optimize com.microsoft.MatMulNbits operator #42612

Job	Run time
Smart_CI	23s
Docker	43s
Build / Build	5m 54s
Debian Packages	0s
Samples	0s
OpenVINO JS API	0s
OpenVINO tokenizers extension	0s
CPU functional tests	0s
PyTorch Models tests	0s
ONNX Runtime Integration / ONNX Runtime Integration	3m 51s
C++ unit tests / C++ unit tests	51s
JAX Layer Tests / JAX Layer Tests	1m 19s
Python API tests / Python API tests	0s
Python unit tests / Python unit tests	2m 14s
Pytorch Layer Tests / PyTorch Layer Tests	2m 28s
TensorFlow Layer Tests	0s
TensorFlow Models tests	0s
ci/gha_overall_status_linux_arm64	1s
	17m 44s

Provide feedback