My CLR aggregate produces a lot of intermediate results. Consequently, serialization/deserialization of aggregate during each row processing leads to great perfomance issues. It is possible to eliminate excessive serialization and improve perfomance?
Product Language
Category
Proposed Solution
Primary Benefit
Other Benefits
Virtualization