[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 26 Jul 2013 00:32:37 +0100
From: Rafael Waldo Delgado Doblas <lord.rafa@...il.com>
To: john-dev@...ts.openwall.com
Subject: Re: Rafael's weekly report #6
Hello,
Thank you for the answers, I will work on this task.
First daily report.
Accomplishments:
1. Clean any OpenCL/CPU code from CGMiner.
2. Understand how CGMiner drivers works.
Priorities:
1. Get a detailed decryption of Scrypt.c. (I have sent a mail to cKolivas).
2. Move Scrypt.c to Epiphany.
3. Get one instance of Scrypt runing on one Ephipany core.
3. Expand to the rest of cores.
Regards,
Rafael.
2013/7/25 Solar Designer <solar@...nwall.com>
> Rafael -
>
> On Thu, Jul 25, 2013 at 12:04:37AM +0100, Rafael Waldo Delgado Doblas
> wrote:
> > 2013/7/20 Solar Designer <solar@...nwall.com>
> > > As I mentioned recently, I now think we should do Litecoin mining on
> > > Parallella in cgminer codebase, separately from JtR (at least
> > > initially). We promised Litecoin mining to the Parallella community,
> > > but we did not promise it to be part of JtR, and I think they're not
> > > very interested in it being part of JtR (although we may consider this
> > > later). We may nevertheless use this mailing list, as I expect that
> > > some of what we'd be discussing would also be relevant to Katja's
> > > project, and vice versa. I hope this project won't take too long, and
> > > we'd have time left for some JtR tasks before the end of summer as
> well.
>
> > I need your advice, Do you think that I should go with this task
>
> Yes. Like I said, we promised Litecoin mining to the Parallella
> community, and there's interest in cryptocurrency mining on Parallella
> boards - folks keep asking about expected speeds - e.g., see some of my
> tweets on this topic yesterday and today (@solardiz).
>
> > or do you prefer that I work in something different?
>
> Like what? If something specific is of interest to you, let us know!
>
> I did think of some alternative tasks for you, but they're either more
> about integration than programming (and we already got one of those with
> JtR/cgminer integration - and spent too much time on it already) or are
> not any easier for you than Litecoin mining on Parallella would be.
> Thus, if you manage to implement Litecoin mining on Parallella within a
> week and "complete" it within another week (so two weeks total), then
> you'll also have time for one or more other non-trivial tasks like this.
> If you fail to complete this task before the end of summer, then you'd
> also fail to complete the possible alternative tasks of similar or
> greater complexity. For example, descrypt on Epiphany and/or Zynq FPGA
> would be nice, but I think it'd require more mentoring from me (or from
> someone else), and given my limited availability (especially in the
> following two weeks) it's not a good option now. You have already
> learned a few things related to scrypt and Litecoin, whereas for
> descrypt you'd be starting from scratch and would (more) likely be stuck
> for these two weeks.
>
> > In any way we need to plan
> > milestones for the next task and goals, I think that this is the faster
> way
> > to develop something.
>
> OK, please propose a timeline for Litecoin mining on Parallella.
>
> > Ok, if finally you want that I work on this project, I will work on
> CGMiner
> > fork ignoring JtR code. I will develop a new driver based on thee
> existing
> > ARM driver, I think that it wil be better use it as base even if we don't
> > use the ARM CPU. I will compute different instances of scrypt on each
> > Parallella core.
>
> This sounds fine to me (although I am not familiar with cgminer codebase).
>
> > Anyway tell me if you really want that I work on this project
>
> Yes. Not much choice here right now. Just try to do it fast, and by
> doing so also leave time for JtR-related tasks of similar or greater
> complexity and demonstrate to us your ability to work on such tasks.
>
> > because looks
> > like it's not that much profitable to do Liteconin minning on Parallella
> > and I would be working on CGMiner instead of JtR.
>
> Yes, we realize these drawbacks.
>
> Thanks,
>
> Alexander
>
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.
