Remove is_placed flag from JIT decorator

programming_examples/algorithms/README.md

@@ -59,7 +59,7 @@ my_kernel = ExternalFunction(
 
 # Pass in full size tensors, algorithm will tile it to tile_size
 # tile_size must be passed as a keyword argument
-iron.jit(is_placed=False)(transform)(
+iron.jit()(transform)(
     my_kernel, input, output, factor, tile_size=TILE_SIZE
 )
 ```

programming_examples/algorithms/for_each.py

@@ -35,7 +35,7 @@ def main():
     initial_tensor = tensor.numpy().copy()
 
     # JIT compile the algorithm
-    iron.jit(is_placed=False)(for_each)(lambda a: a + 1, tensor, tile_size=16)
+    iron.jit()(for_each)(lambda a: a + 1, tensor, tile_size=16)
 
     # Check the correctness of the result
     e = np.equal(initial_tensor + 1, tensor.numpy())

programming_examples/algorithms/transform.py

@@ -36,7 +36,7 @@ def main():
     output = iron.zeros_like(input)
 
     # JIT compile the algorithm
-    iron.jit(is_placed=False)(transform)(lambda a: a + 1, input, output, tile_size=16)
+    iron.jit()(transform)(lambda a: a + 1, input, output, tile_size=16)
 
     # Check the correctness of the result
     e = np.equal(input.numpy() + 1, output.numpy())

programming_examples/algorithms/transform_binary.py

@@ -36,7 +36,7 @@ def main():
     output = iron.zeros_like(input0)
 
     # JIT compile the algorithm
-    iron.jit(is_placed=False)(transform_binary)(
+    iron.jit()(transform_binary)(
         lambda a, b: a + b, input0, input1, output, tile_size=16
     )
 

programming_examples/algorithms/transform_parallel.py

@@ -36,9 +36,7 @@ def main():
     output = iron.zeros_like(input)
 
     # JIT compile the algorithm
-    iron.jit(is_placed=False)(transform_parallel)(
-        lambda a: a + 1, input, output, tile_size=16
-    )
+    iron.jit()(transform_parallel)(lambda a: a + 1, input, output, tile_size=16)
 
     # Check the correctness of the result
     e = np.equal(input.numpy() + 1, output.numpy())

programming_examples/algorithms/transform_parallel_binary.py

@@ -36,7 +36,7 @@ def main():
     output = iron.zeros_like(input0)
 
     # JIT compile the algorithm
-    iron.jit(is_placed=False)(transform_parallel_binary)(
+    iron.jit()(transform_parallel_binary)(
         lambda a, b: a + b, input0, input1, output, tile_size=16
     )
 

programming_examples/basic/vector_reduce_min/vector_reduce_min_jit.py

@@ -17,7 +17,7 @@
 from aie.iron import ExternalFunction
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def my_reduce_min(input_tensor, output_tensor):
 
     num_elements = input_tensor.numel()

programming_examples/basic/vector_scalar_mul/vector_scalar_mul_jit.py

@@ -61,9 +61,7 @@ def vector_scalar_mul(input, factor, output):
     # Pass scale kernel to the transform algorithm
     # tile_size is passed as keyword argument
     # iron.jit compiles and runs the program
-    iron.jit(is_placed=False)(transform)(
-        scale_kernel, input, output, factor, tile_size=tile_size
-    )
+    iron.jit()(transform)(scale_kernel, input, output, factor, tile_size=tile_size)
 
 
 def main():

programming_examples/basic/vector_vector_add/vector_vector_add.py

@@ -17,7 +17,7 @@
 from aie.iron.controlflow import range_
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def vector_vector_add(input0, input1, output):
     if input0.shape != input1.shape:
         raise ValueError(

programming_examples/getting_started/00_memcpy/memcpy.py

@@ -27,9 +27,8 @@
 # JIT decorator for IRON
 # Decorator to compile an IRON kernel into a binary to run on the NPU.
 # Parameters:
-#     - is_placed (bool): Whether the kernel is using explicit or deferred placement API. Defaults to True.
 #     - use_cache (bool): Use cached MLIR module if available. Defaults to True.
-@iron.jit(is_placed=False)
+@iron.jit()
 def my_memcpy(input0, output):
     # --------------------------------------------------------------------------
     # Configuration

programming_examples/getting_started/01_SAXPY/saxpy.py

@@ -21,9 +21,8 @@
 # JIT decorator for IRON
 # Decorator to compile an IRON kernel into a binary to run on the NPU.
 # Parameters:
-#     - is_placed (bool): Whether the kernel is using explicit or deferred placement API. Defaults to True.
 #     - use_cache (bool): Use cached MLIR module if available. Defaults to True.
-@iron.jit(is_placed=False)
+@iron.jit()
 def saxpy(input0, input1, output):
     N = input0.shape[0]  # Tensor size
     element_type = output.dtype

programming_examples/getting_started/02_vector_reduce_max/vector_reduce_max_1col.py

