Skip to main content
arro239
arro239
Associate III
Aug 25, 2013 7:51 PM UTC
Question

When are we getting higher frequencies? :)

Posted on August 25, 2013 at 21:51

Looks like 168MHz Cortex-M4 was announced two years ago. Have ST announced any faster controllers/discovery boards since?

    This topic has been closed for replies.

    8 replies

    Tesla DeLorean
    Tesla DeLorean
    Guru
    Aug 25, 2013 8:17 PM UTC
    Posted on August 25, 2013 at 22:17

    They have a geometry shrink on the table.

    http://www.st.com/web/catalog/mmc/FM141/SC1169/SS1577/LN1806

      2MB of FLASH, and 180 MHz

    But, honestly if you want speed there are routes that get you there, and the M3/M4 aren't it. You need something with fast, and native SDRAM, DDRx access, and caching architectures which the the Cortex-Mx parts don't address. The FPU only supports 32-bit floats, there's no MMU, and a restrictive memory map.

    http://www.st.com/web/catalog/mmc/FM169/SC1156

    For cheap and fast processors look at what's being used in phones and tablets, a lot of that business has migrated to China, and the parts have a very short life-cycle.
    Tips, Buy me a coffee, or three.. PayPal Venmo (See Profile) Up vote any posts that you find helpful, it shows what's working..
    postoffice33
    postoffice33
    Visitor II
    Aug 26, 2013 12:03 AM UTC
    Posted on August 26, 2013 at 02:03

    I hope this is going to happen soon

    http://best5.it

    Anyway I share that using 

    http://www.st.com/web/catalog/mmc/FM169/SC1156

     should be ok.

    Tesla DeLorean
    Tesla DeLorean
    Guru
    Aug 26, 2013 12:47 AM UTC
    Posted on August 26, 2013 at 02:47

    Maybe it would be helpful to be specific about what you're doing, the kind of horse power you think you need, and what kind of frequency/bandwidth the processor/memory need to provide.

    Tips, Buy me a coffee, or three.. PayPal Venmo (See Profile) Up vote any posts that you find helpful, it shows what's working..
    Bill Lewis
    Bill Lewis
    Associate III
    Aug 26, 2013 6:43 PM UTC
    Posted on August 26, 2013 at 20:43

    Of course nothing there I'd ever be able to solder to a PCB.

    emalund
    emalund
    Associate III
    Aug 27, 2013 12:20 PM UTC
    Posted on August 27, 2013 at 14:20

    Of course nothing there I'd ever be able to solder to a PCB.

    yes, of course, if the package were such that you were ''ever be able to solder to a PCB'' you would get reflections of a magnitude that would kill the operation instantly.

    ''a wire is NOT a wire''

    Erik

    emalund
    emalund
    Associate III
    Aug 27, 2013 12:27 PM UTC
    Posted on August 27, 2013 at 14:27

    Of course nothing there I'd ever be able to solder to a PCB.

    yes, of course, if the package were such that you were ''ever be able to solder to a PCB'' you would get reflections of a magnitude that would kill the operation instantly.

    ''a wire is NOT a wire''

    Erik

    Bill Lewis
    Bill Lewis
    Associate III
    Aug 27, 2013 3:44 PM UTC
    Posted on August 27, 2013 at 17:44

    The high frequencies involved are inside the chip.  Running the current 168 MHz parts hand-soldered LQFPs is no problem.

    emalund
    emalund
    Associate III
    Aug 27, 2013 8:16 PM UTC
    Posted on August 27, 2013 at 22:16

    <i>The high frequencies involved are inside the chip.  Running the current 168 MHz parts hand-soldered LQFPs is no problem.</i>

    re reflections etc who gives a hoot about the frequency, it is the associated rise and fall times <b>that do not change by using a lower frequency</b> that causes the reflections.

    Erik

    Terms & ConditionsAccessibility statement