GPUs to Mars: Full-Scale Simulation of SpaceX's Mars Rocket Engine
Why Hacker News recommends it
HN points to it as evidence of SpaceX's serious simulation capability and as a showcase of GPU compute for CFD.
What it is
SpaceX engineers present at NVIDIA's GPU Technology Conference on a multi-physics code that simulates the full-scale Raptor methane engine. It covers GPU-based turbulent combustion modeling with adaptive grids and wavelet compression.
Recommended by 5 HN commenters across 5 different months, 2015 to 2024.
What Hacker News said
Strong rec paulsutter · 2024-10 · on “Can SpaceX land a rocket with 1/2 cm accuracy?”
The key was the full scale combustion simulator software that SpaceX developed to design the Raptor. This talk was given at NVIDIA GTC 2015 and inspired me to go into manufacturing [this video] title: "GPUs To Mars: Full Scale Simulation of SpaceX's Mars Rocket Engine"
Recommends T-A · 2015-07 · on “Ask HN: What companies are doing work related to scientific computing?”
Here's a fun recent link: [this video] Plenty more options to explore in the links under the video, starting from [LINK] Then there is NVIDIA's CUDA Zone with lists like [LINK]
Recommends cma · 2015-12 · on “The first plasma: the Wendelstein 7-X fusion device is now in operation”
There are bigger optimizations than 10X with some of the adaptive grid / wavelet compression stuff. I think these are still 2D, but pretty amazing: [this video]
Recommends cma · 2015-12 · on “The first plasma: the Wendelstein 7-X fusion device is now in operation”
There are bigger optimizations than 10X with some of the adaptive grid / wavelet compression stuff. I think these are still 2D, but pretty amazing: [this video]
Recommends swimfar · 2019-05 · on “Trump Administration Might “Re-Examine” Climate Modeling”
I remember that. Fluid dynamics is a very tricky thing to model and the work they did was really cool. There's a good video presentation on it here: [this video]
Recommends danielvf · 2022-09 · on “Some things I realized about AI while contemplating slide rule prices on eBay”
Or maybe it is that SpaceX is doing simulations on a level ahead of what everyone else does. Here's one of several amazing videos from SpaceX simulations. (There are more, but I don't have the link to the others) [this video]
Cites DiThi · 2017-10 · on “Hyperloop One Becomes ‘Virgin Hyperloop One’”
How about you also copy the sentence that lies just between those two? "As land slowly settles to a new position over time, the damper neutral position can be adjusted accordingly." And if that doesn't work, the tube can have some kind of rails or just to roll over wheels on top of the pylons. (my own thoughts as non expert) Rocket engineers with advanced simulation tools[0] have been working on this for 5 months before releasing the paper. Do you really think they somehow missed this obvious issue? [0] [this video] Edit: Further in the paper it says: "The tube will be supported by pillars which constrain the tube in the vertical direction but allow longitudinal slip for thermal expansion as well
Cites Const-me · 2018-03 · on “Tesla GPU Accelerator Bang for the Buck, Kepler to Volta”
> that would be complex and severely impact performance on GPU architectures Agree about the complexity. However, adaptive grids can be a huge win in terms of performance. The GPU doesn’t need to simulate the whole 8G grid nodes, it only simulates much fewer nodes, only in the areas of large gradients / higher frequencies of the simulated fields. Here’s an old SpaceX presentation about their GPU simulations of rockets: [this video]