I'm not confused. You brought it up for no reason. Whatever the reason was, I hope I helped you understand something new because the only possible reason I could imagine you bringing it up at all was because you didn't know already. And now you do. You're trying to keep the mistake relevant by suggesting that Tesla create a new MCU3 that is a sort of hybrid of MCU1 and MCU2 but to my knowledge nothing like what you're suggesting has ever been seen and you're the first to make that suggestion. Can you link to where you read about or saw the new hybrid you're talking about?
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Tesla infotainment system upgradeable from MCU1 to MCU2
- Thread starter MurrayJimW
- Start date
SucreTease
Teslarian
My original response to you was to correct a erroneous claim that you made. I am not claiming that such an "MCU3" "hybrid", as you call it, exists; rather I am only stating that such an approach is technologically possible. The fact that you have not seen one is not evidence that such a part is not being developed, nor is its absence evidence that your erroneous claim is true. There are a lot of people on these forums who make statements that purport to be factual statements, but are, instead, mere opinion.
@wk057 or @verygreen could probably confirm if they wanted to but my understanding from digging in and chatting with others is that it would be plug and play right now if you swap both MCU and IC. With MINOR software tweaks it would be plug and play for just an MCU swap too. No need for a hybrid part.... the IC communicates with the MCU over the ethernet bus which still exists in the same way on MCU2 as it does on MCU1, it's needed for other things like communicating with the gateway and the autopilot computer. As far as the new bluetooth and WiFi features go well those just wouldn't work. So no 5Ghz WiFi or bluetooth phone as key...
The reason this doesn't happen is just like others have said. Service is already inundated and can't take any additional load. I'm pretty sure they are still trying to get enough mobile techs ready to do all the APE 3 retrofits.
So... could they... yes
Should they? yes
Will they? probably not
The reason this doesn't happen is just like others have said. Service is already inundated and can't take any additional load. I'm pretty sure they are still trying to get enough mobile techs ready to do all the APE 3 retrofits.
So... could they... yes
Should they? yes
Will they? probably not
new IC2 does not need ethernet, it is jsut a dummy screen and the same 4 wire cable is repurposed as lvds screen signal carrier instead. This is why you need to replace the IC in addition to MCU
@verygreen please help kill the idea that mcu1 and mcu2 are different resolutions. I know this myth was started based upon the MCU unit browser self-reported data, but I believe you've confirmed mcu2 lcd screen is actually exact same resolution as mcu0/1, right?
not over that lvds port I imagine, but in theory you can make the two talk. Tesla would need to do the fixing though.
yeah, it's part of the firmware. 1200x1920 on the big screen (both mcu2 and mcu1), 1920x720 on the IC.
ucmndd
Well-Known Member
Makes perfect sense. There has never been any indication throughout yellow-screen-gate that the physical LCD assemblies for MCU1/2 are any different.
I think the confusion some have is that in the model 3 they changed the orientation and used 120px for the controls. This left 1920x1080 for the "screen". This was done on purpose I imagine to facilitate Netflix/other streaming service playback. S and X on the other hand will have no such luck... Gonna be a small viewing window lol
I don't have to worry much about that though. I highly doubt they enable it for MCU1.
I don't have to worry much about that though. I highly doubt they enable it for MCU1.
Only if they didn't also provide a connector for connecting it to an Ethernet bus.
I would bet he has already taken a good look at it given his other work/research.
Isn't it most likely that they just leave the ethernet API used by the IC1 on MCU1 as it is in the MCU2 fw-build?
API will be unused in the native MCU2 cars, but allows an MCU1 -> MCU2 upgrade without replacing the IC too. Most likely the API is already there on MCU2 or it's a simple matter of removing some #ifdef's.
no. the current build does ic display from MCU itself as a second screen. they reused the same connector that was used for ethernet before for the display interface now so you can just pop out mcu1 and ic1 and pop in mcu2 and ic2 and not mess with the harness at all (also makes for a very small change at the factory line - pretty much just make sure you switched both parts)
SucreTease
Teslarian
Doesn't that connector/cable plug into the MCU? If so, it implies that it would be feasible to build an ugrade-version of MCU2 with an IC1-compatible connector/interface that drives IC1 in the same manner that MCU1 did, which, in turn, should enable merely swapping out the CPU daughter board. Little new software needs to be written, as the original display driver could be used with this ugrade-version of MCU2.
mcu1 did not drive IC1 though. IC1 is its own linux computer that talks to mcu1 over ethernet.
as such I am not sure what you are talking about as the daugherboard. if oyu mean mcu1 tegra daughterboard to be replaced with an atom one - that's probably technically possible, but I don't think anybody seriously plans to do anything like it.
I've probably said this before, but I think that any upgrade path that requires any non-trivial amount of engineering work is just not going to happen. We can come up with all sorts of clever technical ways of making MCU2 drive IC1 the same way as IC2, etc. etc. etc. but any approach that requires product development is going to be seen as a distraction from other things that Tesla has to do to ensure its survival and long-term viability. Moreover if these upgrades do happen, they'll be creating new, little-tested car configurations, and I'd rate the potential for bugs higher compared to standard factory builds. I know that Elon has said it's possible, but just saying it doesn't make it so.
Just my opinion. I have no insider information, and I hope I'm wrong.
Bruce.
Just my opinion. I have no insider information, and I hope I'm wrong.
Bruce.
I remember hearing someone mention in an MCU1 related video it should be possible to simply replace MCU1s Tegra3 by a Tegra4 pretty much without any software changes.
Begs the question how complicated that would be too put in/onto the MCU1 circuit board, as my assumption is there aren't Tegra4 boards out there which are identical to the MCU1 board?
And even if, would the difference in performance between Tegra3 and 4 even make a noticeable difference?
Begs the question how complicated that would be too put in/onto the MCU1 circuit board, as my assumption is there aren't Tegra4 boards out there which are identical to the MCU1 board?
And even if, would the difference in performance between Tegra3 and 4 even make a noticeable difference?
The strange hangup that you need to make mcu2 work with IC1 is so strange. you jsut replace both and there's no problem. yes, that's like $300 more, but the customer is paying so who cares.
Similar threads
