diff --git a/speck-simd.cpp b/speck-simd.cpp
index f22a88bf..03145de0 100644
--- a/speck-simd.cpp
+++ b/speck-simd.cpp
@@ -844,6 +844,239 @@ inline void SPECK64_Dec_6_Blocks(__m128i &block0, __m128i &block1,
 
 // ***************************** Power8 ***************************** //
 
+#if defined(CRYPTOPP_POWER8_AVAILABLE)
+using CryptoPP::uint8x16_p;
+using CryptoPP::uint32x4_p;
+using CryptoPP::uint64x2_p;
+
+using CryptoPP::VectorAdd;
+using CryptoPP::VectorSub;
+using CryptoPP::VectorXor;
+
+// Rotate left by bit count
+template<unsigned int C>
+inline uint32x4_p RotateLeft32(const uint32x4_p val)
+{
+    const uint32x4_p m = {C, C, C, C};
+    return vec_rl(val, m);
+}
+
+// Rotate right by bit count
+template<unsigned int C>
+inline uint32x4_p RotateRight32(const uint32x4_p val)
+{
+    const uint32x4_p m = {32-C, 32-C, 32-C, 32-C};
+    return vec_rl(val, m);
+}
+
+void SPECK64_Enc_Block(uint32x4_p &block0, uint32x4_p &block1,
+        const word32 *subkeys, unsigned int rounds)
+{
+#if defined(CRYPTOPP_BIG_ENDIAN)
+    const uint8x16_p m1 = {7,6,5,4, 15,14,13,12, 23,22,21,20, 31,30,29,28};
+    const uint8x16_p m2 = {3,2,1,0, 11,10,9,8, 19,18,17,16, 27,26,25,24};
+#else
+    const uint8x16_p m1 = {3,2,1,0, 11,10,9,8, 19,18,17,16, 27,26,25,24};
+    const uint8x16_p m2 = {7,6,5,4, 15,14,13,12, 23,22,21,20, 31,30,29,28};
+#endif
+
+    // [A1 A2 A3 A4][B1 B2 B3 B4] ... => [A1 A3 B1 B3][A2 A4 B2 B4] ...
+    uint32x4_p x1 = vec_perm(block0, block1, m1);
+    uint32x4_p y1 = vec_perm(block0, block1, m2);
+
+    for (int i=0; i < static_cast<int>(rounds); ++i)
+    {
+        const uint32x4_p rk = vec_splats(subkeys[i]);
+
+        x1 = RotateRight32<8>(x1);
+        x1 = VectorAdd(x1, y1);
+        x1 = VectorXor(x1, rk);
+
+        y1 = RotateLeft32<3>(y1);
+        y1 = VectorXor(y1, x1);
+    }
+
+#if defined(CRYPTOPP_BIG_ENDIAN)
+    const uint8x16_p m3 = {19,18,17,16, 3,2,1,0, 23,22,21,20, 7,6,5,4};
+    const uint8x16_p m4 = {27,26,25,24, 11,10,9,8, 31,30,29,28, 15,14,13,12};
+#else
+    const uint8x16_p m3 = {3,2,1,0, 19,18,17,16, 7,6,5,4, 23,22,21,20};
+    const uint8x16_p m4 = {11,10,9,8, 27,26,25,24, 15,14,13,12, 31,30,29,28};
+#endif
+
+    // [A1 A3 B1 B3][A2 A4 B2 B4] => [A1 A2 A3 A4][B1 B2 B3 B4]
+    block0 = (uint32x4_p)vec_perm(x1, y1, m3);
+    block1 = (uint32x4_p)vec_perm(x1, y1, m4);
+}
+
+void SPECK64_Dec_Block(uint32x4_p &block0, uint32x4_p &block1,
+        const word32 *subkeys, unsigned int rounds)
+{
+#if defined(CRYPTOPP_BIG_ENDIAN)
+    const uint8x16_p m1 = {7,6,5,4, 15,14,13,12, 23,22,21,20, 31,30,29,28};
+    const uint8x16_p m2 = {3,2,1,0, 11,10,9,8, 19,18,17,16, 27,26,25,24};
+#else
+    const uint8x16_p m1 = {3,2,1,0, 11,10,9,8, 19,18,17,16, 27,26,25,24};
+    const uint8x16_p m2 = {7,6,5,4, 15,14,13,12, 23,22,21,20, 31,30,29,28};
+#endif
+
+    // [A1 A2 A3 A4][B1 B2 B3 B4] ... => [A1 A3 B1 B3][A2 A4 B2 B4] ...
