Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

This site may earn commission on affiliate links.
so far Tesla have said (quite recently) that HW3 will be capable of FSD, and about 200%-300% better than a human, with HW4 being about double that again.

the more likely scenario in my mind is that it will come down to regulators and the consequences should HW3 not meet their needs but HW4 would.

The main regulatory aspect here I think is redundancy. HW3 has traded redundancy for capability, which may mean full unsupervised driving will never be on the cards, whereas HW4 has restored that redundancy so might be fine in these aspects.

From a shareholder risk perspective, revenue has only so far been recognised on what has been delivered, so should there be any refunds, they would only hurt the cash side and not the financials. On the other hand, the small print has always been "subject to regulatory approval" so Tesla may be fine, even if there are disgruntled customers - I think should that happen then the ability to pass the FSD functionality/licence to a new HW4 Tesla might retain some customer satisfaction here.

Going forward, with rate of production increase, and assuming HW4 becomes common across all models, then by the end of 2024 there will be more HW4 on the roads than HW3 anyhow, so any potential downside risks and liabilities diminish significantly with time. Indeed should HW4 accelerate the FSD journey and make RoboTaxis happen soon, then there is only upside in my view.

Looking at this technicall - I'd hope TEsla has the ability to let FSD run on a unified 360° field of view that the neural nets are building with whatever camera stack is on the car - with this they could deal with all vehicles having different camare configs and only modifying the "360° stitching layer" for new layouts - this would mean that HW3 has less visibility range and angles but still same features on top of the stitching - so less reliability and assertness but still same features.

I hope they will tell us new radar is within image space, increasing forward detail and range but not different from the stack ontop of 360° view ... that would make the radar part clear as well.
 
Looking at this technicall - I'd hope TEsla has the ability to let FSD run on a unified 360° field of view that the neural nets are building with whatever camera stack is on the car - with this they could deal with all vehicles having different camare configs and only modifying the "360° stitching layer" for new layouts - this would mean that HW3 has less visibility range and angles but still same features on top of the stitching - so less reliability and assertness but still same features.
that would make a lot of sense.

with HW4 having more cameras and also likely to have higher resolution, it can see more/further within that stitch/range of view.

On that assumption the HW4 extra compute is needed only for the increase in cameras and resolution, rather than building in further new nets unavailable to HW3.
I hope they will tell us new radar is within image space, increasing forward detail and range but not different from the stack ontop of 360° view ... that would make the radar part clear as well.

My guess is that radar will give safer operations when view is impeded, such as in heavy rain, fog or snow. So would give fewer "FSD is unavailable" notifications - essential for RoboTaxi operations and longer term potentially moving things closer to level 5.

Will be interesting to also learn the range on the HD radar, i.e. whether it can see further than the forward facing vision systems, and also the placement of the sensor. We assume this to be under the bumper - but could windscreen placement alongside the forward facing cameras be an option too?
 
But...BMW has massive....

Was just about to say, when the "competition" looks like this and is becoming more and more bloated and disconnected from the driving feel, it can't be that hard to surpass them

1676569372961.png
 
Looking at this technicall - I'd hope TEsla has the ability to let FSD run on a unified 360° field of view that the neural nets are building with whatever camera stack is on the car - with this they could deal with all vehicles having different camare configs and only modifying the "360° stitching layer" for new layouts - this would mean that HW3 has less visibility range and angles but still same features on top of the stitching - so less reliability and assertness but still same features.

I hope they will tell us new radar is within image space, increasing forward detail and range but not different from the stack ontop of 360° view ... that would make the radar part clear as well.
Triggered a thought. What if the radar is not a sensor but just a bright light just outside human vision, but clear as day to the new cameras? Possible encoded flood light. Crazy?

(Edit: Not well thought out... the best kind! Radar: Radio Detecting and Ranging. This says nothing about what is receiving these pulses of radio. I bet it's POSSIBLE the cameras can see a lot more than we know here. Just a strobe here and there, to save power. Pure speculation on my part.)
 
Last edited:
I believe both FSD versions will succeed; newer version will always be safer.
I want to say Tesla redefined FSD a couple years back to be an ADAS and not true Robotaxi. Seems to me they are on track to hit that goal.

No Robotaxi for HW3 though so no big appreciation for vehicles which are HW3 capable. Also, unlikely full release will trigger a flood of new FSD purchases.

Not sure how many people have this big landslide of value appreciation modeled in, if you do, you might want to back it out. Aside from that, I think they will deliver what they’ve sold as FSD (recently). Not sure about those older cars where they essentially promised Robotaxi in the FSD sale. Fortunately that’s a fairly small percentage of vehicles.
 
  • Like
Reactions: SOULPEDL
CNBC trumpeting FSD software update as a recall🤦🏼‍♂️🤦🏼‍♂️
Stock seems to be dropping because of this. 362,758 cars being recaled 'by over the air update' as required by NHTSA because FSD may cause a crash due to allowing the vehicle to "exceed speed limits or travel through intersections in an unlawful or unpredictable manner increases the risk of a crash".
 
