Update Speculation thread

In my opinion, electric Pigeon and Wigeon are logic

 

I did a little bit of performance testing in the dx11 version of Beamng and concluded that with traffic enabled, the GPU is the biggest bottleneck. I have a 9900k i9 processor and a 2070s GPU 16gb ram and ssd, about whats realistic for a mid-high end gaming PC for the next few years. I Can manage about 15-16 cars with traffic and traffic pooling at mostly stable 60 fps, In my test scenario i was CPU bottlenecked, as i used lowest settings and resolution, and the FPS was higher (100fps) when the Physics are paused. The recent CPU performance improvements with vulcan are great, dont get me wrong, But the potential GPU performance improvements need to at least match the CPU improvemts to benefit an actual user. So here is hoping for that to happen.

 

That depends largely from the map and even spot on the map, difference between Vulkan and DX11 is going to be greatest on areas of map or maps using huge number of static objects, especially ones with separate textures each.

General rule for DX11 is more texture maps = more drawcalls = more CPU limited. General rules are annoying as they are so wildly inaccurate in some situations, but there is plenty of situations where your CPU is limiting factor in DX11 and with Vulkan you get to be GPU limited and with faster GPUs more so.

Stuff becomes funny with cars, that uses several layers of textures, well used to use, then there are no LODs and with different vehicles you don't gain from batching, so when using single type and model of vehicle, you get batching, which means CPU is used less for graphical stuff, but when using different type of vehicles you don't get batching to help and CPU is used more.

So Vulkan helps a lot with huge number of different kinds of vehicles to run at map locations with lot of static objects, like Italy towns.

Another part where it helps is with lot of CPU cores, like 16 cores / 32 threads that have not been possible to be fully utilized.

Easy rules of thumbs are somewhat annoying as they are always wrong in complex matters like performance limits.

My gtx1080 can be 30% usage or 100% usage (watts) it really depends so much from the map, spot on map, which vehicles I spawn, but with Vulkan that 30% never happens and can get more than twice the FPS in CPU bounded situation.
With another lower spec system difference would be even greater and with my R9 5900X system probably lot less as that is kinda beast compared to old Intel systems.

But yeah, GPU optimizations are going to be thing too in a future I guess, shader stuff and what gets rendered, playing with Intel graphics profiling tools years ago I noticed that in Beam more of what is behind of camera got rendered compared to some other DX11 games, not sure if it is same still.

Not sure if LODs for the cars will happen, that might help with massive gridlocks, but how it would really work out when damage moves vertices based on node position, after all less vertices on lower LOD would just be spiky mess, I guess.


Sometimes hotfixes bring some new content or configs, I think that Wigeon might have possibility to get some extras?

 

But who will be the one to do it

 

This can be arranged...

 

I’d be willing to give it a try although I haven’t had luck with electric swaps in the past

 

So it is possible we may see a Wigeon with an electric motor in the next update/hotfix. Not guaranteed yet however.

 

No, just because something is possible to do doesn't mean that it will be added as vanilla content.

 

Because we have new vehicles coming more than before, How about something like the Hirochi AT5 ( Cross-trek based ) It could be a great addition using the main body frame / jbeam as the sunburst with only small changes added to the rear end, like just a small liftback add-on with a roof extension. This would basically be an all-terrain sunburst and would be quite nice as a smaller alternative to the Hopper, with less capability but a smaller size and still could do some basic offroading.



i wasnt sure what to post this on so i picked this thread

 

Update speculation, not update suggestions. This thread would be a better fit. Vehicle suggestions (See first post) | BeamNG

 

The clutched electric is back in the days of Electric Sunburst, when no official electric cars was present. The electric Sunburst, technically speaking, is very similar to an electric conversion where the transmissions are with torque converters and CVTs. However, all electric motors with clutch are without regenerative braking.
--- Post updated ---

Any electric swap is possible, for example the Citybus.
--- Post updated ---
Also I found a bug on the electric motor(s), to replicate the bug situation, please do the following.
Use the Engine Debug UI app and watch the Flywheel Power.
Drive the eSBR, any spec is ok. Accelerate to a sufficiently high speed such as 120km/h and lift off the throttle.
Compare the power meter on the dashboard(use comfort mode) to the flywheel power, the regen efficiency is very good, note that even the instantaneous power is smoothed and you may momentarily find the recharge power displayed higher than Flywheel Power, this is not a big problem.
The problem is, when you hit reset (R key) and do the same above, the regen efficiency is very very poor.
However if you hit reload (Ctrl+R key) and to the same, the regen efficiency will be higher again.

This is because the energy reservoir is incorrectly reset
\BeamNG.drive\lua\vehicle\powertrain\electricMotor.lua
There are two counters, storageWithEnergyCounter and storageCounter, the motor drains energy from the former and regen the latter, to make sure the regen can charge the battery when the battery is fully discharged. But, in the "local function reset(device, jbeamData)" at line 311, only the "storageWithEnergyCounter" was subject to the reset, the "storageCounter" isn't there. To solve the problem simply add a

Code:

  device.storageCounter = 0

right below the "storageWithEnergyCounter" at line 345

 

That reminds me, what do you think would be the next vehicle to get a Major remaster? I honestly think it would make sense to finish off the H series and Grand Marshall. The Grand Marshall would be a great platform for an late 70's- 80's Body but this would require a new engine family (Retextured Barstow engines would work well.), Plus these would fit well in the earlier D Series as realistically ohc V8's only became a thing in American pickups around the late 90's.

The H Series obviously could do with face lift variants, Passenger vans, a camper van upfit, a custom conversion variant, Bus upfit (as the Wentward can be a bit too big for some routes), and more.

The T series could do a remaster (especially with it's engine) and it's frame can lend well for a COE cab too.


Yeah yeah suggestion rather than speculation but the more we talk about the more we can unintentionally provide possible suggestions.

 

You need to also change the rear suspension type to an independent because a live suspension isn't exactly compatible with an electric motor

 

So you dont think one can replace the ICE and have the rest of the drivetrain unchanged?

 

or you could just strap an electric motor to the solid axle.

 

So, now that we have the last car from The List™ as well as the unexpected Wigeon in-game, what kind of car do you think we will see next? Do you think the Beam team will finally give in to the requests & finally add some sort of a supercar? Will it maybe be a hybrid? Or something as unexpected as the aforementioned Wigeon?

 

I smell vehicle suggestions
Considering the wenover came with the legran remaster, and the wigeon came with the pigeon remaster, I could see it becoming a common thing to make a new car that platform shares with the car that's being remastered;
-With the moonhawk, they may just do more body styles like a sedan/wagon/coupe utility, they may enlarge it, convert it to front-wheel drive and make an earlier wendover, or even a compact like a vega/gremlin/pinto.
-The semi could get the long-suggested cab-over variant as a separate vehicle,
-But what about the bolide? they could go front-engine and make a larger, possibly v12 GT car, or squash the roof down and fit a v12 in the back as a flagship supercar, they might stretch it out and make an uracco/mondial style MR 2+2, though the bolide is already abnormally long as it is. they may throw a complete curveball and go for something like a lamborghini LM002. (or it's possible that given the extent of work planned for the bolide, it's just going to get a heap of content on its own.)