27 thoughts on “WearOS – The Next Generation Sports Watches Start Now – What? Really?”

1. It seems that the gps is single frequency, because if it was dual frequency they would make big claims.
   I’m curious to see if they can use simultaneously all the 4 satellite constellations (gps, galileo, beidou and glonass) or if the user must choose a maximum of two constellations, and to see how the terrestrial signals (cellular and wifi) will help gps.
   Do you know when will appear the first WearOS smartwatches with this new Qualcomm platform? This year or next year?
   They are also using the 28 nanometers production process for the chipset, like Broadcom with the BCM47755. Let’s see if Mediatek is following…

   1. Montblanc Summit 2 is due out next month. In the press slides, they sent out they hinted that the sports mode feature will be rolled out with a future sports brands. When or who that is who knows.

      1. I glanced at that but missed the sports mode inclusion…this is getting annoying for me now as I’m going to have to start looking at LOTS more devices that call themselves sports watches grrrrr 😉

2. There isn’t as much new here as the marketing may suggest. Processor throttling is nothing new and older chips have had this for decades. Smaller manufacturing process is more of a gain for battery these days as we’ve hit the limits of what throttling can do to help. This one is trickier to gauge because the same ARM chip may be available in a 7nm process and a 13nm process, with the latter drawing far more power for the same result (although likely cheaper, since the fab will charge a premium for new processes).

   1. Interesting, so the future is around the corner.
      They could in two or three years produce the GNSS (GPS) chipset with a 7 or 13 nanometers production process instead of the 28 nanometers they are now beginning to use.
      The GNSS (gps) chipset could be MORE powerful with the same power consumption.
      Maybe when Galileo will be ready by 2020 with the E6 frequency and his free PPP service with 20 cm accuracy (if it will be free without commercial fee), there is a possibility that PPP (Precise Point Positioning) or RTK enabled gps receiver will be available also for wearables and sport watches?
      Yes, I know the5krunner, that for instant pace Stryd it’s easier and there is no need of PPP for sport watches 😉

      1. the future is always round the corner 😉

   2. Maybe the new 28 nanometer production process is responsible also of the long battery life of the Suunto9.
      Should we expect the new 28 nm production process also for the other watches, and so long battery life also for the “soon to be released” Polar Vantage and for the “hopefully to be released in 2019” Garmin Fenix?

      1. I mean Garmin Fenix 6

      2. Intel had 22nm in 2011 so don’t think of 28nm as new, it’s very, very, very old in chip fab terms. In other words, it’s likely that we could make a watch with 100 hour battery today quite easily. You’re on the upgrade treadmill and it has years of life in it yet. Every two years you’ll see a “new” processor with “better battery life”. With current technology available today that can last for over 10 years of “upgrades” compared to this 28nm part. 22, 14, 10 and 7nm are available now with 5nm planned by 2020. A good example is the lack of memory in the Fenix 5 – the primary reason the 5 Plus was an upgrade. We all knew that the Fenix 5 should have had more when it was released based on current tech at the time. Never trust a company with good sales volumes to innovate.

      3. sony chip + new algorithms i reckon play biggest part.

   3. yep it’s mostly just better battery life stuff. is it 13nm? i didn’t even thnk it was that 28 IIRC???

      1. In Qualcomm PDF file they say that the wearos platform is 28 nm.
         Real products in shops are always behind the more recent production process.
         It would be nice to have now a superpowerful 11 or 13 nm fenix6!

      2. Still better a 5 nm fenix6

      3. Sorry, completely wrong information from my side. I believed that Qualcomm WearOS 3100 was 28 nm, but I think I remembered wrong. I controlled again and I couldn’t find any indication about nanometers in the pdf file. I tried to find correct information, but in these days the brands give very few technical details. In short: I don’t know what is the production process of the Qualcomm WearOS 3100 (I don’t know the number of nanometers) and I can’t find it anywhere.
         Sorry again for the wrong information.

