diff --git a/src/frontend/A64/decoder/a64.inc b/src/frontend/A64/decoder/a64.inc
index a9e1b00c..d7460c37 100644
--- a/src/frontend/A64/decoder/a64.inc
+++ b/src/frontend/A64/decoder/a64.inc
@@ -863,7 +863,7 @@ INST(SM3TT2B,                "SM3TT2B",                                   "11001
 //INST(SHA512SU1,              "SHA512SU1",                                 "11001110011mmmmm100010nnnnnddddd")
 INST(RAX1,                   "RAX1",                                      "11001110011mmmmm100011nnnnnddddd")
 //INST(SM3PARTW1,              "SM3PARTW1",                                 "11001110011mmmmm110000nnnnnddddd")
-//INST(SM3PARTW2,              "SM3PARTW2",                                 "11001110011mmmmm110001nnnnnddddd")
+INST(SM3PARTW2,              "SM3PARTW2",                                 "11001110011mmmmm110001nnnnnddddd")
 //INST(SM4EKEY,                "SM4EKEY",                                   "11001110011mmmmm110010nnnnnddddd")
 
 // Data Processing - FP and SIMD - Cryptographic four register
diff --git a/src/frontend/A64/translate/impl/simd_sha512.cpp b/src/frontend/A64/translate/impl/simd_sha512.cpp
index 4ccb86c7..acfbb3d0 100644
--- a/src/frontend/A64/translate/impl/simd_sha512.cpp
+++ b/src/frontend/A64/translate/impl/simd_sha512.cpp
@@ -20,4 +20,27 @@ bool TranslatorVisitor::RAX1(Vec Vm, Vec Vn, Vec Vd) {
     return true;
 }
 
+bool TranslatorVisitor::SM3PARTW2(Vec Vm, Vec Vn, Vec Vd) {
+    const IR::U128 d = ir.GetQ(Vd);
+    const IR::U128 m = ir.GetQ(Vm);
+    const IR::U128 n = ir.GetQ(Vn);
+
+    const IR::U128 temp = ir.VectorEor(n, ir.VectorRotateLeft(32, m, 7));
+    const IR::U128 temp_result = ir.VectorEor(d, temp);
+    const IR::U32 temp2 = [&] {
+        const IR::U32 rotate1 = ir.RotateRight(ir.VectorGetElement(32, temp, 0), ir.Imm8(17));
+        const IR::U32 rotate2 = ir.RotateRight(rotate1, ir.Imm8(17));
+        const IR::U32 rotate3 = ir.RotateRight(rotate1, ir.Imm8(9));
+
+        return ir.Eor(rotate1, ir.Eor(rotate2, rotate3));
+    }();
+
+    const IR::U32 high_temp_result = ir.VectorGetElement(32, temp_result, 3);
+    const IR::U32 replacement = ir.Eor(high_temp_result, temp2);
+    const IR::U128 result = ir.VectorSetElement(32, temp_result, 3, replacement);
+
+    ir.SetQ(Vd, result);
+    return true;
+}
+
 } // namespace Dynarmic::A64