LOGIN   :::   RECOVER PASS   :::   GET ACCOUNT    
Browse
  • Projects
  • Code (CVS)
  • Forums
  • News
  • Articles
  • Polls
  •  
    OpenCores
  • FAQ
  • CVS HowTo
  • Mission
  • Media
  • Tools
  • Advertise
  • Mirrors
  • Logos
  • Contact us
  • Job Opportunity
  •  
    Tools
  • Search
      
  • Download Cores (CVSGet)
  •  
    More
  • Wishbone
  • Perlilog
  • EDA tools
  • OpenTech CD
  •  
    Navigation: All forums > Cores > Message List > Message Post

    Message

    Reply | Reply all
    Date Prev | Date Next | Thread Prev | Thread Next Date Index | Thread Index

    From: z3qmtr45 at iinet.com<z3qmtr45@i...>
    Date: Tue May 27 00:39:03 CEST 2008
    Subject: [oc] IP Core design in the 21st Century
    Top
    Hello,

    I started my career as a digital designer in the 70's. Back then all
    you really needed was a good schematic capture system to enter your
    designs. That
    lasted until the 90's when chips had grown to the point where you
    simply could not plop down gates and hook them up fast enough to meet
    a reasonable schedule so we switched to creating the design in RTL and
    then synthesizing it into gates.

    Well chips haven't stopped growing and we are now approaching the
    point were you won't be able to write RTL fast enough to create all
    the code that you need to fill a modern asic.

    There are a lot of people in the industry trying to figure out the
    next big quantum leap that will save us again but so far there is no
    agreement as to what that will be.

    I would like to start an ongoing discussion on this forum as to what
    future ip cores will look like. Hopefully this will lead to a
    component design standard and opensource EDA toolsets that could be
    used to quickly and easily assemble a group of cores into a working
    system.

    Most of the cores available on this site are small enough that
    RTL/Synthesis still works fine. The problem is that if you want to use
    one of these cores in a larger asic then it takes a good deal effort
    to figure out where everything is at and what works or doesn't work.

    One of the most promising possible solutions is IP-XACT. This is a
    soon to be IEEE spec that tells the ip core designer how to design a
    core so that it
    is (dare I say it?) Plug-and-Play to add it into SOC.

    The nice things about this spec is that you really don't have to
    change anything that you are already doing. You create a IP-XACT file
    that describes your core and include that with your ip. Any IP-XACT
    enabled tool can then read your IP and use it.

    The file uses XML that is a open format so anyone can write a tool
    that can use the information from the IP_XACT file without have to
    resort to hand writing your own scripts.

    There are commercial ($$$$) tool sets out there using this. I would like
    to start creating some opensourced tools that could be used to assemble
    opencores ip into SOC's.

    Anyone want to help or have any suggestions?

    John Eaton













    Follow upAuthor
    [oc] IP Core design in the 21st CenturyGünter Dannoritzer

     
    Copyright (c) 1999 OPENCORES.ORG. All rights reserved.