thread 2: 7 hashes, 21.45 hashes/s [2014-05-16 13:55:01] accepted: 54/54 (100.00%), 110.46 hashes/s (yay!!!) [2014-05-16 13:55:02] thread 3: 159 hashes, 32.07 hashes/s [2014-05-16 13:55:02] thread 0: 40 hashes, 30.40 hashes/s [2014-05-16 13:55:02] thread 2: 7 hashes, 24.43 hashes/s [2014-05-16 13:55:02] accepted: 55/55 (100.00%), 117.80 hashes/s (yay!!!) [2014-05-16 13:55:03] accepted: 56/56 (100.00%), 117.80 hashes/s (yay!!!) [2014-05-16 13:55:03] thread 2: 7 hashes, 21.39 hashes/s [2014-05-16 13:55:04] thread 0: 40 hashes, 27.27 hashes/s [2014-05-16 13:55:04] thread 1: 88 hashes, 28.45 hashes/s [2014-05-16 13:55:04] accepted: 57/57 (100.00%), 109.18 hashes/s (yay!!!) [2014-05-16 13:55:05] accepted: 58/58 (100.00%), 109.18 hashes/s (yay!!!) [2014-05-16 13:55:05] thread 2: 43 hashes, 37.23 hashes/s [2014-05-16 13:55:05] accepted: 59/59 (100.00%), 125.02 hashes/s (yay!!!) [2014-05-16 13:55:06] accepted: 60/60 (100.00%), 125.02 hashes/s (yay!!!) [2014-05-16 13:55:06] thread 3: 162 hashes, 33.04 hashes/s [2014-05-16 13:55:08] thread 3: 23 hashes, 21.37 hashes/s [2014-05-16 13:55:08] thread 2: 43 hashes, 23.45 hashes/s [2014-05-16 13:55:08] accepted: 61/61 (100.00%), 100.53 hashes/s (yay!!!) [2014-05-16 13:55:09] accepted: 62/62 (100.00%), 100.53 hashes/s (yay!!!) [2014-05-16 13:55:09] thread 3: 23 hashes, 27.27 hashes/s [2014-05-16 13:55:10] accepted: 63/63 (100.00%), 106.44 hashes/s (yay!!!) e5 [2014-05-16 13:55:40] thread 9: 11 hashes, 14.14 hashes/s [2014-05-16 13:55:41] accepted: 83/83 (100.00%), 196.97 hashes/s (yay!!!) [2014-05-16 13:55:41] accepted: 84/84 (100.00%), 196.97 hashes/s (yay!!!) [2014-05-16 13:55:41] thread 4: 13 hashes, 15.87 hashes/s [2014-05-16 13:55:42] accepted: 85/85 (100.00%), 200.53 hashes/s (yay!!!) [2014-05-16 13:55:42] thread 8: 66 hashes, 14.64 hashes/s [2014-05-16 13:55:43] thread 10: 86 hashes, 17.07 hashes/s [2014-05-16 13:55:43] thread 11: 91 hashes, 17.54 hashes/s [2014-05-16 13:55:43] accepted: 86/86 (100.00%), 198.35 hashes/s (yay!!!) [2014-05-16 13:55:43] thread 3: 93 hashes, 17.88 hashes/s [2014-05-16 13:55:43] thread 2: 81 hashes, 15.38 hashes/s [2014-05-16 13:55:43] thread 7: 89 hashes, 16.61 hashes/s [2014-05-16 13:55:43] thread 0: 85 hashes, 15.82 hashes/s [2014-05-16 13:55:43] accepted: 87/87 (100.00%), 195.78 hashes/s (yay!!!) [2014-05-16 13:55:43] thread 0: 2 hashes, 13.63 hashes/s [2014-05-16 13:55:43] thread 6: 89 hashes, 15.28 hashes/s [2014-05-16 13:55:44] accepted: 88/88 (100.00%), 191.45 hashes/s (yay!!!) [2014-05-16 13:55:44] thread 0: 2 hashes, 13.00 hashes/s [2014-05-16 13:55:44] thread 1: 35 hashes, 16.89 hashes/s [2014-05-16 13:55:45] accepted: 89/89 (100.00%), 189.38 hashes/s (yay!!!) [2014-05-16 13:55:45] thread 0: 2 hashes, 12.28 hashes/s [2014-05-16 13:55:45] thread 3: 22 hashes, 12.19 hashes/s [2014-05-16 13:55:45] accepted: 90/90 (100.00%), 182.96 hashes/s (yay!!!) [2014-05-16 13:55:46] accepted: 91/91 (100.00%), 182.96 hashes/s (yay!!!) [2014-05-16 13:55:47] accepted: 92/92 (100.00%), 182.96 hashes/s (yay!!!) [2014-05-16 13:55:47] thread 0: 2 hashes, 12.71 hashes/s [2014-05-16 13:55:48] thread 11: 64 hashes, 14.06 hashes/s [2014-05-16 13:55:48] thread 5: 77 hashes, 15.58 hashes/s [2014-05-16 13:55:48] accepted: 93/93 (100.00%), 180.41 hashes/s (yay!!!) [2014-05-16 13:55:48] thread 0: 2 hashes, 10.20 hashes/s
