I have analyzed table FND_FLEX_VALUES, it contains thousands of values, like currency, account, etc I just started the course I think it also contains vendor, customer etc so my question is this type of structure look like slow down once table size increase, means we are going to search any value from this table will it take time in future when table size increase?
in a nutshell how you technically analyze flexfield values table and over all structure as for as performance is concern?

When you are looking into fnd_flex_values table, always join it with flex_value_sets table to view meaningful results. In reality following three tables must be joined:

Select flex_Value, c.description
from
fnd_flex_Value_Sets a,
fnd_flex_values b,
fnd_flex_values_tl c,
where a.flex_value_Set_id = b.flex_value_set_id
and b.flex_value_Id = c.flex_value_id
and a.flex_value_Set_name = '<THE VALUE SET NAME YOU WISH TO EXPLORE'>'
and c.lang = '<YOUR LANGUAGE>'

The TL is a language table. Each flex value could have one entry in TL table for each language.

Once the values table is joined with value sets table, the performance is not really an issue.

Also regularly run optimizer. This will store statistics in the system tables and will improve system performance. These statistics are used by the query engine to bring results faster.

If any table will cause performance issue that will be gl_je_lines, gl_balances and even ap_invoice_distributions_all tables etc. These tables contain transactions (not master data like fnd_flex_Values).