Setting the FTZ and DAZ Flags
In Intel® processors, the flush-to-zero (
DAZ) flags in the
MXCSRregister are used to control floating-point calculations. Intel® Streaming SIMD Extensions (Intel® SSE) and Intel® Advanced Vector Extensions (Intel® AVX) instructions, including scalar and vector instructions, benefit from enabling the
DAZflags. Floating-point computations using the Intel® SSE and Intel® AVX instructions are accelerated when the
DAZflags are enabled. This improves the application's performance.
[Q]ftzoption to flush
denormalresults to zero when the application is in the gradual underflow mode. This option may improve performance if the
denormalvalues are not critical to the application's behavior. The
[Q]ftzoption, when applied to the main program, sets the
DAZhardware flags. The negative forms of the
-no-ftzfor Linux* and macOS*, and
/Qftz-for Windows*) leave the flags as they are.
The following table describes how the compiler processes
denormalvalues based on the status of the
When set to ON, the compiler...
When set to OFF, the compiler...
denormalresults from floating-point calculations to zero.
...does not change the