Are 3D simulation results shared globally in an efficient manner?

David,
I came across an old video on YouTube that refers to PSAAP.  Here is the link:
Towards air-breathing hypersonic vehicles through numerical simulations – PSAAP at Stanford by Ivan Bermejo-Moreno
In this simulation, at two minutes, it is showing cross sections.  I was wondering if, in any of these 3D simulations you do, it is possible to export volumes with layers like that?  I wanted to look at the details of each layer, individually, and then in context of the “real” volume. To look at that 3D volume at that moment in time.
Ivan has gone to USC Viterbi School of Engineering at https://viterbi.usc.edu/directory/faculty/Bermejo-Moreno/Ivan and I doubt he has kept those datasets.  I am copying him anyway.
For the Internet Foundation, I am interested in what PSAAP shares of the data you generate, and what might be useful for training globally, and in breaking apart problems.  I have spoken with a few exascale groups, but nothing really serious, as few supercomputer groups work on soft problems like the Internet.
But for this particular thing I was personally curious about those shapes.  If they were normalized and simplified, then it should be possible to measure and control them in real time now.  So it would help to know the absolute units and properties of the flow.
Is it possible there are cross sections of turbulent flows lying around?  I guess any 3D snapshot would do.  As long the real world units were available.
Here is one I was looking at earlier today.
Visualizing the World’s Largest Turbulence Simulation by Leibniz Supercomputing Centre
It seems there is a lot of this going on, but I don’t see sharing on the general Internet (about 4.8 Billion users), but only a tiny few who have access to computers to run these things, and then they just put the results on the shelf and a few pretty pictures and videos.  I think there might be useful datasets for the roughly 25,000 colleges and universities globally, not to leave out high schools.  I might just be missing what is going on.  I have been studying global communities on the Internet for 23 years, and one thing I have learned, is that if I can think of it, then there are almost always 10,000 people and groups out there who have been working on it hard for many years.  So sharing of pieces of these huge simulations – in pieces small enough to move around on the ordinary Internet should be happening a lot.  I found many climate models, and they are sharing bits of that, but it does not seem to be organized or available to everyone.  I have to look at it as a whole, since I am trying to optimize the whole of the Internet.
But the only sharing I see is mostly pretty pictures in lossy formats.  NASA is still sharing images in lossy formats.  Most sites do. Look at these videos – nice pictures, no background information, data, programs, community.  If anyone is trying to find new people or connections, there are good ways to do that.
Richard Collins, Director, The Internet Foundation
My idea was to find a suitable orthogonal representation for each slice, then stitch them together. Using that representation (a simplification but analytic) solve for the optimal sensor and control systems needed for a real world environment where this operates efficiently.  It is an engine, and the input flow is turbulent and varied,  But with lidar and other electromagnetic sensors there are plenty of clicks to modify the flow with electromagnetic microscopic changes. And in stationary test to change the input flow to match the idealized (chosen for its properties and easy of handling) patterns and flows in the engine.  It sounds long and complicated, but I found many people who do that sort of thing routinely — if they can get all the necessary pieces together and deal with the many computer, human, organizational and mathematical languages involved.  I call it “not hard,just tedious”.
Richard K Collins

About: Richard K Collins

The Internet Foundation Internet policies, global issues, global open lossless data, global open collaboration


Leave a Reply

Your email address will not be published. Required fields are marked *