Author

Topic: [ANN][BURST] Burst | Efficient HDD Mining | New 1.2.3 Fork block 92000 - page 965. (Read 2171083 times)

full member
Activity: 224
Merit: 100
Which stagger size I should use with 32gb ram on 3tb hdd?

If you are on Windows 8191 is the max size. If you are on linux go with 100000 (100000 x 256) to be safe
legendary
Activity: 1512
Merit: 1000
quarkchain.io
Which stagger size I should use with 32gb ram on 3tb hdd?
legendary
Activity: 1582
Merit: 1019
011110000110110101110010
Is there a way I can generate plots on a cloud server? How and where?
Yeah, you can, but storage ain't cheap..

How and where please? I would like to look at it from top to bottom and do a cost benefit analysis.
Check them out at http://serverbear.com/

there is something for cpu, i need to plot faster, i mean soemthing like betarig, but for cpu, i don't want to deal with all the bullshit of address/card holder and those shit

would this work? http://www.cpusage.com/

Remember that you need to transfer the generated stagger from RAM to disk. So every time a stagger is generated it would have to be written somewhere.
You will rent cpu power, but where are you writing the files?

They had a demo video there of an app that scraped website data to Amazon storage using multiple CPU instances. I think it would work but I also think it would be too costly unless I am missing something.
full member
Activity: 224
Merit: 100
Is there a way I can generate plots on a cloud server? How and where?
Yeah, you can, but storage ain't cheap..

How and where please? I would like to look at it from top to bottom and do a cost benefit analysis.
Check them out at http://serverbear.com/

there is something for cpu, i need to plot faster, i mean soemthing like betarig, but for cpu, i don't want to deal with all the bullshit of address/card holder and those shit

would this work? http://www.cpusage.com/

Remember that you need to transfer the generated stagger from RAM to disk. So every time a stagger is generated it would have to be written somewhere.
You will rent cpu power, but where are you writing the files?
legendary
Activity: 1582
Merit: 1019
011110000110110101110010
Is there a way I can generate plots on a cloud server? How and where?
Yeah, you can, but storage ain't cheap..

How and where please? I would like to look at it from top to bottom and do a cost benefit analysis.
Check them out at http://serverbear.com/

there is something for cpu, i need to plot faster, i mean soemthing like betarig, but for cpu, i don't want to deal with all the bullshit of address/card holder and those shit

would this work? http://www.cpusage.com/
legendary
Activity: 3248
Merit: 1070
Is there a way I can generate plots on a cloud server? How and where?
Yeah, you can, but storage ain't cheap..

How and where please? I would like to look at it from top to bottom and do a cost benefit analysis.
Check them out at http://serverbear.com/

there is something for cpu, i need to plot faster, i mean something like betarig, but for cpu, i don't want to deal with all the bullshit of address/card holder and those shit
member
Activity: 112
Merit: 10
Question.

Is there any difference in terms of mining (not plot) when using stagger 1000 VS stagger 8000 ?
BIIIGGG difference, bigger stagger requires more RAM, but lowers disk stress.

i am talking about mining , is that true for mining ? I am not asking for plot

He said the mining

I plot way faster with 1000 stagger than 8000 stagger. What does disk stress actually means ? I have to sacrifice plotting performance for less disk stress ?

The idea behind the plots is that if you use 8000 stagger the disk will seek once. If you use 1000 stagger the disk will seek 8 times in the same file of 8000 nonces.
Thats the difference regarding disk stress.
The real difference in mining is that with a larger stagger you will be able to submit shares faster.
There is no memory usage difference in mining, unlike irontiga said.

All miner only submit share once, is there any benefit for submitting share faster other than if ur share is a block find ? Even if it took me 10 sec to submit share , as long as within that 10 sec no block is found , it's no difference submitting share in 1 sec or 10 sec. Am i right ?

If i am right, lower stagger actually work way better if CPU is not i7 , 8 gb ram because i can plot faster and there is no difference in mining.

I guess you are right. But for example, if I had a 6 Tb drive, and for the sake of this example I have 80 plots with 1000 stagger. Lets say 1 second per file, although it takes more than 2 (depending on size). Thats 80 second right there. If you have 10 files with 8000 stagger the same operation would take 10 seconds.

luckily block time target is 4 mins
I was giving you an example where the stagger size mattered.
I am making the plots at 40000 stagger. Cheesy

Let's get some straight answer feedback from everyone:

- Is a stagger of 1000 - with everything else held constant - faster at plot generation of 500GB plots than a stagger size of 8000?

