feat: add d4r inference engines and Windows parity tooling

Add native K/M/RR inference engines, HIP kernels, replay and validation
harnesses, and Windows DLSS setup and runtime support. Include GPU
profiling tools, build and packaging scripts, platform integration
patches, and implementation documentation.

tests/windows/test_rr_fp8_conversion.py

@@ -0,0 +1,23 @@
+import unittest
+
+import numpy as np
+
+from engine.rr.rrmath import q8
+
+
+class F16ToE4M3Tests(unittest.TestCase):
+    def test_special_values_and_signed_nan(self):
+        bits = np.array([0x0000, 0x3C00, 0x7C00, 0xFC00, 0x7E00, 0xFE00], dtype=np.uint16)
+        values = bits.view(np.float16)
+        self.assertEqual(q8(values).tolist(), [0x00, 0x38, 0x7E, 0xFE, 0x7F, 0x7F])
+
+    def test_every_half_pattern_maps_to_byte(self):
+        values = np.arange(65536, dtype=np.uint16).view(np.float16)
+        codes = q8(values)
+        self.assertEqual(codes.shape, (65536,))
+        self.assertEqual(codes.dtype, np.uint8)
+        self.assertTrue(np.all(codes[0x7C01:0x8000] == 0x7F))
+
+
+if __name__ == '__main__':
+    unittest.main()