While we would have liked to get a few of these Teslas, they wouldn’t get our Asimov any closer to the moon, and after all that’s why we are really doing this. So we are even happier to have gotten some of these Teslas, because they help Asimov get to the moon and help us get a look at what’s on the moon.
Our software team will run many simulations on them. For example, we will simulate the translunar injection route our rover and lander will take to the moon, and probably a couple of million times, each with slightly different parameters for payload weight, total fuel, timing and duration of thruster burns. Each new simulation will have some parameters altered according to trends in the previous simulations. The Tesla is what lets us do millions of simulations in a reasonable amount of time.
No, that’s not the Tesla we got.The same can be done for the landing. With a particular lunar orbit and a target landing site, we can test different thruster burn schedules or even an autonomous control landing algorithm, which leads to safer landings.
Our regular computing hardware without the Teslas would let us do maybe 100,000 simulations before we need to launch, and while that’s not terrible, it would tie up our computing resources for a long time, and we’re always looking for better approaches to our problems that give our mission the greatest chance of success.
Once the rover is safely landed, its cameras will be recording stereoscopic, HD-resolution images and transmit then back to Earth. There we will be processing them to extract terrain features and generate a very highly detailed topographic map of the rover’s surroundings. The NVIDIA Teslas will be doing all that work, possibly in real-time, alerting us to the terrain features around the rover and letting the remote driver avoid obstacles or dead ends.
Again, without the Tesla, our computing power would not get us to real-time or close (there is always the Earth-Moon transmission delay), and we would have to drive with a 2D image and no topographic map. As with the simulations above, it is likely possible to complete the mission with 2D image and no maps, but we were seeking a better solution, and found in NVIDIA a gracious partner that believes we can put their products to good use and supplies us with the computing capacity to take our mission to the next level.
So thanks for the Teslas.
The Part-Time Scientists
If we caught your attention, follow us
, visit our Team Blog or




















