[ANN][BURST] Burst | Efficient HDD Mining | New 1.2.3 Fork block 92000

[DrTrouble]

Hot deal alert!  4TB external for $100.  Or, $25/1TB. I picked up 3 myself - http://www.ebay.com/itm/SAMSUNG-D3-Station-4TB-USB-3-0-3-5-Desktop-External-Hard-Drive-STSHX-D401TDB-Bl/291274747882?afsrc=1

[1715675367]

1715675367

[1715675367]

1715675367

[1715675367]

1715675367

[1715675367]

1715675367

[1715675367]

1715675367

[mczarnek]

I don't understand why does 4096 optimize it for hdd?

For each nonce/accountId combination a 256 kb chunk of data is pre-generated during the plotting process.
You can specify any ranges of nonces, just make sure they don't overlap. Otherwise same work will be done twice.
Plot files serve as a giant lookup table for further computation. Plot files can be stored on ANY device, hdds, sdds etc. Doesn't matter which device, only the available space matters.

For each block only 64 bytes (scoop) from each 256 kb chunk are really used. These 64 bytes are selected depending on block id (which is not known beforehand).

It's not possible to generate these 64 bytes on the fly. Only the entire 256 kb block can be generated, but only 64 bytes will be used. This makes the algo terribly inefficient for CPU/GPU mining.

As a result of final computation (during mining) a deadline is produced. A person with the shortest deadline will announce the next block once this deadline expires.
I hope this helped to clear things up a bit.

Also, have any coders looked through this coded and approved or has only one guy written it?  People analyzed the algorithm looking for flaws?

I did, haven't found any flaws. The algorithm looks very solid, couldn't find a way to abuse it.
I'm not a cryptography expert though.

Thank you, now how do you verify that the correct person generated the block?

And what if multiple people have their plots tell them that they can generate a block at the same time?

[FakeAccount]

Just came back from vacation, was gone a full week.  You idiots are still wasting your time on this crap coin?

only idiots take a full week off

[DrTrouble]

Just came back from vacation, was gone a full week.  You idiots are still wasting your time on this crap coin?

only idiots take a full week off

***Please do not feed the animals...***   

[mafostedu]

@ Blago,

Does your latest miner supports special characters in pass phrase for solo ?
If not, can you add it in the future version ?

Thanks.

[Blago]

@ Blago,

Does your latest miner supports special characters in pass phrase for solo ?
If not, can you add it in the future version ?

Thanks.

Yes, support.   https://github.com/Blagodarenko/miner-burst/blob/master/miner.cpp  string #517

[mafostedu]

@ Blago,

Does your latest miner supports special characters in pass phrase for solo ?
If not, can you add it in the future version ?

Thanks.

Yes, support.   https://github.com/Blagodarenko/miner-burst/blob/master/miner.cpp  string #517

Oh OK that's great.
One last question, if I have say 5 rigs & all rigs are connected to the same internet connection as individual system (Not like client-server scenario),
can I mine solo with wallet running on each system & your miner running individually on each system ? or should I create client-server setup & use your latest miner with running the wallet on server (1 of the 5 rigs) & other 4 systems connected as a client to it. Thanks for your great work & prompt reply.

[Irontiga]

@ Blago,

Does your latest miner supports special characters in pass phrase for solo ?
If not, can you add it in the future version ?

Thanks.

Yes, support.   https://github.com/Blagodarenko/miner-burst/blob/master/miner.cpp  string #517

Oh OK that's great.
One last question, if I have say 5 rigs & all rigs are connected to the same internet connection as individual system (Not like client-server scenario),
can I mine solo with wallet running on each system & your miner running individually on each system ? or should I create client-server setup & use your latest miner with running the wallet on server (1 of the 5 rigs) & other 4 systems connected as a client to it. Thanks for your great work & prompt reply.

Yes, will work, but wallet only needs be on 1 server. Just edit nxt-default.properties

[katlogic]

BTW: I believe a top end i7 can do 7000 nonces/min, which, correct me if i'm worng, = 1.7gb/minute....but that don't sound right.....maybe it's 170mb...

