Welcome guest, is this your first visit? Click the "Create Account" button now to join.
Results 1 to 9 of 9
  1. #1
    GPSPower Helper


    Join Date
    Apr 2011
    Posts
    302
    Rep Power
    294

    Default 3490 X 3790 Benchmarks and more

    Many already know that the processor of 3400 family is different and faster that the 3700 family. The purpose of this topic is to show not only the difference in performance between these families, but especially try to find differences within the families themselves, based in judging the performance of several other gps. The more accurate and reliable for this type of test is to use the set of benchmark available on the Garmin 2012 firmwares in hidden menu Diagnostic Page. I'm not sure that this same benchmark is available in 3700, so I'm using the 3700 with the 3400 firmware to match the conditions.
    I hope that with the assessments of benchmarks we can prove or not if there are differences between the models (early, recent or of distinct regions, Asia, USA ..). If such differences may help confirm hardware differences not identifiable, which may have other implications, as to infer that such differences can also be related to the accuracy of the converted 3700 (real hardware difference l and not perceived), among other possibilities (gps malfunctioning...) we'll know after analyzing various benchmarks and for it is desirable the participation of stakeholders...



    for those who do not know:
    To access diagnostic pages: Press and hold your thumb on the speedometer trip computer page (press speed in the map)
    To capture the screen: Setting / Display / Screenshot (the images are in the Screenshot folder of the internal memory of the gps)
    Hardware Version: Setting / Device / About


    NUVI 3790 (-->3490) HW Ver.: V8 8GB 035 DA TC HP GM (SW 7.80)

    4kr

    4qln

    jsan



    NUVI 3490 - HW Version: V7 2GB 036 DS TC GTL (SW 7.90)

    70w


    pejl

    v1
    Nuvi 2689; 3490; Etrex 30; Oregon 200 (from Approach S5)

  2.    Advertissements


  3. #2
    Garmin/GPS Systems GMod.

    Neil's Avatar
    Join Date
    Aug 2011
    Location
    Oz.
    Posts
    7,398
    Rep Power
    1434

    Default

    Quote Originally Posted by angelozip View Post
    ..... I'm not sure that this same benchmark is available in 3700, so I'm using the 3700 with the 3400 firmware to match the conditions.
    ...
    In original 37x0, 'Benchmark' is available in 'Developer Tests' of Diagnostics Page. Is that what you're asking for Angelo?

    If you give me a 'heads up' on what you want etc, i'll do some tests for you on my 'virgin' [never converted] 3790.

    anv
    'Thanking Posts' are banned. To thank someone, and/or to see hidden links and content, use the [Only registered and activated users can see links. ] button below left of the helpful post then refresh your browser [F5 key]. 'Thanking Posts' are banned.
    Please don't spam. Posts serving no purpose other than to thank or to ask about hidden links are trashed or deleted, it's GPSPower's policy. Please don't spam.
    [Only registered and activated users can see links. ] should make their first post as a new Intro Thread in [Only registered and activated users can see links. ].

  4. #3
    GPSPower Helper


    Join Date
    Apr 2011
    Posts
    302
    Rep Power
    294

    Default

    Neil, as I have not been dealing with original firmware of the 3790 and was not willing to reconverted it at the moment, do not remembered if Benckmark of 3790 used the same basis of calculation of 3490, despite knowing of its existence .. But the important thing, in my opinion, would be able to have benchmarks of different units, 3400 or 3700, original or converted for purposes of comparison. For example, it would be interesting to know if your 3790, which seems to be the first series, has indeed differences in performace, for example, in relation to more recent models, that caracterize different hardware and could have other implications.. The idea is very simple, with no further pretensions. Obviously with concrete data we can better support some statements ..
    Last edited by angelozip; 20th June 2013 at 02:54.
    Nuvi 2689; 3490; Etrex 30; Oregon 200 (from Approach S5)

  5. #4
    Master
    Join Date
    Aug 2010
    Location
    Russia
    Posts
    130
    Rep Power
    71

    Default

    Try this on 3597. I have only Europe installed and see this result:
    LONG - about 9300, 5400
    Med - 1250, 1200, 1950
    Short - 70, 110, 50

    I think it doesn't work correctly without NA maps.

