[VECTOR_FLOAT16] Implement serialization and deserialization logic #2853
+391
−0
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
For upcoming VECTOR_FLOAT16 support in the Microsoft JDBC Driver for SQL Server, this change introduces complete serialization and deserialization logic for half-precision floating-point values used inside vector columns.
The driver performs IEEE-754–compliant bidirectional conversion between standard Java Float[] (FLOAT32, 4 bytes) and the compact FLOAT16 (2 bytes) wire format. Applications continue interacting with Float[], while the driver transparently handles precision conversion depending on the negotiated vector type.
Data Flow
Float[](4 bytes each)FLOAT16(2 bytes each)FLOAT16(2 bytes each)Float[](4 bytes each)Conversions Added
Motivation
This enhancement enables efficient storage and transmission of VECTOR_FLOAT16 data between Java applications and SQL Server, providing a foundation for vector search, embeddings, and machine-learning workflows that benefit from reduced memory footprint.