Photon is databrick's brand new native vectorized engine developed in C++ for improved query performance (speed and concurrency). It integrates directly with the Databricks Runtime and Spark, meaning no code changes are required to use Photon.
At this point, not all workloads and operators are supported, but you don't have to worry about identifying which workloads can benefit from Photon since its execution is bottom up: meaning it begins at the table scan operator and continues up the DAG (directed acyclic graph) until it hits an operation that is unsupported. At that point, the execution leaves Photon, and the rest of the operations will run without Photon.
More info here and you can also watch the Data and AI Summit release video here