The important thing is that when you want to save multiple images from multiple sensors, it is better to wait for to write them to disk and then continue with the next server tick otherwise you will lose images or get bad images. I was confused regarding the use of fixed _delta_time and synchronous mode until I realized the order of the game loop, which is as follows: world.tick() sending the tick to the server and blocking!, then -> on_tick() callbacks get called and sensor.listen() callbacks get called but these two's order might change, then returning from world.tick()! During these you can run the client loop on any fps if you separate the world.tick()-ing to another thread which is what I did. Then I realized the solution is in setting the simulation time to fixed time setting.
I opened an issue about this ( carla-simulator#2617) and I found that many others had been struggling with this. I struggled with rendering a multicamera setting with high traffic (200 cars 200 pedestrians) on our good GPU-CPU server as it was lagging and consequently the physics calculations and autopilot calculations also deteriorated which led to a terrible experience. If somebody from the community has a reasonable plan of attack to improve performance without losing flexibility I am happy to hear it. We take this problem seriously and we have a team of amazing engineers working on improving CARLA. Designing a multi-server architecture to use multiple GPUs to render multiple sensors.Keep upgrading the version of UE4 until we end up with a decent support for Vulkan (not as mature as you may thing.).Make traffic independent of the server to alleviate computation load.Now, what are we doing to solve all these issues and make CARLA faster? Long story short, rendering large maps over 1 GPU without proper optimization for that map is going to generate a bottleneck. So even if you got a new NVIDIA RTX XYZ, how good is your CPU? Did you get the latest Xeon?ĭown to the technical bottleneck: Unreal Engine 4. We need to provide a general and flexible architecture to allow the extension of the CARLA server via multiple clients, we need to account for a generic mechanism to render multiple sensors others than cameras, and all of that for any map that a user can come up with. CARLA is a general Autonomous Driving simulator, which means that we need to cover many other use cases besides sensor simulation/rendering. Videogames like GTA-V are optimized to render 1 camera in real-time with very large budgets and game engines created ad-hoc for that purpose. However, here is the thing: sensor simulation is computationally very expensive. The faster CARLA can do sensor simulation the better. I am sure is it a technical reason, in which case haven't you considered trying to get to the root of the problem? I am sure if these crazy games today can achieve 30-60 fps on average machines, our beloved car simulator on a strong gpu can be running Thanks for your issue. Perhaps you could explain here what is the core reason for this. Since this project was created with the goal of helping scientists of Autonomous driving, deep learning, image processing research I think the most essential thing is being able to render images with decent quality and frame rate when having GPUs like that. Most of these issues went unsolved for them. #1580 - Low fps on carla 0.9.5 when cpu, gpu and memory usage is low #1207 - Carla 0.9.2 gets slower after every iteration of scenario_runner #1121 - Increase Framerate of the Server (0.9.2)
#TRAINZ SIMULATOR 12 LOW FPS MANUAL#
#606 - Simulator very slow in manual mode This issue has been brought up before many times: I wonder why is this the case? Why games like GTA V, Pubg or anything run with no lag 60fps as expected on a GPU like mine and even on weaker gpus, and a simple CAR simulator runs on 10-15 fps? I know roughly how games work, but I can't think of a reason why this happens. Cars hit each other, they drive nudging, and I can't do a decent record with even one camera when I will need four at least. I spawn 50 cars 50 pedestrians and my car with only one camera on 700x700 and I hit record and the FPS is 15. However when it comes to rendering scenarios it is terrible. Before I decided on this GTA V was my other option, but this seemed better suiting, and honestly I still think so. I have been working on this project for one and a half months now, and I choose CARLA in the beginning for it being the sole simulator I found with two essential requirements I need: custom camera setting, simulated traffic and maybe an API that let's tinker with the world. I am working on a scene-understanding project with all the stuff (stereo image processing, deep learning, etc). First of all I love the goal of the project and I think there is great potential.
0 kommentar(er)