3. https://www.eetimes.com/document.asp?authorpage=0&authorpage=0&authorpage=0&authorpage=0&authorpage=0&authorpage=0&isAjax=true&isAjax=true&isAjax=true&isAjax=true&isAjax=true&isAjax=true&doc_id=1325498&_=1536930999738&piddl_msgpage=1#msgs
   Title of article “Sony Joins FDSOI Club” of Junko Yoshida date 1/30/2015
   Interesting article, especially the comments, where they explain that the new Sony GNSS chipset present in the new Suunto 9 and the new Polar Vantage is 28 nm but uses a process technology called FDSOI (Fully Depleted Silicon on Insulator), that is different than the process used by Intel (FinFET).
   In the comments we can read that FDSOI is particularly good for low power (and wearables) because can deliver 0,6 V, while FinFET can’t give this low voltage.
   There are a lot of comments and the users didn’t agree, but it seems that the advantage of FinFET is that it’s easy to produce also in 22 and 14 nm (Intel is doing it).
   The advantage of FDSOI is the low power and the low voltage, but it seems very difficult to go under 28nm, one user claims that it’s impossible to go under, other users don’t agree.
   So maybe we can’t really compare just with nanometers FDSOI and FinFET, because are different technology.
   In this article about Qualcomm Wearos 3100
   https://linustechtips.com/main/topic/970558-qualcomm-announces-new-snapdragon-wear-3100-for-wear-os-by-google/
   they write that
   1- about the main processor: “So from these two quotes we know that unfortunately the main application processor will be the same quad core A7 design used on the SDW2100 paired with the same GPU as before. That’s a little bit of a letdown. We don’t know for sure that it’s still using the same 28nm process, but based on it being clocked at the same max of 1.2GHz and having no mention of being a Custom or Semicustom design, it’s pretty safe to say. ”
   They are nor sure about the nanometers of the main processor, but they think it’s 28nm.
   2- about the co-processor : ” It’s a Semicustom Cortex M0 meaning that it should be *crazy* power efficient. Some other interesting details. This is from the SDW3100 Product Brief. ”
   They don’t say anything about nanometers of the co-processor.
   3- about the GNSS: ” And the second letdown is that the LTE, Wifi/Bluetooth, RFFE, and GPS/G/B/G radios/controllers are all the same ones in the SD Wear 2100.So that all sounds pretty disappointing so far, but thankfully that’s not where the big focus of these new chipsets have been so it’s not as bad as it sounds. ”
   They don’t say anything about the nanometers of the GNSS chipset. We just know that it should be the old “Qualcomm IZat Gen8c”, and we know that this chip was in the Qualcomm Snapdragon 820 that was 14 nm FinFET. So maybe it’s 14 nm. Probably it’s nothing new and nothing to be exited about.
   But the Qualcomm Snapdragon 845 is 10nm FinFET, so maybe they could in the next generation produce a new Gnss chipset IZAT with 10nm. Probably the GNSS chilpset of the WearOS 3100 is NOT 10 nm FinFET , so there are still margins for greater battery or greater power.

   1. In the first article, the user AKH0 wrote that also with the FDSI technology, that is particularly good for low power devices, it is possible to go under 28nm: “There is in fact interest from several fabless companies for a 22nm or 14nm FDSOI technology. 14nm prototype chips are in the line”

      1. have you seen anywhere where it specifially says which sony chip suunto are using? (polar then probably will use same)

      2. My starting point was your website, I hope you have guessed 🙂
         Unfortunately I don’t have any new information.
         Then I looked at the Sony website https://www.sony-semicon.co.jp/products_en/gps/index.html , but in this site they don’t say which chip Suunto is using.
         If it was the 5603, it would be better, because we can see that it has
         GPS/GLONASS
         Galileo/BeiDou
         I don’t understand if it has maximum 2 constellations and you can choose between GPS/Glonass and Galileo/Beidou, or if it can use all 4 constellations. It is a low power chipset, so I think max 2 constellations.
         The 5600 has only GPS/GLONASS, in an article of 2015 https://tech.nikkeibp.co.jp/dm/atclen/news_en/15mk/020900363/?P=1 we can read that the CXD5600 is older (year 2013) and the 5603 was just announced in 2015.
         Sony in their website writes that 5600 “power consumption @ continuous tracking” is 16 mW (probably GPS+Glonass, should be 10 W only GPS), 5603 6mW (maybe with all 4 constellations), so for sure less than half the power consumption (in the article they say 1/10 power consumption).