|
|
|
|
dreamspark
|
|
May 16, 2014, 09:58:56 AM |
|
^^^ again thats not evidence.
You don't know what coins or processor is doing that mining. Not saying its not true just that that's not real evidence.
|
|
|
|
Ivan166
|
|
May 16, 2014, 09:59:45 AM |
|
Russian thread updated. Does multi language support in monero.cc work properly? I don't think so. Earlier I did Russian translation for main Monero website.
|
|
|
|
lasybear
Member
Offline
Activity: 113
Merit: 14
|
|
May 16, 2014, 10:02:25 AM |
|
^^^ again thats not evidence.
You don't know what coins or processor is doing that mining. Not saying its not true just that that's not real evidence.
Ok. What are you can accept as evidence? Other words, what you want to see?
|
|
|
|
smooth
Legendary
Offline
Activity: 2968
Merit: 1198
|
|
May 16, 2014, 10:04:35 AM |
|
^^^ again thats not evidence.
You don't know what coins or processor is doing that mining. Not saying its not true just that that's not real evidence.
Ok. What are you can accept as evidence? Other words, what you want to see? Source code would be good (and useful!). Alternately, I guess, you could have a trusted person test it.
|
|
|
|
dreamspark
|
|
May 16, 2014, 10:10:18 AM |
|
^^^ again thats not evidence.
You don't know what coins or processor is doing that mining. Not saying its not true just that that's not real evidence.
Ok. What are you can accept as evidence? Other words, what you want to see? A screen shot of simpleminer if its an optimized version would suffice next to screenshot of the system screen showing the CPU. If its a complete separate miner a screen shot of that miner including the command used to start it which should show an MRO address. Failing that, the source code will do
|
|
|
|
lasybear
Member
Offline
Activity: 113
Merit: 14
|
|
May 16, 2014, 10:10:49 AM |
|
^^^ again thats not evidence.
You don't know what coins or processor is doing that mining. Not saying its not true just that that's not real evidence.
Ok. What are you can accept as evidence? Other words, what you want to see? Source code would be good (and useful!). Alternately, I guess, you could have a trusted person test it. Heh ) How are you think, how high difficulty will grow up after? BTW my source additions are too ugly to show others - I'm not a pro ) But it works )
|
|
|
|
eizh
|
|
May 16, 2014, 10:26:47 AM |
|
I agree on the implementation part. They'll have to take complex, obfuscated code and try to put it in without breaking anything. They'll likely need someone who's damn good at higher math, if not a cryptographer.
Yes, I expect this should restrict adoption in Bitcoin-based alts for a significant time to come. CryptoNote was a genuine research effort with professional cryptographers involved. The result is actually a modified version of what exists in academic literature rather than a direct implementation (in order to block double spends). Impressive stuff. Easy to promise, but delivery is going to be tough. I do hope someone tries, though. Maybe we'll actually get something well-documented and understandable.
|
|
|
|
lasybear
Member
Offline
Activity: 113
Merit: 14
|
|
May 16, 2014, 10:27:09 AM |
|
^^^ again thats not evidence.