7000 nonces * 64 bytes scoop = 448000 bytes = 437 kb / min
For 20 min block it's equivalent of 8.5 mb disk space. Uh oh. Is it really that slow?

Wrong.

7000 nonces * 64 bytes per scoop * 4096 scoops = 1.4gb a minute, or better said about 30mbyte/s.

Remember, nonce generates all 4096 possible scoops. Reading 1 out of 4096 applies only to mining, not plotting.

Thank you, now how do you verify that the correct person generated the block?

And what if multiple people have their plots tell them that they can generate a block at the same time?

It works exactly same as NXT forging. Network can (easily) verify that nonce has a short deadline - it's a result of cryptographic hash with inputs outside of miner's control and he cant "fake" a good deadline. Just with more storage he has better chance to find shorter one. This is equivalent of forging balance.

Finally, the block is simply signed by forger's key. "tell them at the same time" is same as with bitcoin and NXT - blockchain will fork and longer branch wins. If you mean that two people can accidentaly have same plot files and same nonces - no thats not possible, plot files are derived from miner's address (which is checked when nonce is submitted). Block is signed by whomever owns that address.

Also, have any coders looked through this coded and approved or has only one guy written it?  People analyzed the algorithm looking for flaws?

In that regard, I'm mostly worried by NXT codebase used. It has a history of ugly vulnerabilities - https://web.archive.org/web/20140815132834/https://nextcoin.org/index.php/topic,3884.0.html

As for the PoW used in BURST, overall the process is called space-time tradeoff, and is used cryptography to offset computational cost:
http://en.wikipedia.org/wiki/Space%E2%80%93time_tradeoff#Lookup_tables_vs._recalculation

In case of BURST, it translates to space = log2(time), ie by reducing space in half you'd have to search twice as powers of 2 keyspace as before.

Either read up on the theory, or check the miner code yourself. If you can't, you'll have to trust our word on it.

[Hamuki]

Does this kind of mining ruin the HDD?

Am I able to delete the plotting so it will be a normal empty HDD when I dont want to mine anymore?

Please answer as fast as possible.

[bobafett]

where i can see, how many coins are minded until now?

[Irontiga]

Does this kind of mining ruin the HDD?

Am I able to delete the plotting so it will be a normal empty HDD when I dont want to mine anymore?

Please answer as fast as possible.

No, it doesn't ruin the hdd if you use a high enough stagger

Yes, you can delete plots any time, you won't lose your burst you've mined

where i can see, how many coins are minded until now?

127.0.0.1:8125/burst?requestType=getState

it's under

totalEffectiveBalanceNXT

[bobafett]

another perhaps stupid question:

What happens to the transaction Fees? In NXT the fee will spent to te people that are forging. What happens with the fee in burstcoin?

[Irontiga]

another perhaps stupid question:

What happens to the transaction Fees? In NXT the fee will spent to te people that are forging. What happens with the fee in burstcoin?

Same, the miners

[Blago]

@ Blago,

Does your latest miner supports special characters in pass phrase for solo ?
If not, can you add it in the future version ?

Thanks.

Yes, support.   https://github.com/Blagodarenko/miner-burst/blob/master/miner.cpp  string #517

Oh OK that's great.
One last question, if I have say 5 rigs & all rigs are connected to the same internet connection as individual system (Not like client-server scenario),
can I mine solo with wallet running on each system & your miner running individually on each system ? or should I create client-server setup & use your latest miner with running the wallet on server (1 of the 5 rigs) & other 4 systems connected as a client to it. Thanks for your great work & prompt reply.

