[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAFYn=yCoj=e+M5=X3+34XV9oc+3RFGFN=pPmYpUn_OR_i97MKQ@mail.gmail.com>
Date: Fri, 26 Jul 2013 17:47:36 -0400
From: Yaniv Sapir <yaniv@...pteva.com>
To: john-dev@...ts.openwall.com
Subject: Re: Parallella: Litecoin mining
Rafael,
As a starting point, consider using (a) neighbor core(s) local memory to
store parts of your big array. It is much faster to retrieve data from an
on-chip memory than the external memory. Obviously, it means that you are
either not going to use the adjacent core for processing in parallel, or
you need to write the program in a way that core groups share the same data
buffer(s) (I am not familiar with the algorithm so I don't know if "V"
needs to be modified during the calculation).
Yaniv.
On Fri, Jul 26, 2013 at 4:57 PM, Rafael Waldo Delgado Doblas <
lord.rafa@...il.com> wrote:
> Hello,
>
> Second daily report.
>
> Accomplishments:
> 1. Start to move Scrypt.c to Epiphany.
>
> Priorities:
> 1. Move Scrypt.c to Epiphany.
> 3. Get one instance of Scrypt runing on one Ephipany core.
> 3. Expand to the rest of cores.
> 4. Get a detailed decryption of Scrypt.c. (I have sent a mail to cKolivas).
>
> Regards,
> Rafael.
>
> Well at this point I know what I should modified in CGMiner (I hope). I
> I'm available to run Scrypt in one core looks really easy expand to the
> rest of cores because of how is CGMiner coded. I didn’t get the detailed
> description of Scrypt yet but even if I don’t have it, I could start to
> work con Scrypt port.
>
> Each epiphany core has 32K and the cKolivas scyrpt implementation needs
> 128,5K plus the executable binary image size. LiteCoin has the next Scrypt
> parameters:
>
> r = 1
> p=1
> N=1024
>
> If I look in the reference implementation: "V" will need 128KB, "XY" 256B
> and "B" 128B then I the reference implementation will need ~128,37KB plus
> the executable binary image size.
>
> The Alexander's escrypt-tmto1 needs ~64K plus the executable binary image
> size then even with this TMTO ratio is not enough to work with Parallella.
>
> I was thinking in raise the CPU use in escrypt-tmto1 but is there a better
> way to get Scrypt on Parallella without raise cpu use? Do I losse
> something? Have I missed anything?
>
> BTW: The only function that we need to port to Epiphany is:
>
> scrypt_1024_1_1_256_sp(data, scratchbuf, ostate);
>
> data is the plaintext
> scratchbuf is a memory area to store V XY and B.
> ostate is the hash.
>
> scratchbuf will be inside of the epiphany core.
>
>
> Regards,
> Rafael.
>
>
--
===========================================================
Yaniv Sapir
Adapteva Inc.
1666 Massachusetts Ave, Suite 14
Lexington, MA 02420
Phone: (781)-328-0513 (x104)
Email: yaniv@...pteva.com
Web: www.adapteva.com
============================================================
CONFIDENTIALITY NOTICE: This e-mail may contain information
that is confidential and proprietary to Adapteva, and Adapteva hereby
designates the information in this e-mail as confidential. The information
is
intended only for the use of the individual or entity named above. If you
are
not the intended recipient, you are hereby notified that any disclosure,
copying,
distribution or use of any of the information contained in this
transmission is
strictly prohibited and that you should immediately destroy this e-mail and
its
contents and notify Adapteva.
==============================================================
Content of type "text/html" skipped
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
