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

[fanepatent]

Got some stuff reconfigured, attached drives to a different PC.  Reinstalled 13.12 drivers.  Now I am trying to use 'GPU plot generator v2.1.1'.  It gets to  --->

Creating OpenCL program
[ERROR] Unable to open the source file

Here is the command I am running -
gpuPlotGenerator.exe generate 0 0 "D:\plots" xxxxAddressxxx 0 32000000 1024 256 4080

Any ideas?

gpuPlotGenerator.exe generate 0 0 "plots" xxxxAddressxxx 0 32000000 1024 256 4080

And make the plots folder where the generator is. Or move the ploter in the root of drive D if you already have the plots folder there

[4x13]

another question about plot creation....

I am still using the original run_generate.sh script to generate plots on a Mac.

what I want to do is to create my plots in 200GB files for easy of management, so based on the posted info, 80000 plots should equal 200GB

so I use the command;

./run_generate.sh <My Address> 0 80001 8191 4

The terminal says it create a plot and goes from 0 - 80000 in the terminal, but the file is only 21GB

next I ran this command

./run_generate.sh <My Address> 80001 160000 8191 4

to get the next plot from 80001 - 160000 which should be another 200GB file but this file is only 43GB but the terminal says it went from 80001 all the way through 240000

why?? how do you control the sizes of the plots accurately and make sure there are no crossovers or missing hashes in the plots??

[Dzus1k]

Are 3000-3500 bursts a day normal for 6TB?

have 6 tb too and have only 1200 coins daily so ye with 3000 you are pretty lucky

[Pilotseye]

another question about plot creation....

I am still using the original run_generate.sh script to generate plots on a Mac.

what I want to do is to create my plots in 200GB files for easy of management, so based on the posted info, 80000 plots should equal 200GB

so I use the command;

./run_generate.sh <My Address> 0 80001 8191 4

The terminal says it create a plot and goes from 0 - 80000 in the terminal, but the file is only 21GB

next I ran this command

./run_generate.sh <My Address> 80001 160000 8191 4

to get the next plot from 80001 - 160000 which should be another 200GB file but this file is only 43GB but the terminal says it went from 80001 all the way through 240000

why?? how do you control the sizes of the plots accurately and make sure there are no crossovers or missing hashes in the plots??

80000 equals 20GB so you need one more zero behind the eight  

[4x13]

ok, but why is there an overlap when I go to create the second and third plots

when I go to create the third plot, I am watching the terminal and there is overlap in the creation from plot #2 to plot #3

./run_generate.sh <My Address> 80001 160000 8191 4


./run_generate.sh <My Address> 160001 240000 8191 4

shouldn't these be consecutive plots? or am I missing something?

and if this is correct, why is the second plot twice the size as the first plot if each plot is only supposed to be 20GB (By my mistake on size)??

[Pilotseye]

ok, but why is there an overlap when I go to create the second and third plots

when I go to create the third plot, I am watching the terminal and there is overlap in the creation from plot #2 to plot #3

./run_generate.sh <My Address> 80001 160000 8191 4


./run_generate.sh <My Address> 160001 240000 8191 4

because it does not create exactly 80000 but 81920 if you use 8192 (= 8GB RAM) for example
that's because the tool adjusts the nonces to match the stagger size, but that's not new  

[pinballdude]

I am solo mining, have been for about 12 hours.  I do not see any transactions on my wallet(s) yet.  Is this normal?  Would a pool present quicker payouts?  

You only get a transaction if you generate the lowest deadline of everyone in the entire network. That takes quite a bit of luck, so even though there is a "lottery" every 4 mnutes, it can take days, even for a relatively large number of terabytes.

I have over 7TB plotted now, and i have hit 3 blocks in like 10 days. The normal day is with no blocks.

At this link, https://bchain.info/BURST/tools/calculator  you can type in how many TB you have, and then see how many coins you get per day. I typed in 7 TB and got 3714 coins.
Now, take the current block reward, which is 9500 and divide by the coins per day from above : 9500/3714
The result, 2.55 is the average number of days you can expect to have to wait between blocks.

You can also take the month number, i get 111414   then divide by 9500 that is about how many blocks you will get in a month. I get 11.7 so during a month i should get about 12 blocks.

As everything is based on randomness, the actual number of days will vary of course, but this is probably close to the best estimates.

I have 20TB on the way, when that is installed and plotted i hope to get to about a block a day.

Solo mining is like watching paint dry..  not that much excitement most of the time.

[Pilotseye]

Got some stuff reconfigured, attached drives to a different PC.  Reinstalled 13.12 drivers.  Now I am trying to use 'GPU plot generator v2.1.1'.  It gets to  --->

Creating OpenCL program
[ERROR] Unable to open the source file

Here is the command I am running -
gpuPlotGenerator.exe generate 0 0 "D:\plots" xxxxAddressxxx 0 32000000 1024 256 4080

Any ideas?

gpuPlotGenerator.exe generate 0 0 "plots" xxxxAddressxxx 0 32000000 1024 256 4080

And make the plots folder where the generator is. Or move the ploter in the root of drive D if you already have the plots folder there

did you install AMD SDK again as well together with 13.12 drivers?

[fivebells]

Got some stuff reconfigured, attached drives to a different PC.  Reinstalled 13.12 drivers.  Now I am trying to use 'GPU plot generator v2.1.1'.  It gets to  --->

Creating OpenCL program
[ERROR] Unable to open the source file

did you install AMD SDK again as well together with 13.12 drivers?

When I get this error, it is because the kernel files are not where the plotter is expecting.  Where is the kernel subdirectory, relative to your current working directory at the time you run the plotter?

[DrTrouble]

Got some stuff reconfigured, attached drives to a different PC.  Reinstalled 13.12 drivers.  Now I am trying to use 'GPU plot generator v2.1.1'.  It gets to  --->

Creating OpenCL program
[ERROR] Unable to open the source file

Here is the command I am running -
gpuPlotGenerator.exe generate 0 0 "D:\plots" xxxxAddressxxx 0 32000000 1024 256 4080

Any ideas?

gpuPlotGenerator.exe generate 0 0 "plots" xxxxAddressxxx 0 32000000 1024 256 4080

And make the plots folder where the generator is. Or move the ploter in the root of drive D if you already have the plots folder there

I moved the plots folder in to the generator folder and it works - thanks!!

[4x13]

ok, so this should be the correct command

100GB plot #1 ./run_generate.sh <My Address> 0 409600 8191 4
100GB plot #2 ./run_generate.sh <My Address> 409601 819200 8191 4
100GB plot #3 ./run_generate.sh <My Address> 819201 1228801 8191 4

etc.. etc.. is this correct for creating consecutive plot files without gaps??

--> (This should create 100GB files)
--> (because I have 4GB of ram spec'd in my generate file is this correct??)
--> (java -Xmx4096m -cp pocminer.jar:lib/*:lib/akka/*:lib/jetty/* pocminer.POCMiner generate $@)

Please let me know if this is correct so I can get started generating my plots,

Thanks..

[Janror]

Some words about the performance of my wplotgenerator:

• All of these benchmark results I’ve posted are real values from actual computers with the listed processors. There is nothing faked or made up, but there was nothing else running and I’ve chosen optimal parameters for performance.
• I used this command line to generate these results:
  wplotgenerator 1 0 9600 1920 %NUMBER_OF_PROCESSORS% /async
• I’ve chosen the strange stagger size of 1920 because of their many prime factors (2*2*2*2*2*2*2*3*5) and the small needed memory footprint (about 1 GB RAM with active /async), so it would be running nice and fast on a bunch of different computers.
• The prime factors of the stagger size are relevant, if you want to achieve maximum performance, since the wplotgenerator is dividing the actual work in many smaller pieces. To get the maximum performance these values have to add up correctly, because with each stagger the calculation is stopping for a brief moment and then it have to calculate the remainder of the outstanding work and to write the stagger to the disk. While plotting, you will see this stagger change in short drops of your current CPU usage in task manager.
• Keep in mind, that Windows is caching the written data for a while. So activating /async is only useful, if you are plotting to a network-location. If you are plotting to your local HDD don’t use it, since it is using RAM, which Windows would need for its file cache.
• What hits the performance values for plotting:
  
  • Of course mining does, but since we want the new plots mined as fresh as possible, there is nothing we can do about this.
  • Other active programs. Too allow optimal mining-performance while plotting, the wplotgenerator actively reduces the priority of its process. So pretty much every other program that runs, will take precedence over wplotgenerator.
  • Too high stagger sizes. If your computer is running out of available RAM, it will be swapping, which will hurt the performance a LOT.
  • Virtualization could limit the available code paths, since not all hypervisors are capable to provide the CPUs AVX/AVX2-units to virtual machines.
  • Not optimal RAM configuration. Not all computers are built for optimal RAM performance. Always try to install matching pairs of memory modules for all of your memory channels. http://en.wikipedia.org/wiki/Multi-channel_memory_architecture
• The optimal stagger size for PLOTTING performance should be a multiply of your chosen thread count, a CPU factor of the available code path (factor 4 for SSE4/AVX and factor 8 for AVX2) and a memory factor to use about half of your free RAM.
• To calculate the memory factor use your free RAM in megabytes, divide it by 2 to use only the half, then divide it by the count of the threads you’d like to use (you could take all your cores of your CPU, since wplotgenerator is backing down) and at last divide it by 4 for SSE/AVX or 8 if your CPU is capable of AVX2. Round the resulting value (the rounding is important, because that’s the reason, we are doing all of this). In “short”:
  [memory factor] = round([free memory in MB] / 2 / [thread count] / [CPU factor])
• To calculate the optimal stagger size for optimal plotting performance you’ve to multiply these factors back up:
  [optimal performance stagger size] = [rounded memory factor] * [CPU factor] * [thread count] * 4 (plots per MB)
• For example if you have 12.3 GB (or about 12'595 MB) of free RAM and an Intel i7 3770 this would be:
  memory factor = 12'595 / 2 / 8 (threads) / 4 (only AVX) = 196.8 --> rounded to 197
  optimal performance stagger size = 197 * 4 * 8 * 4 = 25'516
  With this stagger size the plotter is using 6'304 MB for plots and keeps the CPU as busy as possible. The rest of the RAM is used for the Windows file cache to keep the time to pause for writing the plots to the disk as short as possible.

I’ve used a small test setup, to create my reference performance values. You can download it from here (updated with version 1.16): https://mega.co.nz/#!D5lmAJaT!4SpnRR_gvkqTPzZMhDqxjcdSe4X1xeEFFoElo6kVGzA

Instructions: Download the file, extract it to a location where you have at least about 2 GB of free disk space (I know, these locations are getting pretty rare in the last weeks ) and launch the Test.bat file. Since the batch file is running CPU-Z to output some CPU details, you possible need to approve the UAC prompt if you would like to see them. The batch file is using the included program “File Checksum Integrity Verifier” (fciv) to calculate a SHA1-hash of the resulting plot file. This is an example of one of my machines:

Windows Burstcoin plot generator V1.15
Creating plots for nonces 0 to 9600 (2 GB) using 960 MB memory and 12 threads with async writing - using AVX codepath
Setting priority class 'below normal'.
100 Percent done. Average 15346 nonces/minute.
Finished plotting.
//
// File Checksum Integrity Verifier version 2.05.
//
2b96345e9dff2fa2fe2c6e995f8ed091ae970a44 plots\1_0_9600_1920
        Codename                Sandy Bridge-EP/EX
        Specification           Intel(R) Xeon(R) CPU E5-1660 0 @ 3.30GHz
        Instructions sets       MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, EM64T, VT-x, AES, AVX

The created plot file and the temporary CPU-Z output will be automatically deleted at the end. The printed SHA1 hash should ALWAYS be 2b96345e9dff2fa2fe2c6e995f8ed091ae970a44.

As always, please consider donating some of your newly mined Bursts to support further development or instructions: BURST-LNVN-5M4L-S9KP-H5AAC

And of course, happy mining!
Janror

[enta2k]

Are 3000-3500 bursts a day normal for 6TB?

have 6 tb too and have only 1200 coins daily so ye with 3000 you are pretty lucky

Good to know, im wondering, the pool is pretty stuffed, every 2-3 submitting fail.
Maybe i try another one, with better performance.

Is there any server, that don´t fails every 2 tasks?

[alfbugbug]

Hey dev,

Where do the next developments on the coin stand?
There are some interesting features in the list,
would love to hear about the progress.

[OmegaReaper]

Are 3000-3500 bursts a day normal for 6TB?

have 6 tb too and have only 1200 coins daily so ye with 3000 you are pretty lucky

Good to know, im wondering, the pool is pretty stuffed, every 2-3 submitting fail.
Maybe i try another one, with better performance.

Is there any server, that don´t fails every 2 tasks?

http://burstpool.ddns.net/

[Irontiga]

Hey dev,

Where do the next developments on the coin stand?
There are some interesting features in the list,
would love to hear about the progress.

I heard that escrow was getting close

[enta2k]

Are 3000-3500 bursts a day normal for 6TB?

have 6 tb too and have only 1200 coins daily so ye with 3000 you are pretty lucky

Good to know, im wondering, the pool is pretty stuffed, every 2-3 submitting fail.
Maybe i try another one, with better performance.

Is there any server, that don´t fails every 2 tasks?

http://burstpool.ddns.net/

Gonna give it a try, im on it now.

Wohoo, 3 shares in a row, unbelievable

Edit: Okay this pool works 100 times better, if the payment is good, im gonna stay

[Avaahnaa]

ok, so this should be the correct command

100GB plot #1 ./run_generate.sh <My Address> 0 409600 8191 4
100GB plot #2 ./run_generate.sh <My Address> 409601 819200 8191 4
100GB plot #3 ./run_generate.sh <My Address> 819201 1228801 8191 4

etc.. etc.. is this correct for creating consecutive plot files without gaps??

--> (This should create 100GB files)
--> (because I have 4GB of ram spec'd in my generate file is this correct??)
--> (java -Xmx4096m -cp pocminer.jar:lib/*:lib/akka/*:lib/jetty/* pocminer.POCMiner generate $@)

Please let me know if this is correct so I can get started generating my plots,

Thanks..

No this is not correct. See below.

ok, but why is there an overlap when I go to create the second and third plots

when I go to create the third plot, I am watching the terminal and there is overlap in the creation from plot #2 to plot #3

./run_generate.sh <My Address> 80001 160000 8191 4


./run_generate.sh <My Address> 160001 240000 8191 4

shouldn't these be consecutive plots? or am I missing something?

and if this is correct, why is the second plot twice the size as the first plot if each plot is only supposed to be 20GB (By my mistake on size)??

You are getting overlap because you are not following the syntax correctly.


It is:

 <ACC#> <start nonce> <number of nonces> <stagger> <threads>


Not:

 <ACC#> <start nonce> <end nonce> <stagger> <threads>

[enta2k]

i made the same mistake, had to replot 

[zznbey]

WTB burst