Nsa's James Webb Space Telescope delivered the deepest and sharpest infrared image of the distant universe so far on July 12 2022. Webb's First Deep Field image is galaxy cluster SMACS 0723 as it first appeared 4.6 billion years ago, and it is teeming with thousands of galaxies – including the faintest objects ever observed in the infrared.
While the image is already astounding to humans and scientists around the globe on further possibilities that await us from the JWST, the data it gleans will also be feeding a GPU-accelerated AI created at the University of California, Santa Cruz.
Dubbed Morpheus, this deep learning framework classifies astronomical objects, such as galaxies, based on the raw data streaming out of telescopes on a pixel-by-pixel basis.
In fact, Morpheus is already playing a key role in helping scientists understand images taken from the Hubble Space Telescope.
Morpheus is trained on UC Santa Cruz's Lux supercomputer. The machine includes 28 GPU nodes with two Nvidia V100 Tensor Core GPUs each.
Eventually, Morpheus will also be using the images from the $10 billion JWST to better understand what it's looking for, such as looking for habitable planets outside of our solar system. Since the JWST's optics are unique with a vastly different mirror array and better near, mid and far-field infrared cameras, the JWST will be collecting light from galaxies further away than visible on Hubble.
This also means the pictures taken by JWST will be giving up even more secrets in time to come as Morpheus works to unlock what it can process from the vastly higher quality images and new data from the infrared cameras.
Meanwhile, the University of California Santa Cruz made the above image available in a zoomable format, allowing people to hone in on individual galaxies for a closer look.