You don't know what coins or processor is doing that mining. Not saying its not true just that that's not real evidence.
Ok. What are you can accept as evidence? Other words, what you want to see? A screen shot of simpleminer if its an optimized version would suffice next to screenshot of the system screen showing the CPU. If its a complete separate miner a screen shot of that miner including the command used to start it which should show an MRO address. Failing that, the source code will do Sorry, very big... One i5 and One e5 connected to local pool:
|
|
|
|
dreamspark
|
|
May 16, 2014, 10:28:09 AM |
|
Finally, someone who knows their asshole from their elbow when it comes to botnet administration. However, you are wrong about one thing: If I write my miner correctly, the average user will never know it's there.
Hypothetically, were I to write a silent miner for a botnet, I would make it use far less CPU when the user is on it. Maybe 25% to 50% of the CPU. This is not obvious to the user, and when the system goes idle, I simply kick the miner into full gear. You might say Windows defaults cause the system to suspend/hibernate when idle, but I can easily change that from my code, and 99% of users would never think anything of it.
As this author, I know the idea is to keep as many bots as I can for as long as I can. Therefore, I will not abuse my bots by running a miner at 100% on them at any point where I think the user would notice. If someone has a slow computer, they will do a malware scan or just outright reinstall - and if they can't, they will find someone who can. This is why you keep the users happy.
So, correctly implemented mining malware not very easily noticed by most users, therefore, it's easy money.
P.S. About the DDoS thing, most users won't notice the DDoS itself, but the consequences may hit the user right in the face - people being constantly harassed by way of a DDoS have a tendency to start sending notifications of abuse to the ISPs of the bots. If the ISP contacts the user because they keep getting abuse emails, that's something the user would have a hard time ignoring if they tried.
Some very good points. In regards to the silent miner, I think its interesting to speculate what sort of PC's are infected and the hardware that they are likely to be running. From what I can ascertain the majority of botnets are comprised of low hardware PC's. This brings up a number of considerations. One being the actual HR you can get from 25% of a pentium 4 for example. I agree that if you have 1000's of PC's its still likely to = a high HR but not too many botnets particularly private ones that aren't rented out for bigger purposes are that large. Another being that if they are running low powered machines, 25-50% of cpu usage is still likely to make other simple tasks mundanely slow. In addition just because the PC is idle doesn't mean you can't see or hear it being hot or loud.
|
|
|
|
smooth
Legendary
Offline
Activity: 2968
Merit: 1198
|
|
May 16, 2014, 10:30:59 AM |
|
Hypothetically, were I to write a silent miner for a botnet, I would make it use far less CPU when the user is on it. Maybe 25% to 50% of the CPU. This is not obvious to the user, and when the system goes idle, I simply kick the miner into full gear. You might say Windows defaults cause the system to suspend/hibernate when idle, but I can easily change that from my code, and 99% of users would never think anything of it.
If you kick up to full gear, the computer will run loud and hot (especially laptops) and definitely be noticeable. Granny will ask her 12 year old grandkid computer expect why her computer is always making a ton of noise when she's not using it, and the kid will figure it out. This is especially true for the typical consumer grade (not enthusiast) PC that is not thermally optimized. In fact even 25% may be noticeable, especially if such botnets become widespread and people are warned to look for precisely that. Operating systems may adapt too. The new Mac OS has something called App Nap that puts applications to sleep when they don't have a visible window. Apps can request to not be put to sleep and this is currently invisible, but widespread mining bot nets would motivate a feature where any app programmatically consuming CPU must have an icon in the menu bar or some similar visible indicator. Something similar already exists in some mobile OSs. Everyone and everything adapts. You can't assume that bot nets do X, Y, or Z on a widespread basis and nothing else changes. You have to game it out to see what actually might happen, and even that is usually wrong.
|
|
|
|
dreamspark
|
|
May 16, 2014, 10:31:38 AM |
|
^^^ again thats not evidence.
You don't know what coins or processor is doing that mining. Not saying its not true just that that's not real evidence.