         I see that the Suunto9 has only GPS and Polar Vantage GPS+Glonass . If it is the 5600 chipset, the Suunto9 and the Polar Vantage probably won’t ever have Galileo.
         If it is the 5603 chipset, maybe the Suunto9 and the Polar Vantage will have Galileo with a firmware upgrade.

         All the other websites say just Sony chip, but no one tried to guess which chip, you are the only one I think. By the way, I find that is not correct that brands don’t declare to the buyers what they are selling. I remember years ago when I had to buy a personal computer of HP or Dell, I could read all the specifications if I wanted in a clear spreadsheet. But with this new wearable technology, all we can do is just to try to guess.

      3. I asked Polar which one. And was told that the person i asked DID NOT KNOW IF THEY WERE ALLOWED TO SAY (by Sony)? which was unusual. anyway. i only mention that to highlight there may be commercial restrictions either side on not disclosing it.

      4. I’m not a lawyer or a jurist, but I think that Polar has at least a moral obligation to say his customers which chip is inside. It’s like when I go to the supermarket, I have the right to know if I’m buying apples or oranges.
         Then if Sony made a custom chip just for Suunto or Polar, they could at least say : “It’s a custom chip” and not declare anything else.
         I don’t think that all this mystery about the chips that are inside a commercial product is commercially correct either.

      5. yeah, but do you know which pesticides/fertilisers were used on which oranges and which apples.

      6. The starting point was your website, I hope you have guessed 🙂
         I have no new information.
         Then I just looked at the Sony website https://www.sony-semicon.co.jp/products_en/gps/index.html and I saw that there are currently just these two chips.

         The chip CXD5600GF is of 2013 (power consumption@continuous tracking 16mW with 2 constellations GPS+Glonass)
         The chip CXD5603GF is of 2015 (power consumption@continuous tracking 6mW, I don’t know how many constellations, maybe 2 or maybe 4)
         The first doesn’t support Galileo and Beidou, the second yes.
         It would be nice to know which chip is inside the Suunto9 and the Polar VantageM, because if it is the 5603 maybe these watches will have Galileo with a future firmware upgrade.

      7. Always about argument nanometers, the chip of Mediatek Helio P23 for smartphone was 16 nm FinFET.
         The new Helio P60 for the future smartphones is 12 nm FinFET, Mediatek claims “Improved power savings: Compared to MediaTek Helio P23, P60 delivers up to 25% power savings in heavy duty gaming situations and boasts overall power savings of up to 12 %”
         If Garmin will use a similar gnss chipset for the fenix6, they will have also long battery life (but no dual frequency SIGH)
         Multi-GNSS: GPS/GLONASS/Beidou or GPS/GLONASS/Galileo co-receive
         http://i.mediatek.com/p60

      8. slide 16 https://www.u-blox.com/sites/default/files/documents/Analyst_day_2017_website.pdf
         It seems that u-blox still uses 28 nm FinFET technology, if I understand correctly.
         So it is possible that it can produce less power hungry gnss chipset.

      9. I guess if the 5600 is similar in consumption to sirf chips then it MUST be the 5603 that gives the power saving.

      10. I hope that you are guessing, 5603 would be good, ready for Galileo and Beidou 🙂
          This would clear my concern about the absence of Galileo in the Suunto and Polar.

      11. Without power saving mode:
          1- Sony 5603 is 6 mW @ continuous tracking
          2- Sony 5600 is 16 mW @ continuous tracking
          3- Mediatek MT3333 (in Fenix3) is 27 mW
          4- Sirfstar IV is 80 mW
          5- Sirfstar V is 90 mW
          So I would say that both the Sony are good candidate to be inside the Suunto9 and Polar Vantage.
          The Sony 5600 is older but probably costs less, so it could be better for a mass market product. The problem is that the 5600 does NOT have Galileo and Beidou. We should further investigate: I ask the help of experts 🙂

Comments are closed.