diff --git a/gpu_test/test_kernels.py b/gpu_test/test_kernels.py
index 99ab192..dc0bdf7 100644
--- a/gpu_test/test_kernels.py
+++ b/gpu_test/test_kernels.py
@@ -234,6 +234,43 @@ def test_multi_param(kernel_runner: KernelRunner) -> None:
     assert result == [20, 40, 60, 80]
 
 
+# --- Matmul ---
+
+
+def test_naive_matmul_i64(kernel_runner: KernelRunner) -> None:
+    """Naive i64 matmul: C = A(2x4) * B(4x3) -> C(2x3)."""
+    # Work partition: one thread per output element.
+    # GLOBAL-ID maps to (row, col) with row = gid / N, col = gid MOD N.
+    result = kernel_runner.run(
+        forth_source=(
+            "PARAM A 8\n"
+            "PARAM B 12\n"
+            "PARAM C 6\n"
+            "GLOBAL-ID\n"
+            "DUP 3 /\n"
+            "SWAP 3 MOD\n"
+            "0\n"
+            "4 0 DO\n"
+            "2 PICK\n"
+            "I SWAP 4 * +\n"
+            "CELLS A + @\n"
+            "I 3 * 3 PICK + CELLS B + @\n"
+            "* +\n"
+            "LOOP\n"
+            "2 PICK 3 * 2 PICK +\n"
+            "CELLS C + !"
+        ),
+        params={
+            "A": [1, 2, 3, 4, 5, 6, 7, 8],
+            "B": [1, 0, 2, 0, 1, 2, 1, 0, 1, 2, 1, 0],
+        },
+        block=(6, 1, 1),
+        output_param=2,
+        output_count=6,
+    )
+    assert result == [12, 6, 9, 28, 14, 29]
+
+
 # --- User-Defined Words ---
 
 
diff --git a/test/Pipeline/matmul-naive.forth b/test/Pipeline/matmul-naive.forth
new file mode 100644
index 0000000..fe87d58
--- /dev/null
+++ b/test/Pipeline/matmul-naive.forth
@@ -0,0 +1,27 @@
+\ RUN: %warpforth-translate --forth-to-mlir %s | %warpforth-opt --warpforth-pipeline | %FileCheck %s
+\ RUN: %warpforth-translate --forth-to-mlir %s | %warpforth-opt --convert-forth-to-memref --convert-forth-to-gpu | %FileCheck %s --check-prefix=MID
+
+\ Verify that a naive integer matmul kernel survives the full pipeline.
+\ CHECK: gpu.binary @warpforth_module
+
+\ Verify the kernel signature at the memref+gpu stage.
+\ MID: gpu.func @main(%arg0: memref<8xi64> {forth.param_name = "A"}, %arg1: memref<12xi64> {forth.param_name = "B"}, %arg2: memref<6xi64> {forth.param_name = "C"}) kernel
+
+PARAM A 8
+PARAM B 12
+PARAM C 6
+
+\ M=2, N=3, K=4. One thread computes C[row, col] where gid = row*N + col.
+GLOBAL-ID
+DUP 3 /
+SWAP 3 MOD
+0
+4 0 DO
+  2 PICK
+  I SWAP 4 * +
+  CELLS A + @
+  I 3 * 3 PICK + CELLS B + @
+  * +
+LOOP
+2 PICK 3 * 2 PICK +
+CELLS C + !