@@ -22,9 +22,8 @@
 # JIT decorator for IRON
 # Decorator to compile an IRON kernel into a binary to run on the NPU.
 # Parameters:
-#     - is_placed (bool): Whether the kernel is using explicit or deferred placement API. Defaults to True.
 #     - use_cache (bool): Use cached MLIR module if available. Defaults to True.
-@iron.jit(is_placed=False)
+@iron.jit()
 def vector_reduce_max(input0, output):
     element_type = output.dtype
 

programming_examples/getting_started/03_matrix_multiplication_single_core/matrix_multiplication_single_core.py

@@ -20,9 +20,8 @@
 # JIT decorator for IRON
 # Decorator to compile an IRON kernel into a binary to run on the NPU.
 # Parameters:
-#     - is_placed (bool): Whether the kernel is using explicit or deferred placement API. Defaults to True.
 #     - use_cache (bool): Use cached MLIR module if available. Defaults to True.
-@iron.jit(is_placed=False)
+@iron.jit()
 def matrix_multiplication_single_core(input0, input1, output):
     # Problem size
     # - matrix0 shapes: (M, K)

programming_guide/mini_tutorial/aie2p.py

@@ -26,9 +26,8 @@
 # JIT decorator for IRON
 # Decorator to compile an IRON kernel into a binary to run on the NPU.
 # Parameters:
-#     - is_placed (bool): Whether the kernel is using explicit or deferred placement API. Defaults to True.
 #     - use_cache (bool): Use cached MLIR module if available. Defaults to True.
-@iron.jit(is_placed=False)
+@iron.jit()
 def aie2p(input0, output):
     # Dataflow with ObjectFifos
     # ObjectFifos represent a dataflow connection between endpoints in the AIE array.

programming_guide/mini_tutorial/exercise_1/answer_2.py

@@ -16,7 +16,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_1(input0, output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_1/answer_3.py

@@ -15,7 +15,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_1(input0, output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_1/answer_4.py

@@ -16,7 +16,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_1(input0, output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_1/answer_5.py

@@ -16,7 +16,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_1(input0, output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_1/exercise_1.py

@@ -16,7 +16,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_1(output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_2/answer_1.py

@@ -16,7 +16,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_2(input0, output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_2/answer_2.py

@@ -16,7 +16,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_2(input0, output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_2/exercise_2.py

@@ -16,7 +16,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_2(input0, output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_3/answer_2.py

@@ -16,7 +16,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_3(input0, input1, output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_3/exercise_3.py

@@ -16,7 +16,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_3(input0, output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_4/answer_2.py

@@ -23,7 +23,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_4(output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_4/exercise_4.py

@@ -16,7 +16,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_4(output):
     data_size = output.numel()
     element_type = output.dtype

programming_guide/mini_tutorial/exercise_5/exercise_5a/answer_1.py

@@ -17,7 +17,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_5a(input0, output):
     # Define tile size
     tile_height = 3

programming_guide/mini_tutorial/exercise_5/exercise_5a/answer_2.py

@@ -17,7 +17,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_5a(input0, output):
     # Define tile size
     tile_height = 3

programming_guide/mini_tutorial/exercise_5/exercise_5a/answer_3.py

@@ -16,7 +16,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_5a(input0, output):
     data_size = input0.numel()
     element_type = input0.dtype

programming_guide/mini_tutorial/exercise_5/exercise_5a/exercise_5a.py

@@ -17,7 +17,7 @@
 import aie.iron as iron
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_5a(input0, output):
     # Define tile size
     tile_height = 3

programming_guide/mini_tutorial/exercise_5/exercise_5b/exercise_5b.py

@@ -23,7 +23,7 @@
 data_width = 16
 
 
-@iron.jit(is_placed=False)
+@iron.jit()
 def exercise_5b(input0, output):
     # Define tile size
     tile_height = 3