Options that apply to kernel implementation on AI Engine tiles.
- Xchess arg
-
Used to pass kernel specific options to the CHESS compiler that is used to compile code for each AI Engine. The option string is specified as
<kernel>:<optionid>=<value>
. The option string is included during compilation of generated source files on the AI Engine where the specified kernel is mapped.Xchess=main:darts.xargs=-nb
- Xelfgen arg
-
Can be used to pass additional command-line options to the ELF generation phase of the compiler, which is currently run as a
make
command to build all AI Engine ELF files.For example, to limit the number of parallel compilations to four, you write
Xelfgen="-j4"
.Note: If you see errors withbad_alloc
in the log, or if the Vitis IDE crashes, this could be due to insufficient memory on your workstation. A possible workaround (other than increasing the available memory on your machine) is to limit the parallelism used by the compiler during code generation by passing this option to the Elf Generator Makefile (-j1 or -j2).Xelfgen=-j2
- Xmapper arg
-
Can be used to pass additional command-line options to the mapper phase of the compiler. These are options to try when the design is either failing to converge in the mapping or routing phase, or when you are trying to achieve better performance via reduction in memory bank conflict. See the Mapper and Router Options for a list and description of options.
Xmapper=DisableFloorplanning
- Xpreproc arg
-
Pass general option to the PREPROCESSOR phase for all source code compilations AIE/PS/PL/x86sim of the AI Engine component (e.g. -D...).
Xpreproc=-D<var>=<value>
- Xpslinker arg
- Pass general option to the PS LINKER phase of the AI Engine component (e.g. -L...
-l...).
Xpslinker=-L<libpath> -l<libname>
- Xrouter arg
-
Pass general option to the ROUTER phase.
Xrouter=dmaFIFOsInFreeBankOnly
- Xx86sim arg
-
Pass x86sim-specific option to the compiler.
Xx86sim=clangStaticAnalyzer
- fast-floats
-
Enable fast implementation for linear floating point scalar operations like
add
,sub
,mul
andcompare
. Accepted values are true and false. The default is false.fast-floats=true
- fast-nonlinearfloats
-
Enable fast implementation for non-linear floating point scalar operations like
sine
/cosine
,sqrt
andinv
. Accepted values are true and false. The default is false.fast-nonlinearfloats=true
- fastmath
-
Enable fast implementations of
float2fix
,fplt
andfpge
. Accepted values are true and false. The default is false.fastmath=true
- float-accuracy
-
Option available only for AI Engine ML. It selects the required floating-point compute accuracy emulated with
bfloat16
numerical type:float-accuracy=<arg
Available argument values are:
-
safe
: Accuracy is slightly better than FP32. -
fast
: Improved performances with similar accuracy to FP32. -
low
: Best performances with better accuracy than FP16 andbfolat16
.
-