Not many can claim 25 years on the Internet! Join us in celebrating this milestone. Learn more about our history, and thank you for being a part of our community!
This guy asked why Ray Tracing was so slow. Bacterius explains the problem succinctly and then goes on to explain how to fix the problem. He even gave a time-frame for when the solution he proposed would take to come out. Low and behold, he was spot on.
This probably breaks a ton of rules but this guy deserves the credit.
https://www.gamedev.net/forums/topic/635956-why-is-ray-tracing-slow/
Well, now it's 2019. I've got an NVidia 2080 RTX running with Unreal Engine 4 and I can ray trace 1920x1080 scenes in real time at 60 fps. It's been something graphics programmers have dreamed of for 30+ years. For a long time, I thought it would never happen, but here we are. This afternoon I was ray tracing a scene with two metallic spheres and counting the number of bounces in their reflections and modifying it in real time. Wow! A scene like that would have taken a few minutes to render on the CPU 8 years ago, even with multi-threading.
The tech is only 3-4 months old though, so it's easy to run into areas which aren't fully supported yet. It's also easy to crash your video card on accident. But, I think in the next few quarters, we're going to start seeing a lot of these rough edges getting smoothed out. Seeing RTRT in upcoming games may take a lot longer though: release cycles are long and publishers & devs are going to target a market that doesn't have RTRT capable GPU's quite yet. Give it a few years though, and I think ray traced games will become a lot more common.
You can't really tell if he knew what he was talking about. If it had taken 15 years instead of 7, you would have found another post of some other guy who said it would take 15 years until we see that kind of hardware. That said, his explanation is correct, but that's not that impressive, more like common knowlegde if you are an expert in that field.
We often do that with other things, too. People who are good at predicting stock prices are seen as experts, when in reality they are just guessing like everyone else and just happen to be correct by pure luck. If 1000 people guess 1000 different outcomes, some of those predictions are obviously closer to the actual outcome than the rest.