- Is the 64bit version of JAVA in its allowance for higher stagger and maximum memory usage faster at plot generation of 500GB plots than the 32bit version?

- What is the most optimum settings in terms of speed in getting the plots generated for the average user with an i5 3570k or similar processor and 8GB of ram?
sr. member
Activity: 458
Merit: 250
beast at work

The idea behind the plots is that if you use 8000 stagger the disk will seek once. If you use 1000 stagger the disk will seek 8 times in the same file of 8000 nonces.
Thats the difference regarding disk stress.

how would you guys like to have a stagger of 40960 or bigger ? Smiley

i did my plotting on chunks, files of 10Gb (easier to generate in one place and move via network in other places) with a stagger of 8191 then use the merge utility provided by dcct to rearrange/merge the content of the files from a stagger of 8191 to the number of nouances in the file (in my case 40960)
full member
Activity: 238
Merit: 100
it is great that now have 209.16 MH on scrypt multipool , but i was wonder there the multipool dump and buy burst? polo or c-cex
full member
Activity: 224
Merit: 100
Question.

Is there any difference in terms of mining (not plot) when using stagger 1000 VS stagger 8000 ?
BIIIGGG difference, bigger stagger requires more RAM, but lowers disk stress.

i am talking about mining , is that true for mining ? I am not asking for plot

He said the mining

I plot way faster with 1000 stagger than 8000 stagger. What does disk stress actually means ? I have to sacrifice plotting performance for less disk stress ?

The idea behind the plots is that if you use 8000 stagger the disk will seek once. If you use 1000 stagger the disk will seek 8 times in the same file of 8000 nonces.
Thats the difference regarding disk stress.
The real difference in mining is that with a larger stagger you will be able to submit shares faster.
There is no memory usage difference in mining, unlike irontiga said.

All miner only submit share once, is there any benefit for submitting share faster other than if ur share is a block find ? Even if it took me 10 sec to submit share , as long as within that 10 sec no block is found , it's no difference submitting share in 1 sec or 10 sec. Am i right ?

If i am right, lower stagger actually work way better if CPU is not i7 , 8 gb ram because i can plot faster and there is no difference in mining.

I guess you are right. But for example, if I had a 6 Tb drive, and for the sake of this example I have 80 plots with 1000 stagger. Lets say 1 second per file, although it takes more than 2 (depending on size). Thats 80 second right there. If you have 10 files with 8000 stagger the same operation would take 10 seconds.

luckily block time target is 4 mins
I was giving you an example where the stagger size mattered.
I am making the plots at 40000 stagger. Cheesy
hero member
Activity: 868
Merit: 1000
Question.

Is there any difference in terms of mining (not plot) when using stagger 1000 VS stagger 8000 ?
BIIIGGG difference, bigger stagger requires more RAM, but lowers disk stress.

i am talking about mining , is that true for mining ? I am not asking for plot

He said the mining

I plot way faster with 1000 stagger than 8000 stagger. What does disk stress actually means ? I have to sacrifice plotting performance for less disk stress ?

The idea behind the plots is that if you use 8000 stagger the disk will seek once. If you use 1000 stagger the disk will seek 8 times in the same file of 8000 nonces.
Thats the difference regarding disk stress.
The real difference in mining is that with a larger stagger you will be able to submit shares faster.
There is no memory usage difference in mining, unlike irontiga said.

All miner only submit share once, is there any benefit for submitting share faster other than if ur share is a block find ? Even if it took me 10 sec to submit share , as long as within that 10 sec no block is found , it's no difference submitting share in 1 sec or 10 sec. Am i right ?

If i am right, lower stagger actually work way better if CPU is not i7 , 8 gb ram because i can plot faster and there is no difference in mining.

I guess you are right. But for example, if I had a 6 Tb drive, and for the sake of this example I have 80 plots with 1000 stagger. Lets say 1 second per file, although it takes more than 2 (depending on size). Thats 80 second right there. If you have 10 files with 8000 stagger the same operation would take 10 seconds.

luckily block time target is 4 mins
full member
Activity: 224
Merit: 100
Question.

Is there any difference in terms of mining (not plot) when using stagger 1000 VS stagger 8000 ?
BIIIGGG difference, bigger stagger requires more RAM, but lowers disk stress.

i am talking about mining , is that true for mining ? I am not asking for plot

He said the mining

I plot way faster with 1000 stagger than 8000 stagger. What does disk stress actually means ? I have to sacrifice plotting performance for less disk stress ?

