A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 10:45 UTC
[Hardware][PowerPC] Make oneDNN dependency optional for Power (#9039)
Signed-off-by: Varad Ahirwadkar <varad.ahirwadkar1@ibm.com>
Varad Ahirwadkar committed
Oct 4, 2024 at 22:54 UTC
e5dc713c2343b3549b43d6e2764a1036e4052bf8
1 file changed
+6
-1
cmake/cpu_extension.cmake
+6
-1
@@ -84,7 +84,12 @@ endif()
84
85
message(STATUS "CPU extension compile flags: ${CXX_COMPILE_FLAGS}")
86
87
-list(APPEND LIBS dnnl numa)
87
+list(APPEND LIBS numa)
88
+
89
+# Appending the dnnl library for the AVX2 and AVX512, as it is not utilized by Power architecture.
90
+if (AVX2_FOUND OR AVX512_FOUND)
91
+ list(APPEND LIBS dnnl)
92
+endif()
93
94
#
95
# _C extension