  6. #5
    Navigation software expert Kanopus's Avatar
    Join Date
    Jan 2010
    Location
    Internet
    Posts
    481
    Rep Power
    249

    Default

    It looks like routes on CN North America but since we don't know what map it was, we can't repeat this benchmark. Other factor could be whether map was on SD or internal flash.
    Thanks, Kanopus

  7. #6
    GPSPower Helper


    Join Date
    Apr 2011
    Posts
    302
    Rep Power
    294

    Default

    @Kanupus and ASTX,

    None of the gps that I took the test have CN North America installed, only the map of my country, nothing more. So understand that the data for route calculations are standard and are embedded in the firmware, but it may be different in other continents. Anyway if the system presents these routes and do the calculation shows that this embedded in the firmware, independent of maps (perhaps only of the base map).
    The problem in relation to the results is that they vary in the unit itself, just knew it. So the values ​​can not be taken as absolute and are not as stable as I thought. Either way can be good for something, I hope.
    My suggestion for the test: turn on the GPS, enable the GPS Simulation and wait a few minutes for all the internal processes of launch to finish. Thus the processor is completely released for testing. Then make several tests and select the one that has the best values​​.
    Following these directives, redid my tests and values ​​greatly improved. I will present only the first line of each group of routes calculated (the remainder improved in proportion):

    Nuvi 3790: Long: 3836 ms, Medium: 341 ms, Short: 63 ms
    Nuvi 3490: Long: 2231 ms, Medium: 208 ms, Short: 49 ms
    Zumo 350: Long: 2243 ms Medium: 210 ms, Short: 41 ms (350 i V6 2GB 64 MB)


    What I could observe from these measurements is that my 3490 has a much faster processor than the 3790, a relief, because I thought my 3490 was a 3790 remanufactured by Garmin or distributors, lol. Just regrettable that Garmin removed the magnetometer of my series. Does anyone have a 3490 with this item of hardware? (test in Diagnostics Page)
    From the result also gives to infer that the Zumo 350 probably has the same processor of the 3490, as the benchmark values ​​are almost the same, which makes sense in terms of production scale, cost savings and type of gps (road) ..
    Last edited by angelozip; 20th June 2013 at 17:04.
    Nuvi 2689; 3490; Etrex 30; Oregon 200 (from Approach S5)

  8. #7
    Master
    Join Date
    Aug 2010
    Location
    Russia
    Posts
    130
    Rep Power
    71

    Default

    Somthing wrong with it. Same result
    I can't believe that 3597 have a slower processor than 3790. Especially looking on screen when driving

  9. #8
    Navigation software expert Kanopus's Avatar
    Join Date
    Jan 2010
    Location
    Internet
    Posts
    481
    Rep Power
    249

    Default

    Quote Originally Posted by angelozip View Post
    None of the gps that I took the test have CN North America installed
    This test doesn't work correctly without a map, maybe it uses basemap or direct routes?

    CNNA 2014.1 on SD, nuvi 37xx converted to 34xx:
    Long NY->LA no highways: 125s, failure
    Long NY->LA: 20.5s
    Thanks, Kanopus

  10. #9
    GPSPower Helper


    Join Date
    Apr 2011
    Posts
    302
    Rep Power
    294

    Default

    Quote Originally Posted by Kanopus View Post
    This test doesn't work correctly without a map, maybe it uses basemap or direct routes?


    Really is need a map to calculations of Benckmark, but this map can and should be preferably the Basemap to match the test conditions when comparing GPS, reminding that with him the route calculation is Direct set. I also tested the benchmark with City Navigator North America 2014.10: the results were much worse and the calculations of more complex routes had flaws, eg, NY-> LA (no highway). With CNNA the speed of calculation certainly has to do with the complexity and type of route set (faster, shorter ..) and too with the speed of reading of the SD card (variable) among other factors, therefore, can not be standard for comparison purposes. In addition, after testing with the CNNA the benckmark stopped working. Only clearing the non-volatile memory all returned to work. With any other map installed, other than the CNNA, will be used the Basemap by the Benckmark. So if CNNA is present, suggest disabling it to better compare it to other devices.

    Thus, for the purposes of comparison on equal terms, recommend:
    - enable the GPS Simulation and wait a few minutes for all the internal gps processes of launch to finish and if the performance is very bad, so clear Non Vol Memory.

    The route that we may use would be for the comparative Benchmarking could be:

    Long: NY-->LA (no highway)
    Medium: Seathle --> Bainbridge Island (no ferry)
    Short: SFO--> Fisherman’s Wharf



    Quote Originally Posted by Astx View Post
    Somthing wrong with it. Same result
    I can't believe that 3597 have a slower processor than 3790. Especially looking on screen when driving
    Probably the test was done in different conditions. I believe the above explanation may help. No doubt that the 3597 has to be faster than the older models ..
    Last edited by angelozip; 24th June 2013 at 18:15. Reason: more clarity in the text
    Nuvi 2689; 3490; Etrex 30; Oregon 200 (from Approach S5)

 

 

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
This website uses cookies
We use cookies to store session information to facilitate remembering your login information, to allow you to save website preferences, to personalise content and ads, to provide social media features and to analyse our traffic. We also share information about your use of our site with our social media, advertising and analytics partners.