How to read a subset of columns from arrow ipc format file the fastest way?

We are building a high performance training system, and we do care about the performance a lot. We store the training data in arrow ipc format file, say, there are 100M rows and 1000 columns. We just need to read data of 10 columns each time for training.

Seems with the arrow ipc format file, we have to read the whole file first to get the 10 columns. We did to use parquet, because there will be serialize and deserialize and we think arrow ipc format will be faster.

Is there's any suggestion if we can just read 10 columns only to get better performance?

Start by examining how Arrow IPC files handle column access and compare that behavior with the Parquet approach described in the issue. Determine whether selecting 10 columns without reading the full 1000-column file is supported; done should be a clear, evidence-based recommendation or a scoped implementation proposal.

Written by the indexing model from the issue text.