Bit manipulation functions
Applies to: SQL Server 2022 (16.x) Azure SQL Database Azure SQL Managed Instance SQL analytics endpoint in Microsoft Fabric Warehouse in Microsoft Fabric
Bit manipulation functions such as moving, retrieving (getting), setting, or counting single bits within an integer or binary value, allow you to process and store data more efficiently than with individual bits.
A bit has two values (1
or 0
, which represent on
or off
, or true
or false
). A byte is made up of a sequence of 8 bits. Bit manipulation functions in SQL Server treat the "leftmost" bit in a byte as the biggest (the most significant). To the bit manipulation functions, bits are numbered from right to left, with bit 0
being the rightmost and the smallest and bit 7
being the leftmost and largest.
For example, a binary sequence of 00000111
is the decimal equivalent of the number 7
. You can calculate this out using powers of 2 as follows:
00000111 = (2^2 + 2^1 + 2^0 = 4 + 2 + 1 = 7)
What this means in practice is that while SQL Server stores this value as 11100000
(byte-reversed), the bit manipulation functions will treat it as though it's 00000111
.
When looking at multiple bytes, the first byte (reading left to right) is the biggest.
You can use the following images to visualize how SQL Server's bit manipulation functions interpret bit and byte expression values and bit offsets.
int
smallint
Functions
There are five functions available for manipulating bits in SQL Server:
All five functions are intended to operate on tinyint, smallint, int, bigint, binary(n), and varbinary(n) data types.
The following types aren't supported: varchar, nvarchar, image, ntext, text, xml, and table.
Remarks
In the initial implementation, Distributed Query functionality for the bit manipulation functions within linked server or ad hoc queries (OPENQUERY) won't be supported.