Ok. What are you can accept as evidence? Other words, what you want to see? A screen shot of simpleminer if its an optimized version would suffice next to screenshot of the system screen showing the CPU. If its a complete separate miner a screen shot of that miner including the command used to start it which should show an MRO address. Failing that, the source code will do Sorry, very big... Thats more like it Are you mining to your own pool?
|
|
|
|
smooth
Legendary
Offline
Activity: 2968
Merit: 1198
|
|
May 16, 2014, 10:33:46 AM |
|
lazybear are you interested in a bounty to release the source code (maybe cleaned up a bit?) your optimized miner?
If not, I'll probably play around with the code myself tomorrow and see if I can come up with something, or maybe Noodle Doodle will take an interest.
|
|
|
|
Buratino
Legendary
Offline
Activity: 1151
Merit: 1003
|
|
May 16, 2014, 10:34:11 AM Last edit: May 16, 2014, 10:46:32 AM by Buratino |
|
Hehe, seems lasybear is that "someone C++ skilled". Пpивeт, пoдeлиcь cвoими нapaбoткaми cooбщecтвy, дyмaю, зa этo тeбe пoлaгaeтcя пpиз. Я cкинycь нa бayнти, ecли ты coглaceн. [Hi, contribute please your work to community, i think you should be rewarded. I also will send donation to you, if you agree.]
|
|
|
|
dreamspark
|
|
May 16, 2014, 10:41:52 AM |
|
I do have to question some peoples thought process when they claim that a GPU coin is much much fairer than a CPU coin. Do they not think that there are huge GPU farms that rape the GPU coins as well. The main combatant is pools which allow the big farms to solo mine and then everyone else to mine on pools.
Come on, think about why that reasoning is faulty for a minute. No, I'm serious. Think about it before you read the rest of this post. Let's say, for the purposes of this discussion, that most regular people have 2 CPUs working for them. Maybe they're high end, maybe low end, but let's just look at the number of CPUs at this point. Regular miner has his 2 CPUs, now botnet miner likely has tens or hundreds of thousands. Really good bot herders can even break the 1M bots mark. You can see that even if we consider the number of bots the botner miner has as half of the total - to account for shut down PCs, loss of bots, internet connection issues on the bots, etc.- and even to tip it more in the home user's favor, let's say he has 2 of the latest i7 Extreme Editions, or hell, even high end Xeons. Even if those bots are all rocking low-end Pentiums and Semprons, the botnet miner will kill the average users by a LARGE margin. Now, why won't this happen with GPUs? Well, for starters, botnet mining on GPUs is not advisable. Very few bots will have decent GPUs, even fewer will have high-end ones, and the user will likely notice the fan going at 100% if he's just in the room. So even if you disable it when he's on the system, his chances of noticing something is amiss are too damned high. Okay, but what about rich GPU farmers? Well, GPUs cost a lot. Not per unit, but to maintain. The more of them you have, you start having big problems. Insane, ungodly amounts of heat. Ridiculous power bills. And then you need someone who is very skilled with computers - and mining, too - to maintain the farm, because God knows GPUs that are left overclocked and running at 100% for weeks/months at a time are not stable. In short, GPUs do not scale well. A few are easy to manage, try too many, and it will cost you a fortune. As an aside, this is why ASICs are awful: They are EXTREMELY expensive compared to GPUs, locking out the average person, they entail greater risk, as they cannot be resold if their mining usefulness ends before you thought it would, and they are plug and play, so you can hire unskilled labor to keep them running - and you'll need less labor, because they are far more stable. Again good points I can't really refute many of them as at the end of the day they are just guess work. Particularly in regards to the hardware that they are running and the size of the botnets. Think about why some of that reasoning is faulty as well though. I think the most important thing to constantly remember is your talking about bot admins with 1000's or 10,000's of bots. There are hundreds of things more profitable with a bot net near that size than mining monero Look at how much coin you could feasibly have sold, there is no serious liquidity in the market. Or are you suggesting that a bot admin would hold random crypto hoping it goes up when he could use is 1m bots to make some real money?
|
|
|
|
eizh
|
|
May 16, 2014, 10:44:29 AM |
|
lazybear are you interested in a bounty to release the source code (maybe cleaned up a bit?) your optimized miner?
