how detailed can a jbeam be

I am quite a noob when it comes to how the physics skeleton in BeamNG works , i only know there are beams and nodes and they form a physical skeleton shaped like the car itself.

But when i play BeamNG and only crash my car at a small tree at like 3 mph it just doesnt look right , if understand that right because of the insufficient number of nodes and beams.

How much more detail could be added to Jbeams without my pc melting to a red hot puddle of metal?

 

Depends on the PC and size of the car I guess. I think mine could handle a car 3x as detailed as the moonhawk and stay over the 20fps deathpoint, but you could make a 1x1m box insanely detailed and have ok fps.

 

My PC is pretty equal to yours . Processor is a bit better , but i only have a gtx 660. Someone should just make a car that detailed , want to see what happens. I dont get your box argument tho. shouldnt a box with tons of beams and nodes also slowly kill your Cpu?

 

Well I need to know your specs dang it!

But lets say a car sized, completely flat box at 4x2x0m (I know thats not right) has 1 node on the edge every 1m. It has 8 nodes and poor deformation. Now the 1x1x0 box can have 2 nodes for every meter, have much better deformation, but the same performance.

EDIT: My math is screwed up. Ignore the values and take the idea.

 

i7 4770
16 gb ddr3 ram
gtx 660 2gb
Windows 8.1 (want win 7 back )
wouldnt a box like this just be a rectangle? Ill just assume you mean 4x2x1. now i see what you mean noone needs math btw

 

If each vertex on the car (usually 20,000 40,000) was converted to a node and beam and programmed to perfection, the crash quality would be on the order of magnitude better. But at that point you would need a super powerful computer to run it. and the car file itself would probably be about 400 MB each jbeam file would have thousands of lines of code for each beam and node. Again to run all those jbe file calculations would cripple a 5,000 dollar PC, you would need to spend upwards of 15 grand to build a PC capable of running such a car.


Sent from my iPhone using Tapatalk

 

I don't think so. BNGs physics engine is set up to run one car on one core. I do not think the world has a core fast enough for this.

 

true in its current build it wouldn't be possible, but with OpenGL or cuda it could be run on an extreme PC

 

I was wondering, is there anything stopping someone using the JBeam to simulate stuff on an atomic level? Other than insane processing power I mean.

 

This current physics engines gets unstable past certain values. The underlying physics (beams and nodes) is fairly accurate down to the atomic level. Every state of matter relies on points of mass and bonds between them as an oversimplification.

 

Accuracy of a floating point is not high enough to go that small

 

Lets just wait what intel does , there will be a time where it just has to increase clockspeeds in order to make Cpus faster.
they wont be able to shrink the internals of their cpus forever

 

They'll have to switch to industrial diamond CPUs to reach higher clock speeds, it's been discuses by them at one point.

Proof of concept: http://www.geek.com/chips/81ghz-diamond-semiconductor-created-551147/


Sent from my iPhone using Tapatalk

 

sonunds nice , but how will they cool this? or incase it? wouldnt metall , or the plastic parts of the MB melt?

 

Its thought carbon nanotube transistors could be utilised for a CPU clocking in the terahertz region. There are many alternative substrates to silicon, gallium carbide or germanium for example.
The RAD750 processor used aboard the curiosity rover (and its predecessor is used on many other spacecraft and satellites, POWER Architecture btw as in the old PowerPC macs, nintendo gamecube/wii/wiiU, xbox 360 and PS3) itself does not have a spec of silicon within it, although this is for radiation hardening purposes *not* performance reasons, its manufactured on a 130mm^2 die in a 250nm lithography, much larger than we would conventionally use.

Radiation is one reason we are currently limited to 20nm or so chips right now lithography wise. The smaller you go the more susceptible to radiation the chip becomes and as it stands right now we are approaching the limits before background radiation kicks in depending on where you live although alternate substrates again help this (hence why curiosity uses them).

 

CPUs clocked into the terahertz region would be awesomer than sharks with friggin lasers on their heads, but aren't possible thanks to light being slow as shit. Electricity (or rather, the electromagnetic wave) travels at the speed of light (~3x10[SUP]8 [/SUP]m/s) minus whatever resistance its medium provides. 1THz is 1x10[SUP]12[/SUP] (one trillion) cycles per second, so each signal has a trillionth of a second (a picosecond) to travel about the motherboard and get processed or whatever. Assuming they travel at the speed of light, we can acquire the distance a signal can propagate in a picosecond by dividing the speed of light by one trillion. c / 1x10[SUP]12[/SUP] gives us ~3x10[SUP]-4[/SUP]​m, which is about 0.3mm.
THz clocked CPUs might be possible in the future, but I can't really think of an application unless it, the board, and all other components fit on a pinhead.

Excuse any mistakes in my maths/physics/English/logic; it's nearly 1am at time of posting and I'm really quite done with thinking, haha.

 

The flow of electrical current is not via an electromagnetic wave but is instead the physical flow of electrons

 

Shit, my bad. Pretty sure the rest of it is legit though.

 

Is there anything faster than the speed of light? What about Q-bits ( I think that's what it's called) the concept is that this element is linked no matter the distance to its other element, meaning if the - element is on Pluto and the + element is on earth and the - element on Pluto changed its state from 1 to 0 then the + element on earth would change it's state from 1 to 0 at the same (mirrored) time no speed or travel distance involved, only a link on the atomic level.

Take this with a grain of salt, I may have dreamed it or something but in almost certain it's a theory.


Sent from my iPhone using Tapatalk

 

In theory no:

we have the kinetic energy formula: 0.5 · m · v2 (0.5 x the mass of the object x speed squared).

If we increase the speed to the speed of light we have: 0.5 x m x c2

But what happens if we still give the object more energy? As speed can't be any greater, then we have to increase mass, and that is exactly what happens, if an object reaches the speed of light and you still give it more energy to go faster, you will only make the object more massive (E = mc2)

Quantum entanglement.