Yes, will work, but wallet only needs be on 1 server. Just edit nxt-default.properties

Or you can run wallet for each rigs and set
   "Mode" : "solo",
   "Server" : "127.0.0.1",
   "Port": 8125,
   "UpdaterAddr" : "127.0.0.1",
   "UpdaterPort": 8125,
   "EnableProxy": false,

[Blago]

I tried to generate nonces on-fly, the result is very bad 1100 nonce / min (FX6300, 6 cores), because per one scoop miner must execute 4098 Shabals.
https://github.com/Blagodarenko/miner-burst/blob/master/miner.cpp line 626

[Irontiga]

I tried to generate nonces on-fly, the result is very bad 1100 nonce / min (FX6300, 6 cores), because per one scoop miner must execute 4098 Shabals.
https://github.com/Blagodarenko/miner-burst/blob/master/miner.cpp line 626

Nice! I think that is a very good thing to have, otherwise there is no proof that PoC is truly just PoC!!!

Besides, i have a much better cpu than i have hdd's, you know, help that net hash, lol, jk :/

[tex81]

1TB of plots are equivalent to 1 000 000 nonces per minute. So real-time plotting for mining is useless (with current software and speed).

[unsoindovo]

Someone want to check my calculations?

I'm calculating that assuming Bitcoin gets 280,000,000 GH/s, then assuming it uses the most energy and cost efficient miner I could find, then it costs $2 per GH and uses 0.5W/GH.

This means that the cost to buy the equipment to power the network is $560,000,000.
Also, the Bitcoin network uses 0.5 W per GH.  In other words 280,000,000 * 0.5 = 140,000,000 W


Now I went with the most energy efficient drive after quickly looking around, I found this: http://www.tomshardware.com/reviews/4tb-3tb-hdd,3183-15.html

The HITACHI Deskstar 5K4000:  http://www.amazon.com/HITACHI-0F14697-Deskstar-5K4000-INTERNAL/dp/B00B6TMG7O

This costs you $130 for 4TB.  So $560,000,000 will buy you ~4,300,000 4TB drives.

4,300,000 drives at 6W per drive gets you 25,800,000 W

Or in other words, ~4 times more energy efficient than Bitcoin..  which is nice but considering how much cheaper it is to forge for Nxt than mine for Bitcoin.. maybe not as much potential as I initially was thinking.  Especially if new Bitcoin mining equipment truly runs at 0.19 W instead of 0.5 W as used for the above calculation.. at which it's maybe half as energy efficient as Bitcoin.

See here: http://blogs.wsj.com/digits/2014/08/15/german-startup-says-its-new-chip-halves-bitcoin-mining-energy/

So this would use half as much energy as a similarly sized Bitcoin network.

Here is my much longer analysis of the Nxt network: https://docs.google.com/document/d/1J8uhdshu9epGRrQHBaloGc4itdvuAHZDAUtNDjOhz-8/edit?usp=sharing

It's still a cool idea and hard drives are more reusable though..  idk.  I'll probably still buy a little but it's not as exciting as I was initially thinking unless he can figure out a way to use that data to store something instead of simply mine with it.


Btw, FakeAccount, you may want to read this article on RAM optimizers:
http://www.howtogeek.com/171424/why-memory-optimizers-and-ram-boosters-are-worse-than-useless/

Mining using user-submitted data would not work well. Miners would likely submit files made of deterministically generated data they could re-create when needed, and use those to mine with, turning it into POW. I noticed you linked the Permacoin paper a few pages back, which only gets around this problem by using a central authority to decide which files are able to be mined with.

Where this is more efficient, is for people mining with existing hardware they're also using for other purposes. The difference in power between 'normal usage' of a computer, and 'normal usage' + mining would be a much more interesting comparison.

i think the real question it is highlighted above!!!
this is the real difference between PoW and PoC...
and obviously, PoC is much more better!!!
 

[uray]

1TB of plots are equivalent to 1 000 000 nonces per minute. So real-time plotting for mining is useless (with current software and speed).

wrong, its 4*1024*1024 , 1 nonce = 256KB (0.25MB)
if you can hash burst algo by 4.19 MH/s , its equal to 1TB of disk capacity
which is diffcult to achieve, because burst use 4098 round of hashing, compared that to X11 algo, which use 11 round, burst in term of hashing complexity is 372 times more secure than darkcoin!