Scatter optimization : Eliminate 64b integer divide. (#662)

Launch 2D grid to eliminate divide and mod in device code,
since 64b integer division is very expensive.

Github Issue #506

Co-authored-by: Vijay Krishnamoorthy <vijay_krish@apple.com>

benchmarks/python/time_utils.py

@@ -1,4 +1,4 @@
-# Copyright © 2023 Apple Inc.
+# Copyright © 2023-2024 Apple Inc.
 
 import time
 
@@ -20,3 +20,15 @@ def time_fn(fn, *args, **kwargs):
 
     msec = 1e3 * (toc - tic) / num_iters
     print(f"{msec:.5f} msec")
+
+
+def measure_runtime(fn, **kwargs):
+    # Warmup
+    for _ in range(5):
+        fn(**kwargs)
+
+    tic = time.time()
+    iters = 10
+    for _ in range(iters):
+        fn(**kwargs)
+    return (time.time() - tic) * 1000 / iters