Memory size limits

Any chance the data from this chart for the Biker SDK could be updated to include the Edge 820?
http://developer.garmin.com/index.php/blog/post/connect-iq-2-biker-monkey

Or maybe it would be more useful in the long run to have that information on the device page:
https://developer.garmin.com/connect-iq/compatible-devices/
That way as new devices come out then we can see the memory limits in a central place. (thouh that page looks like it hasmany errors with regard to "Available ANT+ Profiles" like the 735 not supporting power)

Also,is there any indication of CPU speed? As in from a connect IQ app perspective how to tell how much processing power is available. For example I'm guessing the Edge 1000 has much more processing power then the 920XT but how does the Edge 820 compare to the Edge 1000? So far only seen a teardown of the FR 230 to know what CPU is in that so no way of knowing CPU spec differences.

Thanks
  • Following this thread, to me the best test would be a watch-app and not a data field. As a data field, it's dealing with things like compute() only running every second, and other things happening in the watch/FW at the time for the activity.

    IMHO, Let the test run in a watch-app and see what the numbers are, as that reflects what CIQ sees when the FW is running at full speed with only that app.


    I agree with Jim. I got different results in my watches than the ones in the app description (sometimes 10% to 15% lower), the result also changes depending on the number of data fields on screen (https://apps.garmin.com/en-US/apps/91534331-1228-47cd-8156-68458b597a7d)

    I would love to see the results in an app.

    Anyway, just FYI, my app crashes in the 235 only when I save a session and start another immediately (that is required for the multi sport feature). The same code runs perfectly in other devices. I will never know if the issue is related to CPU, IO or something different in the firmware. Right now I simply disabled this feature in my app for 235 users.


    Cheers,

    Robinson
  • Watch-app can be applied only to several devices. Data field supported by all the product range. I tested the code as watch-app and found exactly the same numbers.
    Pi calculation test performed at init, so every second only texts change, no calculations.
    Different values depend on timer precision, that is why they may differ from time to time - OS switches between processes and can handle, i.e. garbage collector. It is not possible to re-run tests or run it longer to have more statistics and averaging - Garmin terminates long-running app (watchdog timer).
  • Former Member
    Former Member over 8 years ago
    In an app, you could start a timer, and run the computation each time the timer triggers, and continuously accumulate a run time. You will only be terminated for executing for too long on a single event, so triggering repeated timer events allow you to continuously compute without tripping the execution limit.
  • Watch-app can be applied only to several devices. Data field supported by all the product range.


    Actually, watch-apps work on all the watches, and for the non watches, watch-apps are in the works for some or all of them.

    In the SDK 2.1.x announcement, it does talk about apps on Edge devices

    http://developer.garmin.com/index.php/blog/post/connect-iq-2-widgets-and-apps-on-edge
  • Ok. I've uploaded an app for that with the same functions. :)
    https://apps.garmin.com/en-US/apps/7d836871-cf67-4963-b56b-1d81a7808ba0
    Timer idea is good, I will try with the next version.
  • I'm running the app on a va-hr with the 2.74 FW and the 2.1.1 CIQ VM and here are my numbers:
    5.7/21.3 Pips

    How is that comparing to others that have run the app (not the data filed) on other devices?
  • I've tested and got exactly the same numbers for FR235 and epix. I will find VAHR to test with 2.74b, but I see your score looks the same too.
    epix 3.20, Connect IQ 1.2.1 | 1.0/15.4 Pips
    forerunner 235 5.07b, Connect IQ 1.3.0 | 0.4/10.0 Pips
    vivoactive HR 2.71b, Connect IQ 2.1.0 | 6.1/21.3 Pips


    I propose to switch all discussion to the https://forums.garmin.com/showthread.php?357784-App-Connect-IQ-benchmark thread.
  • This is pretty cool.

    FR 630, fw 5.4, CIQ 1.3.1, 0.5 / 8.8 Pips
    Epix, fw 3.2, CIQ 1.2.1, 1.0 / 13.3Pips

    Interesting how much faster the newer devices are. Fenix really does need an update.

    Another interesting test would be graphics speed. Drawing text and/or lines.

    On a different subject, does anybody know if there is a table showing how many data fields you can have on specific devices?
  • Another interesting test would be graphics speed. Drawing text and/or lines.

    see https://forums.garmin.com/showthread.php?358810-Tree-Benchmark
    On a different subject, does anybody know if there is a table showing how many data fields you can have on specific devices?


    It's in one of the appendixes in the UX guide that's part of the SDK.