Stock seems to be dropping because of this. 362,758 cars being recaled 'by over the air update' as required by NHTSA because FSD may cause a crash due to allowing the vehicle to "exceed speed limits or travel through intersections in an unlawful or unpredictable manner increases the risk of a crash".
As expected, that ridiculous sell-off is already short-lived... buyers jumping in
 
Stock seems to be dropping because of this. 362,758 cars being recaled 'by over the air update' as required by NHTSA because FSD may cause a crash due to allowing the vehicle to "exceed speed limits or travel through intersections in an unlawful or unpredictable manner increases the risk of a crash".
If the premise is "may cause crashes" then wouldn't all self driving software be recalled from everyone?
 
Not sure about those older cars where they essentially promised Robotaxi in the FSD sale. Fortunately that’s a fairly small percentage of vehicles.
I think when those cars were traded in to Tesla, they would remove the original FSD, and if they add FSD again in resale it would carry the new definition.
So, I am actually wondering how many of those OG FSD are still out there never traded in to Tesla, and what would happen to them eventually.
 
If the premise is "may cause crashes" then wouldn't all self driving software be recalled from everyone?
Hard to know if this is an initial 'shot over the bow' from NHTSA's engineering review or just a simple and minor tweak to the software to resolve a particular identified issue.

Edit: Some more details here. Looks like a software tweak to FSD beta.

FpGzb71WAAA1fwp
 
Triggered a thought. What if the radar is not a sensor but just a bright light just outside human vision, but clear as day to the new cameras? Possible encoded flood light. Crazy?

(Edit: Not well thought out... the best kind! Radar: Radio Detecting and Ranging. This says nothing about what is receiving these pulses of radio. I bet it's POSSIBLE the cameras can see a lot more than we know here. Just a strobe here and there, to save power. Pure speculation on my part.)
Completely wrong wavelength.
Cameras can see near-IR if not filtered, but that would blind oncoming cameras.
 
  • Informative
Reactions: wtlloyd
I want to say Tesla redefined FSD a couple years back to be an ADAS and not true Robotaxi. Seems to me they are on track to hit that goal.

No Robotaxi for HW3 though so no big appreciation for vehicles which are HW3 capable. Also, unlikely full release will trigger a flood of new FSD purchases.

Not sure how many people have this big landslide of value appreciation modeled in, if you do, you might want to back it out. Aside from that, I think they will deliver what they’ve sold as FSD (recently). Not sure about those older cars where they essentially promised Robotaxi in the FSD sale. Fortunately that’s a fairly small percentage of vehicles.
You could be right on newer FSD as ADAS.
I still have the original FSD purchase on the oldest Model 3's (unless I accepted newer terms through the car?). Maybe these early ones are a bit sticky legally, probably not too many considering.
 
  • Like
Reactions: navguy12 and Ogre
This was already done at the last earnings call. And I think they squared that circle pretty sufficiently:

"Hardware 3 will not be as good as Hardware 4, but I'm confident that Hardware 3 will so far exceed the average -- the safety of the average human. So how do we get ultimately to -- let's say, for argument's sake, if Hardware 3 can be, say, 200% or 300% safer than human, Hardware 4 might be 500% or 600%. It will be Hardware 5 beyond that. But what really matters is are we improving the average safety on the road."

Yes, HW3 is inferior to HW4, but both will be safer than human drivers. If Tesla can deliver FSD with HW3, there are no broken promises and no need for an expensive retrofit.
Agreed. Purchasing a Tesla is purchasing a computer. Everyone knows that the next major revision will be better than the last, have more functionality, and will not necessarily be compatible with some of the new software/drivers/hardware. Pretending otherwise or being unhappy that this happens is not very realistic.
 
Hard to know if this is an initial 'shot over the bow' from NHTSA's engineering review or just a simple and minor tweak to the software to resolve a particular identified issue.
It's a rabbit hole when you can make a case that every wrong maneuver from FSD beta can increase the chance of a crash. Maybe they are nerfing the ability for the car to go above the speed limit.
 
Stock seems to be dropping because of this. 362,758 cars being recaled 'by over the air update' as required by NHTSA because FSD may cause a crash due to allowing the vehicle to "exceed speed limits or travel through intersections in an unlawful or unpredictable manner increases the risk of a crash".
This concerns me.

Our society has normalized over-the-speed-limit driving. Seems like NHTSA is slowly making FSD unappealing by legal nannying it to death.

Are Blue Cruise and SuperCruise put under similar constraints? For that matter every cruise control mechanic on the planet. Most of modern bersions are aware of speed and this constraint could easily apply to them. There is no speed limit above 75 MPH in the US…
 
Last edited: