Date: Tue, 25 Jun 2013 13:00:00 +0100 From: Rafael Waldo Delgado Doblas <lord.rafa@...il.com> To: john-dev@...ts.openwall.com Subject: Re: Parallella: scrypt Alexander, As you told me, I'm posting a copy of the mail that I sent you: I just finished the implementation of Scrypt format. Now I have some questions: The SSE implementation have a really poor performance on my i5-3570K only 30 c/s and it's really worst in Parallella where the ref implementation only gets 0.7, I have attached the scrypt_fmt.c (I'm using the same code from escrypt that you attached in you first mail). I'm working on my own git repo(https://github.com/LordRafa/jonh1.8.0), should I move to another one? Also I need to know the if this are the targets for my gsoc project: For the first half of the summer: First, implement scrypt in host mode. Second, experiment with scrypt time-memory tradeoff. Third, move it to epiphany. For the second half of the summer(early August or maybe before?) : Fourth, implement epiphany based Litecoin mining. Fifth, implement descrypt on Parallella. in oder to fix the gsoc project description. Furthermore I need to know the next task, yesterday you told me about interleaving and opencl, may you tell me more about those tasks? is it time-memory tradeoff related? My apologies, I'm a little bit lost but now that I finished my degree I can work on this project the whole day. Best regards, Rafa. 2013/6/24 Rafael Waldo Delgado Doblas <lord.rafa@...il.com> > Sorry, I forget to tell that I'm using only host cpu, working with the > reference implementation contained in escrypt-0.0.35.tar.gz. > > > 2013/6/24 Yaniv Sapir <yaniv@...pteva.com> > >> Rafael, >> >> It is really hard to tell if this is an expected result when we don't >> know much about *how* you integrated the Epiphany in (if at all... are you >> running on the host alone?) >> >> >> On Mon, Jun 24, 2013 at 3:54 PM, Rafael Waldo Delgado Doblas < >> lord.rafa@...il.com> wrote: >> >>> Finally, I have added the scrypt format based on escrypt sources. I get >>> a really poor performance on parallella, 0.7 c/s, is it normal?. >>> >>> Best regards, >>> Rafael >>> >>> >>> 2013/6/19 magnum <john.magnum@...hmail.com> >>> >>>> On 12 Jun, 2013, at 14:27 , magnum <john.magnum@...hmail.com> wrote: >>>> > We actually have Colin's reference implementation in bleeding-jumbo, >>>> added by Dhiru for the "scrypt" format (with format_name django-scrypt). >>>> And Jim optimized it to 2x. I haven't looked at it but maybe it should be >>>> renamed to django-scrypt? >>>> >>>> I renamed it as above now. The new format can be called scrypt and >>>> if/when it supports django hashes too, we can delete the django-scrypt >>>> format. >>>> >>>> magnum >>>> >>> >>> >> >> >> -- >> =========================================================== >> 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 View attachment "scrypt_fmt.c" of type "text/x-csrc" (6445 bytes)
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.