• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 98 subscribers
  • Views 259 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

About the platform_test project comes with EVM6678

may may92122
Expert 1030 points

When we got the EVM6678, after installed ccs 5.0 and BiosSDK, there is a platform_test project  under the install filefolder. It has some code to initialize EVM6678, such as PLL configuration. I want to know that under these configuration, under which main frequency 6678 runs? Is it 1G?

Looking forwards some body can help me!

  • 0
    Genius 3770 points

    Hi May,

    Yes, if you didn't make any change to the gel file, the main freq of core is @1Ghz.

    This is done by the Set_Pll1() function in evmc6678l.gel

    When you connect to the target, you can have a look at the console output for the actual frequency.

     

    Allen

  • 0 in reply to Allen Lee
    Expert 1030 points

    Thanks Allen,

    I asked that question because I tested some code on 6713 and one core of 6678, the running time on 6713 is almost two times of one core of 6678. Once a TI engineer told that under same main freq of core, the performance of 6678 is at least 4 times of 6713. So I doubt whether the core of 6678 runs under 1GHz.

  • 0 in reply to may may92122
    Genius 3770 points

    Hi May,

    C6713 is a floating-point DSP which is based on a C67x CPU, and C6678's ISA is composed of C64x+ and C674x.

    So if the test code is floating-point and they are running at same freq, the performance should be similar because the FLOP ability of C67x and C674x doesn't differ too much.

    And if the code is fixed-point, one core of C6678 should two times than C6713 because the fixed-point processing ability of one C64x+  and one C67x is equivalent to two C67x approximately. There is no evidence of 4 times difference.

    Allen

  • 0 in reply to Allen Lee
    Expert 1030 points

    I know, thanks Allen.