I've been working to develop and populate my new EV comparison site for true nerds:
evbench.pluginpath.com
It's still a rough work in progress as I feel out what questions it needs to answer, but I would love feedback on any and all aspects!
122
Agreed. Unsure if it's native in the model or they're doing something "simple" like rendering a 3D sphere and then having an image model make it 3D.
8
Never heard of DLSS?
Gaussian splatting doesn't morph shapes like this, typically, and it can't fill in information on the back sides of objects.
It may be using gaussian splatting as a basis, but (IMO) it's *definitely* got an AI rendering over the top.
1
24
Replying to @mweinbach @brandenflasch
That was my first guess, but the weird morphing of objects and shapes makes me thing it's a high speed AI generated image/video based off of the live feeds from the cameras.
1
118
Cost savings is $0.4 to $1 per load on drying (~3kWh less).
And it may cost process time, but it saves human time (and a gating one at that!) - no transfer step.
So if you average 1 load of laundry a day, you're saving $150-300+ per year. And you don't have to go transfer to progress from cleaning to drying.
1
Replying to @JordanEVGuy
Reasons I love driving an EV:
No need to "warm up"
Instant acceleration
Fully charged every morning
Quiet & smooth
Reasons I miss ICE cars:
Manually shifting
V8 rumble
22
Yes.
"Tested range" is using (in this case) the recent OoS 70mph range tests, corrected to the same temperature.
1
1
24
"everyone ignores"
Literally every time I see anyone bring up these that's the first complaint.
1
13
Replying to @AC__3_3 @elektrowoz_pl
I've already accounted for the shape and efficiency in my prior post.
And I specifically noted we're talking about driving range/time.
The iX5 has solid driving assistance on the highway, but yes FSD is better.
5
Replying to @AC__3_3 @elektrowoz_pl
Bad take (on the charging and range front).
Even with slower charging and much worse efficiency, the iX5 still adds range MUCH faster than the Model Y.
Puke green line is the iX5 with a derated charging speed
Range normalized charging speed (image 1)
Range added vs. time (image 2)
1
17
Replying to @itskyleconner @elektrowoz_pl
Is it though?
I guess I'll have to watch the 6-hour, 3 part series in the coming weeks.
😅
74
Replying to @Mikekantorski
EVs should have a one-button "charge to 100%" in the app, the car, and on the charge port.
Then you can hit one button when you're plugging in or if you realize you need the full pack the next day.
The Rivian and Tesla apps are *pretty* easy, but it could be frictionless.
77
Replying to @R1Stephanie
If you're building a dressable base, definitely Santa Gary.
Otherwise, I think the snowman Gary will be more popular with people not "in the know".
1
1
14
Replying to @Joey_Bagadohnut @AustyUSA
The problem is that now the dots don't correspond to the location. They may appear close to or further from your route than they actually are.
You also don't get a visual sense of density. Just that there are *some*.
1
6
When rendering at 30fps, the minimum theoretical frame latency is 33.33 ms. At 60fps its 16.67 ms.
9
Don't let it get you down.
teapaw.org/pluginpath/status/2124…
It doesn't matter that much.
[But it still sucks!]
Replying to @itskyleconner @AlejandroEV66
So, I re-ran the sim using the derated profile from @AlejandroEV66 scaled up to the iX5's battery:
It's still competitive and makes a joke out of the Model Y L's charging.
Even normallizing for efficiency (image 2), it doesn't charge more slowly than the Y L until 60%.
The main impact is that it makes it more like the Model Y - optimal charge stops are no longer ~200mi, but ~140mi (image 3).
This changes the optimal strategy, but overall time is only impacted by ~30 minutes in the ideal case (image 4). Still over 2 hours ahead of the Model Y L.
[Yes, I know - FSD - we're just worried about the nerd range and charging data here!]
1
10
Replying to @itskyleconner @AlejandroEV66
So, I re-ran the sim using the derated profile from @AlejandroEV66 scaled up to the iX5's battery:
It's still competitive and makes a joke out of the Model Y L's charging.
Even normallizing for efficiency (image 2), it doesn't charge more slowly than the Y L until 60%.
The main impact is that it makes it more like the Model Y - optimal charge stops are no longer ~200mi, but ~140mi (image 3).
This changes the optimal strategy, but overall time is only impacted by ~30 minutes in the ideal case (image 4). Still over 2 hours ahead of the Model Y L.
[Yes, I know - FSD - we're just worried about the nerd range and charging data here!]
30fps to 60fps is an improvement of 16ms. It just makes it a lot more enjoyable.
But the point wasn't that this adds latency (which, it does, painfully - watch the LTT video).
The point is that adding frames never *improves* latency like native rendering does.
I'd never trade any latency for more frames. I'm playing, not watching a show.
1
17










