OMAP3530 versions and OPP modes...

Hi,

On the Gumstix-list we have had a discussion about which OPP modes were to be considered overdrive modes for the different OMAP35xx versions?

For the 600MHz version OPP5 is the highest supported mode.

For the 720MHz version OPP6 have been introduced to allow 720MHz operation.

In the datasheet it's stated that OPP5 and OPP6 are both considered overdrive modes (and both running 1.35V) - It's unfortunately not 100% clear if OPP5 is considered an overdrive mode for the 720MHz version as well - or if this only applies for the 600MHz version?

From the datasheet:

-------------------------------------------------------------------------

To avoid significant device degradation for commercial temperature

OMAP3530/OMAP3525 devices

(0°C < Tj < 90°C), the device power-on hours (POH) must be limited to one of the following:

* 100K total POH when operating across all OPPs and keeping the time spent at OPP5-OPP6 to less than 23K POH.

* 50K total POH when operating at OPP5 - OPP6.

* 44K total POH with no restrictions to the proportion of these POH at operating points OPP1 - OPP6.

To avoid significant device degradation for extended temperature OMAP3530A/OMAP3525A devices

-------------------------------------------------------------------------

Does this basically mean that running the 720MHz version at 600MHz is as "hard" for it as running at 720MHz? This would kind of suggest that the aging is mainly controlled by the operating voltage and not by the frequency - Would this be a correct conclusion? Or should the table be read in a way that it applies OPP5 for the 600MHz and OPP6 for the 720MHz version?

Best regards and thanks - Hoping that somebody can clarify

  Søren

Søren Steen Christensen - SSC Solutions ApS - Web: www.ssc-solutions.dk

  • Søren Steen Christensen
    Does this basically mean that running the 720MHz version at 600MHz is as "hard" for it as running at 720MHz? This would kind of suggest that the aging is mainly controlled by the operating voltage and not by the frequency - Would this be a correct conclusion?

    This has been my interpretation, and it makes some sense that the voltage increase is what is leading to the lowered expected POH for the device, though this is a less professional opinion and more of what I have seen from overclocking x86 chips personally (you rarely damage the CPU permanently from the clock increase, but raising voltage to allow the clock increase to stabilize can lead to damage and reduced life) which I suspect has some application here.

    If there was a difference in POH expectancy between the 720MHz and the 600MHz grade silicon at OPP6 and OPP5 I would have expected the datasheet to point that out explicitly (if I am incorrect here one of the factory folk watching this will correct me and clarify the datasheet). It is possible that the 720MHz grade will still outlast the 600MHz grade at the same conditions (i.e. 600MHz) on average, but I don't believe that we have testing to express that relationship based on the datasheet wording.

    In general, if you want the longest lasting devices, run at the lower voltages and frequencies, if you need the performance than run at the higher voltages and frequencies but keep in mind that you are statistically lowering the life expectancy of the part. Most applications I have seen use the dynamic voltage and frequency scaling capability of the device to only go into overdrive as processing conditions demand, which means most of the time they are not in overdrive anyway which gives them the full POH from the spec (and a much smaller power footprint).

    Søren Steen Christensen
    Or should the table be read in a way that it applies OPP5 for the 600MHz and OPP6 for the 720MHz version?

    I don't believe this is the case or I would have expected separate columns for the speed grades.

  • In reply to Bernie Thompson TI:

    Hi Bernie,

    Thanks for your quick reply - Sorry for not getting back to you prior to now. Your thoughts are basically similar to what I expected, so I'm glad that we share the same view on the topic :-)

     Best regards and Thanks
      Søren

    Søren Steen Christensen - SSC Solutions ApS - Web: www.ssc-solutions.dk