Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/tackling-large-scale-soc-and-fpga-prototyping-debug-challenges.2367/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021770
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Tackling large-scale SoC and FPGA prototyping debug challenges

Daniel Payne

Moderator
I've met and interviewed Brad Quinton at Tektronix here in Beaverton, Oregon and he just wrote an article at EE Times: Tackling large-scale SoC and FPGA prototyping debug challenges. Brad summarizes the core requirements for optimal SoC debugging:

  • Full RTL-level visibility with minimal dependence on FPGA CAD flows
  • Fast, efficient identification of the root cause of bugs
  • Run at full speed to enable verification with real-world signals
  • Time-correlated signals across different clock domains
  • Scale seamlessly to support large, complex designs
  • Enable efficient debugging of firmware and application-level software
  • Verify hardware and software interactions at the system-level

Three approaches have traditionally been applied to SoC debug:

  • Simulation acceleration
  • Emulation
  • FPGA-based prototyping

There are big issues with each of these approaches, however Brad's group has come up with something new for the third approach, FPGA-based prototyping by adding instrumentation efficient enough to allow hundreds of thousands of probes to be used, instead of a thousand probes (ie. ChipScope and SignalTAP).

Read the full EE Times article, or some of the Wiki pages here:

Have you tried FPGA-based prototyping, and if so how was the debug process for you? Fun or painful?

<script src="//platform.linkedin.com/in.js" type="text/javascript"></script>
<script type="IN/Share" data-counter="top"></script>
 
Last edited:
Debug process was mostly painfull. You simply don't have enough resources in an FPGA to log the whole set of signals. When you have limited visibility and Intermittent failures, the process becomes increasingly painfull. I have only used Xilinx tools so Not sure if Sysnopsys FPGA toolset has similar limitations.
 
I am using Synopsys Haps-6x FPGA prototyping boards. Our design was partitioned into 4 Vertex6 by using Synopsys Certify. Many cables among them. And also, I am using Synposys Identify to bring some signals out to watch for debugging. Will move to Vertex7 soon. They have Haps-7 coming out.

Debug process was mostly painfull. You simply don't have enough resources in an FPGA to log the whole set of signals. When you have limited visibility and Intermittent failures, the process becomes increasingly painfull. I have only used Xilinx tools so Not sure if Sysnopsys FPGA toolset has similar limitations.
 
Back
Top