Will it be possible to simulate a complete (at least simple) bacteria atom by atom on an exascale supercomputer? or is it possible already today with the largest systems?
Here, I've read that scientists can simulate an organelle of bacteria. For that, they modeled a system consisting of 100M atoms. How to know which computational capacity was required for that?
Paper: "Atoms to Phenotypes: Molecular Design Principles of Cellular Energy Metabolism"
On Quora, there's an estimation that an E.coli organism would have about 9x1010 atoms (its genome size is about 4 million base pairs).
The smallest known bacterium so far has got about 112K nucleotides in its genome.
In 2014, there has been an atom by atom simulation of the polio virus capsid on the K computer with 10 petaflops. To include enough water molecules to fill and surround the capsid, Okazaki and colleagues needed to model the dynamics of nearly 6.5 million atoms, which they did for a simulated 200 ns.
A 2019 paper claims to simulate a non-biological environment with a total of 20 trillion atoms ... at up to 1.33 PFLOPS, trillion is here 1012. The used tool in this case models fixed molecules, compare to more advanced and complex ab initio simulation.