The idea behind the plots is that if you use 8000 stagger the disk will seek once. If you use 1000 stagger the disk will seek 8 times in the same file of 8000 nonces.
Thats the difference regarding disk stress.
The real difference in mining is that with a larger stagger you will be able to submit shares faster.
There is no memory usage difference in mining, unlike irontiga said.

All miner only submit share once, is there any benefit for submitting share faster other than if ur share is a block find ? Even if it took me 10 sec to submit share , as long as within that 10 sec no block is found , it's no difference submitting share in 1 sec or 10 sec. Am i right ?

If i am right, lower stagger actually work way better if CPU is not i7 , 8 gb ram because i can plot faster and there is no difference in mining.

I guess you are right. But for example, if I had a 6 Tb drive, and for the sake of this example I have 80 plots with 1000 stagger. Lets say 1 second per file, although it takes more than 2 (depending on size). Thats 80 second right there. If you have 10 files with 8000 stagger the same operation would take 10 seconds.
newbie
Activity: 54
Merit: 0
hero member
Activity: 686
Merit: 500
 that site won't work on my phone, can you just tell me  how many coins have been mined so far?

9607 * 10'000 = 96'070'000
newbie
Activity: 54
Merit: 0
 that site won't work on my phone, can you just tell me  how many coins have been mined so far?
hero member
Activity: 868
Merit: 1000
Question.

Is there any difference in terms of mining (not plot) when using stagger 1000 VS stagger 8000 ?
BIIIGGG difference, bigger stagger requires more RAM, but lowers disk stress.

i am talking about mining , is that true for mining ? I am not asking for plot

He said the mining

I plot way faster with 1000 stagger than 8000 stagger. What does disk stress actually means ? I have to sacrifice plotting performance for less disk stress ?

The idea behind the plots is that if you use 8000 stagger the disk will seek once. If you use 1000 stagger the disk will seek 8 times in the same file of 8000 nonces.
Thats the difference regarding disk stress.
The real difference in mining is that with a larger stagger you will be able to submit shares faster.
There is no memory usage difference in mining, unlike irontiga said.

All miner only submit share once, is there any benefit for submitting share faster other than if ur share is a block find ? Even if it took me 10 sec to submit share , as long as within that 10 sec no block is found , it's no difference submitting share in 1 sec or 10 sec. Am i right ?

If i am right, lower stagger actually work way better if CPU is not i7 , 8 gb ram because i can plot faster and there is no difference in mining.
legendary
Activity: 1582
Merit: 1019
011110000110110101110010
WTF is this? I am running a machine with one external drive generating plots just fine but when I try to mine a pool I get a java heap memory error. Online says that there is a leak. This computer has 32 GB of RAM. Any ideas?



Running Win 7 64-bit SP1

you must run too many miners, what -xmx you have set, then * your miner's number, figure out how many memory have you used

I am running one miner and I did not think that you allocated for pool mining like you do for solo mining. My second computer has 4 HD generating plots and mining perfectly using the same bat:

Pool Mining BAT: C:\Windows\SysWOW64\java -cp pocminer_pool.jar;lib/*;lib/akka/*;lib/jetty/* pocminer_pool.POCMiner mine http://127.0.0.1:8125 http://178.62.39.204:8121

So are you saying I should have C:\Windows\SysWOW64\java -Xmx1000m -cp pocminer_pool.jar;lib/*;lib/akka/*;lib/jetty/* pocminer_pool.POCMiner mine http://127.0.0.1:8125 http://178.62.39.204:8121



what's your stagger size, it is used for temp memory swap for the disk, if you set 1000, then when mining, miner will read 1000 from the disk to memory

It is set to 1000 for plot generating. I went ahead and set the pool bats to C:\Windows\SysWOW64\java -Xmx1000m and no more crashes. So it seems that was my problem.
full member
Activity: 224
Merit: 100
Question.

Is there any difference in terms of mining (not plot) when using stagger 1000 VS stagger 8000 ?
BIIIGGG difference, bigger stagger requires more RAM, but lowers disk stress.

i am talking about mining , is that true for mining ? I am not asking for plot

He said the mining

I plot way faster with 1000 stagger than 8000 stagger. What does disk stress actually means ? I have to sacrifice plotting performance for less disk stress ?

The idea behind the plots is that if you use 8000 stagger the disk will seek once. If you use 1000 stagger the disk will seek 8 times in the same file of 8000 nonces.
Thats the difference regarding disk stress.
The real difference in mining is that with a larger stagger you will be able to submit shares faster.
There is no memory usage difference in mining, unlike irontiga said.
newbie
Activity: 54
Merit: 0
 how many coins have been mined so far?
Jump to: