Donations to open source a GPU Protoshares miner (PTS)

[archit]

Got it to compile in Visual Studio 12 but it isn't working ATM

[ptsrush]

beeeer.org itself very very bad lucky. 5 hours and no block found.

Start more miners! :)

beeeer.org is very very poor.

since another cuda miner with xpt protocol released, don't use this any more.

[5chdn]

beeeer.org itself very very bad lucky. 5 hours and no block found.

Start more miners! :)

[ptsrush]

beeeer.org itself very very bad lucky. 5 hours and no block found.

[phrozenspite]

dga
are you expecting that AMD cards will be able to mine this faster once someone ports the code like they are with other gpu mined coins?

[ksi]

How to compile on windows? I know almost nothing about UNIX.

[ptsrush]

Hi all,

the cudapts run great on my GT 560 at 540 cpm, gentoo AMD64.

Anyway, I can not found my shares on http://ptsweb.beeeeer.org/user/<my pts address>.

Is it really working?

Yes the user stats does work. If you only just started mining then you'll have to wait till a block is found by the pool.

got it.

thank you very much.

Update: get my first 0.04848 PTS now.

[hammurabi]

This code is for NVidia cards.
Do we have opensource gpu miners for radeons yet?

[ryen123]

Hi all,

the cudapts run great on my GT 560 at 540 cpm, gentoo AMD64.

Anyway, I can not found my shares on http://ptsweb.beeeeer.org/user/<my pts address>.

Is it really working?

Yes the user stats does work. If you only just started mining then you'll have to wait till a block is found by the pool.

[ptsrush]

Hi all,

the cudapts run great on my GT 560 at 540 cpm, gentoo AMD64.

Anyway, I can not found my shares on http://ptsweb.beeeeer.org/user/<my pts address>.

Is it really working?

[earthbound]

The following is unscrupulously copied/pasted from my erred compile attempt on Cygwin. I "boostrapped" and compiled the boost library from source, and I don't know whether I did that properly. Also, I'm reading in some places that cygwin won't compile with Windows' network sockets, and that boosts' sockets (?) will more reliably compile and work? I don't know how to switch things around to make that the case. At least its' compiling .o files, though . . .

http://pastebin.com/Sk0c2VjG

[Edit: oops! posted that as "code," heh heh]

Also, I've installed MSVC express (which I'm not actually sure cygwin is using?) and CUDA Toolkit 5.5, and I set the paths for those before invoking the cygwin shell, with the following batch:

Code Select Expand


REM NOTES:
REM A lot of prerequisites, including installing boost on cygwin; re:
REM http://www.boerngen-schmidt.de/2013/07/installing-boost-1-54-0-c-libraries-in-cygwin/

@echo off
call "C:\Program Files\Microsoft Visual Studio 8\Common7\Tools\vsvars32.bat"
REM start cmd /c "C:\Program Files (x86)\Microsoft Visual Studio 11.0\Common7\IDE\WDExpress.exe"

REM Additional environment variables for UCDA/MSVC/GCC compilation;
REM re: https://bitsharestalk.org/index.php?topic=2114.msg25184#msg25184
REM export PATH=$PATH:CUDADIR/bin
REM export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:CUDADIR/lib64
REM
REM OR, adapted:

PATH=%PATH%;C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v5.5\bin
LD_LIBRARY_PATH=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v5.5\bin

REM On 32-bit OS:
REM chdir C:\cygwin\bin
REM On 64-bit OS:
chdir C:\cygwin64\bin
bash --login -i


[gordonhucn]

Great work!! Just read some code in github, if we use gpu to cal sha512 in the protoshares_revalidateCollision function, it might give a little bit more performance.

[ripplexiaoshan]

Marked, waiting for the window version...

[dga]

Update from about 15 minutes ago:  I just committed another change to the repository that eliminates a few more errors when trying to compile under Cygwin.  It also boosts performance by a fairly negligible amount, at least on compute_35 devices.  I didn't bother benchmarking the difference on other cards, because the gain was only about 20 c/m on a Tesla.  But, hopefully, this is one step closer to something that will build happily on Windows.

[dga]

..So I guess that means "no" ?

It means "complicated"

Here's an example:  One of my systems that I've tested on is a new, nice Haswell-based machine from Dell.

CPU:  i7-4770
GPU:  GTX 650Ti

CPU hashing gets somewhere in the 330 c/s range using Yam.  Maybe more, maybe less, but in that range.
GPU hashing gets almost exactly the same.  I think I can sneak it a little higher in the next release, but for now, they're nearly identical.

The CPU costs $300.  Not counting motherboard and memory.

The GPU costs about $140, 1GB of DRAM included.

The CPU draws about 70-80W when running full tilt.  The GPU uses about 15-30W when running my code full speed.

You need one motherboard (~$120) and DRAM (~$40) for every CPU or two.  You can instead fit 3-6 GPUs into a motherboard, reducing those costs further.

Also, my code doesn't really take advantage of the zoominess of something like the 650Ti.   People have reported very good numbers from older cards.  The memory bandwidth and number of memory channels is pretty important here, and that's not increasing as radically fast as core counts.

In total, running PTS on a GPU probably costs 1/4th as much as doing it on a CPU.  That's not insignificant, but it's also not earth-shattering.  I suspect that given the current price of PTS and the hash rates, mining PTS on a CPU will actually remain power-profitable, which is more than you can say for most other options (DOGEcoin on a Haswell CPU isn't too bad, though.  *grin*)

As I said initially:  This doesn't (yet) change the game as radically for PTS as it did for things like Bitcoin.  The amount of memory accesses help level the game to a decent degree.  But I'll note that I think there's more room for optimization in the GPU code than the CPU code, so the gap will widen a bit more.