[CPU][Feat] Enable KleidiAI accelerated int4 dynamic quant with BF16 activations on Arm CPUs (#33122)

Signed-off-by: Fadi Arafeh <fadi.arafeh@arm.com>

vllm/model_executor/layers/quantization/kernels/mixed_precision/dynamic_4bit.py

@@ -11,6 +11,18 @@ from vllm.scalar_type import scalar_types
 from .MPLinearKernel import MPLinearKernel, MPLinearLayerConfig
 
 
+# This implementation is for the KleidiAI-accelerated w4a8int quantization
+# scheme on Arm CPUs:
+# torch.ops.aten._dyn_quant_matmul_4bit performs dynamic quantized matmul
+#   it takes:
+#       - int4 weights packed along with bias/scales by
+#         torch.ops.aten._dyn_quant_pack_4bit_weight
+#       - float32/bfloat16 activations
+#   then it leverages KleidiAI ukernels that:
+#       - dynamically quantize the activations to int8
+#       - unpack the int4 weights to int8
+#       - perform int8 x int8 -> int32 matmul
+#       - dequantize the int32 output to float32/bfloat16 outputs
 class Dynamic4bitLinearKernel(MPLinearKernel):
     SUPPORTED_QUANT_TYPES = [scalar_types.int4]
 
@@ -29,9 +41,14 @@ class Dynamic4bitLinearKernel(MPLinearKernel):
             and c.act_type
             not in [
                 torch.float32,
+                torch.bfloat16,
             ]
         ):
-            return False, "Dynamic4bitLinearKernel on Arm requires Float32 activations"
+            return (
+                False,
+                "Dynamic4bitLinearKernel on Arm requires Float32 or"
+                " BFloat16 activations",
+            )
         if c.full_weight_shape[0] % c.group_size != 0:
             return (
                 False,