diff --git a/libspu/mpc/ab_api.cc b/libspu/mpc/ab_api.cc
index f298b36f..8dcacff4 100644
--- a/libspu/mpc/ab_api.cc
+++ b/libspu/mpc/ab_api.cc
@@ -118,6 +118,10 @@ Value mul_aa(SPUContext* ctx, const Value& x, const Value& y) {
   TILED_DISPATCH(ctx, x, y);
 }
 
+Value mul_aaa(SPUContext* ctx, const Value& x, const Value& y, const Value& z) {
+  TILED_DISPATCH(ctx, x, y, z);
+}
+
 Value mul_aa_p(SPUContext* ctx, const Value& x, const Value& y) {
   TILED_DISPATCH(ctx, x, y);
 }
diff --git a/libspu/mpc/ab_api.h b/libspu/mpc/ab_api.h
index 8bb954eb..3a39dab2 100644
--- a/libspu/mpc/ab_api.h
+++ b/libspu/mpc/ab_api.h
@@ -41,6 +41,7 @@ OptionalAPI<Value> add_av(SPUContext* ctx, const Value& x, const Value& y);
 
 Value mul_ap(SPUContext* ctx, const Value& x, const Value& y);
 Value mul_aa(SPUContext* ctx, const Value& x, const Value& y);
+Value mul_aaa(SPUContext* ctx, const Value& x, const Value& y, const Value& z);
 Value mul_aa_p(SPUContext* ctx, const Value& x, const Value& y);
 
 Value square_a(SPUContext* ctx, const Value& x);
diff --git a/libspu/mpc/ab_api_test.cc b/libspu/mpc/ab_api_test.cc
index aaee85d3..cb36e44f 100644
--- a/libspu/mpc/ab_api_test.cc
+++ b/libspu/mpc/ab_api_test.cc
@@ -329,6 +329,37 @@ TEST_P(ArithmeticTest, MulAA) {
   });
 }
 
+TEST_P(ArithmeticTest, MulAAA) {
+  const auto factory = std::get<0>(GetParam());
+  const RuntimeConfig& conf = std::get<1>(GetParam());
+  const size_t npc = std::get<2>(GetParam());
+
+  utils::simulate(npc, [&](const std::shared_ptr<yacl::link::Context>& lctx) {
+    auto sctx = factory(conf, lctx);
+
+    auto p0 = rand_p(sctx.get(), kShape);
+    auto p1 = rand_p(sctx.get(), kShape);
+    auto p2 = rand_p(sctx.get(), kShape);
+
+    auto v0 = p2v(sctx.get(), p0, 0);
+    auto v1 = p2v(sctx.get(), p1, 1);
+    auto v2 = p2v(sctx.get(), p2, 2);
+
+    auto a0 = v2a(sctx.get(), v0);
+    auto a1 = v2a(sctx.get(), v1);
+    auto a2 = v2a(sctx.get(), v2);
+
+    auto prod = mul_aaa(sctx.get(), a0, a1, a2);
+    auto p_prod = a2p(sctx.get(), prod);
+
+    auto s = mul_pp(sctx.get(), p0, p1);
+    auto s_prime = mul_pp(sctx.get(), s, p2);
+
+    /* THEN */
+    EXPECT_VALUE_EQ(s_prime, p_prod);
+  });
+}
+
 TEST_P(ArithmeticTest, MulAAP) {
   const auto factory = std::get<0>(GetParam());
   const RuntimeConfig& conf = std::get<1>(GetParam());
diff --git a/libspu/mpc/kernel.cc b/libspu/mpc/kernel.cc
index c43a7d17..1e9e8d96 100644
--- a/libspu/mpc/kernel.cc
+++ b/libspu/mpc/kernel.cc
@@ -66,6 +66,21 @@ void BinaryKernel::evaluate(KernelEvalContext* ctx) const {
   ctx->pushOutput(WrapValue(z));
 }
 
+void TernaryKnernel::evaluate(KernelEvalContext* ctx) const {
+  const auto& x = ctx->getParam<Value>(0);
+  const auto& y = ctx->getParam<Value>(1);
+  const auto& z = ctx->getParam<Value>(2);
+
+  SPU_ENFORCE(x.shape() == y.shape(), "shape mismatch {} {}", x.shape(),
+              y.shape());
+  SPU_ENFORCE(x.shape() == z.shape(), "shape mismatch {} {}", x.shape(),
+              z.shape());
+
+  auto out = proc(ctx, UnwrapValue(x), UnwrapValue(y), UnwrapValue(z));
+
+  ctx->pushOutput(WrapValue(out));
+}
+
 void MatmulKernel::evaluate(KernelEvalContext* ctx) const {
   const auto& lhs = ctx->getParam<Value>(0);
   const auto& rhs = ctx->getParam<Value>(1);
diff --git a/libspu/mpc/kernel.h b/libspu/mpc/kernel.h
index 8db9b546..a9010029 100644
--- a/libspu/mpc/kernel.h
+++ b/libspu/mpc/kernel.h
@@ -52,6 +52,13 @@ class BinaryKernel : public Kernel {
                           const NdArrayRef& rhs) const = 0;
 };
 
+class TernaryKnernel : public Kernel {
+  public:
+  void evaluate(KernelEvalContext* ctx) const override;
+  virtual NdArrayRef proc(KernelEvalContext* ctx, const NdArrayRef& x,
+                          const NdArrayRef& y, const NdArrayRef& z) const = 0;
+};
+
 class MatmulKernel : public Kernel {
  public:
   void evaluate(KernelEvalContext* ctx) const override;
diff --git a/libspu/mpc/shamir/arithmetic.cc b/libspu/mpc/shamir/arithmetic.cc
index fb137584..2a6e3de7 100644
--- a/libspu/mpc/shamir/arithmetic.cc
+++ b/libspu/mpc/shamir/arithmetic.cc
@@ -36,6 +36,10 @@ NdArrayRef wrap_a2p(SPUContext* ctx, const NdArrayRef& x) {
   return UnwrapValue(a2p(ctx, WrapValue(x)));
 }
 
+NdArrayRef wrap_negate_a(SPUContext* ctx, const NdArrayRef& x) {
+  return UnwrapValue(negate_a(ctx, WrapValue(x)));
+}
+
 // Generate zero sharings of degree = threshold
 NdArrayRef gen_zero_shares(KernelEvalContext* ctx, int64_t numel,
                            int64_t threshold) {
diff --git a/libspu/mpc/shamir/arithmetic.h b/libspu/mpc/shamir/arithmetic.h
index 608f810d..25113242 100644
--- a/libspu/mpc/shamir/arithmetic.h
+++ b/libspu/mpc/shamir/arithmetic.h
@@ -147,6 +147,18 @@ class MulAA : public BinaryKernel {
                   const NdArrayRef& rhs) const override;
 };
 
+class MulAAA : public TernaryKnernel {
+ public:
+  static constexpr const char* kBindName() { return "mul_aaa"; }
+
+  ce::CExpr latency() const override { return ce::Const(2); }
+
+  ce::CExpr comm() const override { return ce::K() * 4; }
+
+  NdArrayRef proc(KernelEvalContext* ctx, const NdArrayRef& x,
+                  const NdArrayRef& y, const NdArrayRef& z) const override;
+};
+
 class MulAAP : public BinaryKernel {
  public:
   static constexpr const char* kBindName() { return "mul_aa_p"; }