If not, I'll probably play around with the code myself tomorrow and see if I can come up with something, or maybe Noodle Doodle will take an interest.
smooth, NoodleDoodle just said on IRC his latest optimizations are 4x faster on Windows. Untested on Linux so far but he'll push the source to the git repo soon. We'll be at 1 million network hashrate pretty soon.
|
|
|
|
smooth
Legendary
Offline
Activity: 2968
Merit: 1198
|
|
May 16, 2014, 10:46:00 AM |
|
Let's say, for the purposes of this discussion, that most regular people have 2 CPUs working for them. Maybe they're high end, maybe low end, but let's just look at the number of CPUs at this point. Regular miner has his 2 CPUs, now botnet miner likely has tens or hundreds of thousands. Really good bot herders can even break the 1M bots mark. You can see that even if we consider the number of bots the botner miner has as half of the total - to account for shut down PCs, loss of bots, internet connection issues on the bots, etc.- and even to tip it more in the home user's favor, let's say he has 2 of the latest i7 Extreme Editions, or hell, even high end Xeons. Even if those bots are all rocking low-end Pentiums and Semprons, the botnet miner will kill the average users by a LARGE margin.
You misunderstand stand the point of egalitarian mining. Yes, someone with more CPUs will mine more than someone with fewer CPUs. That is obvious. But as long as it stays linear the return on investment is mostly constant, in fact the PC user has an advantage that is very similar to a botnet owner: his PC is already paid for, he didn't buy it for mining, so his capital cost of mining is essentially zero. You can't expect to be miner extraordinaire with a PC but with linear scaling, you get your share. You can get a few coins out of it every now and then, maybe make a little bit over the cost of electricity (which only costs around 30c/day). At this scale its a hobby or even less. But there is at this point no reason to turn off the wallet miner, you might as well accept the free coins. It is only once the relationship becomes non-linear (where buying specific GPUs or worse ASICs means high capital costs but massively higher hash rates than regular PCs) that this sort of casual mining is useless and it does make sense to turn off the wallet miner, and then everything becomes concentrated. BTW, GPUs are definitely scalable, though not necessarily the way hobbyists do it.. Dga did his nv GPU mining using AWS GPU instances. Hundreds of them I think. As an aside, this is why ASICs are awful: They are EXTREMELY expensive compared to GPUs, locking out the average person, they entail greater risk, as they cannot be resold if their mining usefulness ends before you thought it would, and they are plug and play, so you can hire unskilled labor to keep them running - and you'll need less labor, because they are far more stable.
Yes, that's right. ASICs are at one extreme, CPUs are at the other. GPUs are somewhere in between. If CPU mining manages to survive on this coin we will get something that is pretty much the opposite of what ASICs did. We'll see if that happens.
|
|
|
|
smooth
Legendary
Offline
Activity: 2968
Merit: 1198
|
|
May 16, 2014, 10:49:56 AM |
|
You think the average person will recognize the very slight increase in noise that would come from a CPU getting a little warm and the fan speed going up? Most of those cases are not designed to evacuate heat and suck in cool air, they're designed to be quiet - even under load.
100% not true for laptops. They get very loud and hot under load. That represents a large and growing portion of all PCs. And its not true for cheap desktop PCs running at high load either. I've mined on them, so I can tell you, they get loud.
|
|
|
|
lasybear
Member
Offline
Activity: 113
Merit: 14
|
|
May 16, 2014, 10:50:03 AM |
|
lazybear are you interested in a bounty to release the source code (maybe cleaned up a bit?) your optimized miner?
If not, I'll probably play around with the code myself tomorrow and see if I can come up with something, or maybe Noodle Doodle will take an interest.
0. I'm not a git and programming friendly. 1. There're AES and SSE hard coded, without choice. 2. There're 3 files with a bunch of indish code. 3. Tested only on linux x64 (Ubuntu 14.04) gcc 4.8.2, boost 1.55 Ok?
|
|
|
|
|