Billing headache is sensible reason. I still think 3G is likely very cheap being old tech and probably have better coverage during the Model S release cycle.
On 4G HSPA+, are we there yet? or still on 3G?
Yes. Depending on your definition of 4G.
Based on some sleuthing on
TM's forums, the chipset Tesla is using is the Sierra Wireless AR8550 (or at least was/is in North America). It supports HSPA+, HSDPA, EDGE, GPRS, and GSM. There is some conflicting information online with regards to what bands it supports; one source says 1700 (AWS) and AT&T itself says it supports 2100 MHz. This is important because the majority of HSPA networks outside of the Americas support 2100 MHz. (sometimes, AWS is known as 2100 because it technically operates in 1700/2100 paired spectrum, making things more confusing). I'm assuming they are using the ROW (rest of world) version of the chipset for Europe, the AR8552.
As far as speeds go, "HSPA" can be as misleading as "4G". The chipset above (if it's what they are still using) supports 14.4 Mbps on the downlink, and 5.76 Mbps on the uplink. Those are maximum speeds. Real speeds on a congested network will most likely be in the 1 to 3 Mbps range on the downlink, peaking at 6 Mbps is you're lucky.
So, it may be "4G" at it's maximum theoretical rate, but not in real use.
I do think and agree getting LTE is probably unnecessary and won't be beneficial for a number of reasons
- music+maps+telemetry don't need it. These are the primary functions in a car.
- browser is the only one that can use it and maybe video streaming if Tesla were to add it. I can think of 2 factors that prevents this area from improving.
I don't have my car yet, but I've read a lot of complaints about map tiles loading slowly. This is caused by network latency (see my other post) more than throughput. Google Maps uses a series of tiles images to display maps. I'd argue that LTE is necessary to improve this performance in any meaningful way. I personally don't care about video streaming, but I'm sure some others do (maybe while stationary/parked I could see this being handy).
As I said in my other post, LTE is less about today, and more about tomorrow.
A. smartphone/Tablet/computers's fast browsers comes from a ton of software tuning work by the browser developer. I doubt Tesla software team will dedicate the necessary effort to tune up the browser. Browser is also a constantly evolving standard with lots of plugins which all requires tuning as well. Takes a lot to keep up. I would guess the "browser" feature on Tesla will fade into the sunset (probably have already given its current capability) unless they piggy back onto standard products like Apple or Android.
Tesla is utilizing WebKit, which, is an open sourced project Apple's Safari browser is built on (and Chrome, previously - Google has since forked WebKit). So, Tesla is basically doing what you said: piggybacking on a standard product.
At first glance, the reason doesn't appear to be hardware or 3G
I'd argue that it's exactly hardware and 3G that is causing the poor performance (that, and as you mention, any "tuning" of their implementation of WebKit).
Making video go fast require dedicated hardware for video playback. The reason our latest smartphones+tablets do video so well is because they put great video hardware in it. The evolution was really fast. Video hardware improved by at least 2X/year in the last few years. Not sure what is the vintage of Tesla's entertainment computer but surely <= 2011 hardware given the car's design cycle. In addition, Tesla has no business evolving the entertainment computer rapidly as they should be focused on building great cars. It is impossible to keep up the smartphone/tablet's improvement rate anyway.
Tesla is using nVidia's Tegra VCM SoC to drive infotainment functions. The Tegra system-on-a-chip (SoC) integrates eight specialized processors, including a multi-core ARM CPU, a GPU, and dedicated audio, video, and image processors. So, the system has the dedicated hardware you speak of.
Even if we get LTE, Tesla's browser is likely still slow unless they put a lot more software work into it.
I agree 100% with this statement!
B.. Carriers fundamentally wants to see additional revenue streams for bigger mobile screens such as tablets. They probably don't want to give it away for cheap in cars.
Tesla is AT&T's customer in this case. AT&T is just providing a pipe, and have no interest in building additional revenue streams given the current model (although I'm sure they'd love to). Tesla's focus is different than AT&T's, and it's clear they're not overly interested in value added services in this realm - at least not yet. Maybe in 4 years time. I'm glad Tesla has abstracted the network provider from us. It shows their focus is in the right place, at least at the moment.