Is this why Apple needs its own 5G chips?

A flight of fantasy found on one of my favorite Apple investor boards:

Nobody has more chip design prowess on this earth! 2000 engineers on board from intel, who was very close to what aapl wanted or they wouldn’t spend 1Billion! And all them engineers will be led by the best chip design leaders in the free world, to quote Neil Young.

Me thinks you are all understanding the very high r&d increase. Aapl 5G modems equal to qcom by 2021 at the latest. And do not forget, the license for 7 years from qcom, Tim knew wtf he was doing.

Next up, you all do realize aapl has many very highly touted satellite rocket scientist onboard, right! Think 1500 5G satellites all aapl owned, and musk will place them with pin point precision, then it is T and VZ that are fubared. Ok, Sprint and tmobile, too

Yall do not understand the cost and time to upgrade their networks, while AAPL does… and AAPL has FU$$$$$$$$$$$$, and Tim et al long term plans make President XI in China, with their long term world domination, look foolish. AAPL long term plans extend decades, and they are laser focused. — “MacEntropist”

My take: I leave it to readers to unpack the acronyms and apply their own smell test.

7 Comments

  1. Mordechai Beizer said:

    Not a great idea. There’s lag due to speed of light. In a geosynchronous orbit you would have about 300ms transmission delay. Closer in there would be less transmission delay but it would be complicated to manage. Also, think of how many satellites it would take. In 2016 there were 300,000 cell towers in the US. How many satellites would it take to replace that?

    3
    August 9, 2019
      • Mordechai Beizer said:

        Irridium had world wide coverage and very few customers. The assumption here is that the fantasy 5G satellite network would support everyone with a 5G phone. And we’re talking data transmission, not voice. Irridium was designed for voice. You only need 64K capacity for a high quality voice transmission.

        Each original Irridium satellite could handle only 1100 simultaneous connections. Next generation handles more but nowhere near what would be needed for 5G service.

        0
        August 9, 2019
  2. Mordechai Beizer said:

    There are so many other things wrong this idea. One reason there are so many cell towers is to increase overall network capacity. Each connection takes a finite amount of bandwidth so that bandwidth can be re-used if the phone signal only goes a short distance. If the connection had to be maintained over a 100 mile distance (assume that the satellites are close in so that there is less transmission delay) then you would give up a lot of capacity over a signal that only has to be heard for a few miles.

    If my phone had to transmit 100 miles to reach a satellite, how strong a signal would that take and how much power. I don’t want my phone emitting that strong a signal while it sits next to the family jewels. Nor do I want to recharge every hour.

    Finally, if the satellites are close in then they would have to cover the entire world since the satellite mesh network would be constantly entering and leaving the area above the US. So if it were to take 3,000 satellites to provide coverage for the US (a number pulled out of my **s) then you would need 50 times as many satellites to provide continuous coverage (the US surface area is 4mi sq miles, the globe is 200mi sq miles).

    0
    August 9, 2019
    • Turley Muller said:

      Never work. Wouldn’t even be able to hold a conversation with latency that poor. You’re hearing what the other person said 5 seconds ago, The goal of 5G networks is sub-1ms latency. It’s the zero delay response time that’s the game changer. And, some of that involves moving centralized cloud data centers to the edge of the network. Anything that shortens geographical distance is always ideal. –

      0
      August 9, 2019
  3. Gregg Thurman said:

    What would happen if the carriers used the 700 (MHz?) band (that had been used by analog broadcast TV) to retransmit to satellites?

    The 700 MHz band can carry TONS of data (the lower the freq the greater the capacity over a very long, greater than horizon to horizon, range) with extremely low latency.

    This suggestion is getting out there and I bring it up because the carriers bought the rights to it, yet haven’t used it for anything as yet (afaik).

    0
    August 9, 2019
    • Turley Muller said:

      They have repurposed the 700MHz- all around that range – for cellular. LTE runs on it. It has the best range and propagation. Speed is based on the bandwidth or channel size- such as 700- 710. That’s only 10MHz. 5G is rolling on on the LTE bands (low freq) for coverage, which doesn’t offer any real improvement over advanced LTE, cause essentially the same thing.

      But for 5G mmWave you can have a signal that is 38,000-38,000 or 800MHz block, you know versuses the 10’s or 20’s available on the low-band. That’s just a product of there being so much in unused spectrum in the mmWave. Problem is he only travel few hundred meters and doesn’t like trees or windows. In addition, since antennas are directly proportional to the wavelength, you can have tons of antennas for mmWave which enables 256x MIMO. The low-band wavelength is 100’s times longer than the high-band. So, you can only fit about 2-4 antennas in a smartphone. Latest iPhones have 4×4 MIMO capable of gigabit LTE.

      Until the wireless carriers can get their hands on the 3.5GHz – 5GHz stuff (mid-band) for 5G, the low-band that TMUS, AT&T, and US Cellular (sub-1000MHz) are currently rolling out 5G on really will not be any faster. It’s intended for coverage due to the range and propagation. Mid-band is great balance of coverage and speed. Unlike the others, Verizon is holding off on nationwide 5G until MB is available, stating it isn’t interested in taking LTE network and rebranding it 5G and it not be transformative. In essence, advanced LTE = 5G E = low-band 5G.

      VZ is primarily focused on mmWave 5G which really only has industrial/enterprise applications- must easier to monetize but more costly to build. Opposite is true for sub-6GHz 5G. Any cell site that has been upgraded to the advanced classes of LTE in the last couple years will just require a software update for 5G. However, meaningful improvements won’t be seen until carriers can feed their networks more spectrum. FCC been working on freeing up some MB which has been slow. A promising solution is DSS- dynamic spectrum sharing where you can run 5G on any and all bands a carrier has. Spectrum bands are dedicated to specific networks- meaning more you allocate to LTE or 3G etc is less you have to use for 5G. With DSS, you can use bands from any network for 5G concurrently and bundle them together to create a much larger bandwidth carrier signal.

      3
      August 10, 2019

Leave a Reply