+    uint32x4_p x1 = vec_perm(block0, block1, m1);
+    uint32x4_p y1 = vec_perm(block0, block1, m2);
+
+    for (int i = static_cast<int>(rounds-1); i >= 0; --i)
+    {
+        const uint32x4_p rk = vec_splats(subkeys[i]);
+
+        y1 = VectorXor(y1, x1);
+        y1 = RotateRight32<3>(y1);
+
+        x1 = VectorXor(x1, rk);
+        x1 = VectorSub(x1, y1);
+        x1 = RotateLeft32<8>(x1);
+    }
+
+#if defined(CRYPTOPP_BIG_ENDIAN)
+    const uint8x16_p m3 = {19,18,17,16, 3,2,1,0, 23,22,21,20, 7,6,5,4};
+    const uint8x16_p m4 = {27,26,25,24, 11,10,9,8, 31,30,29,28, 15,14,13,12};
+#else
+    const uint8x16_p m3 = {3,2,1,0, 19,18,17,16, 7,6,5,4, 23,22,21,20};
+    const uint8x16_p m4 = {11,10,9,8, 27,26,25,24, 15,14,13,12, 31,30,29,28};
+#endif
+
+    // [A1 A3 B1 B3][A2 A4 B2 B4] => [A1 A2 A3 A4][B1 B2 B3 B4]
+    block0 = (uint32x4_p)vec_perm(x1, y1, m3);
+    block1 = (uint32x4_p)vec_perm(x1, y1, m4);
+}
+
+void SPECK64_Enc_6_Blocks(uint32x4_p &block0, uint32x4_p &block1,
+            uint32x4_p &block2, uint32x4_p &block3, uint32x4_p &block4,
+            uint32x4_p &block5, const word32 *subkeys, unsigned int rounds)
+{
+#if defined(CRYPTOPP_BIG_ENDIAN)
+    const uint8x16_p m1 = {7,6,5,4, 15,14,13,12, 23,22,21,20, 31,30,29,28};
+    const uint8x16_p m2 = {3,2,1,0, 11,10,9,8, 19,18,17,16, 27,26,25,24};
+#else
+    const uint8x16_p m1 = {3,2,1,0, 11,10,9,8, 19,18,17,16, 27,26,25,24};
+    const uint8x16_p m2 = {7,6,5,4, 15,14,13,12, 23,22,21,20, 31,30,29,28};
+#endif
+
+    // [A1 A2 A3 A4][B1 B2 B3 B4] ... => [A1 A3 B1 B3][A2 A4 B2 B4] ...
+    uint32x4_p x1 = (uint32x4_p)vec_perm(block0, block1, m1);
+    uint32x4_p y1 = (uint32x4_p)vec_perm(block0, block1, m2);
+    uint32x4_p x2 = (uint32x4_p)vec_perm(block2, block3, m1);
+    uint32x4_p y2 = (uint32x4_p)vec_perm(block2, block3, m2);
+    uint32x4_p x3 = (uint32x4_p)vec_perm(block4, block5, m1);
+    uint32x4_p y3 = (uint32x4_p)vec_perm(block4, block5, m2);
+
+    for (int i=0; i < static_cast<int>(rounds); ++i)
+    {
+        const uint32x4_p rk = vec_splats(subkeys[i]);
+
+        x1 = RotateRight32<8>(x1);
+        x2 = RotateRight32<8>(x2);
+        x3 = RotateRight32<8>(x3);
+
+        x1 = VectorAdd(x1, y1);
+        x2 = VectorAdd(x2, y2);
+        x3 = VectorAdd(x3, y3);
+
+        x1 = VectorXor(x1, rk);
+        x2 = VectorXor(x2, rk);
+        x3 = VectorXor(x3, rk);
+
+        y1 = RotateLeft32<3>(y1);
+        y2 = RotateLeft32<3>(y2);
+        y3 = RotateLeft32<3>(y3);
+
+        y1 = VectorXor(y1, x1);
+        y2 = VectorXor(y2, x2);
+        y3 = VectorXor(y3, x3);
+    }
+
+#if defined(CRYPTOPP_BIG_ENDIAN)
+    const uint8x16_p m3 = {19,18,17,16, 3,2,1,0, 23,22,21,20, 7,6,5,4};
+    const uint8x16_p m4 = {27,26,25,24, 11,10,9,8, 31,30,29,28, 15,14,13,12};
+#else
+    const uint8x16_p m3 = {3,2,1,0, 19,18,17,16, 7,6,5,4, 23,22,21,20};
+    const uint8x16_p m4 = {11,10,9,8, 27,26,25,24, 15,14,13,12, 31,30,29,28};
+#endif
+
+    // [A1 A3 B1 B3][A2 A4 B2 B4] => [A1 A2 A3 A4][B1 B2 B3 B4]
+    block0 = (uint32x4_p)vec_perm(x1, y1, m3);
+    block1 = (uint32x4_p)vec_perm(x1, y1, m4);
+    block2 = (uint32x4_p)vec_perm(x2, y2, m3);
+    block3 = (uint32x4_p)vec_perm(x2, y2, m4);
+    block4 = (uint32x4_p)vec_perm(x3, y3, m3);
+    block5 = (uint32x4_p)vec_perm(x3, y3, m4);
+}
+
+void SPECK64_Dec_6_Blocks(uint32x4_p &block0, uint32x4_p &block1,
+            uint32x4_p &block2, uint32x4_p &block3, uint32x4_p &block4,
+            uint32x4_p &block5, const word32 *subkeys, unsigned int rounds)
+{
+#if defined(CRYPTOPP_BIG_ENDIAN)
+    const uint8x16_p m1 = {7,6,5,4, 15,14,13,12, 23,22,21,20, 31,30,29,28};
+    const uint8x16_p m2 = {3,2,1,0, 11,10,9,8, 19,18,17,16, 27,26,25,24};
+#else
+    const uint8x16_p m1 = {3,2,1,0, 11,10,9,8, 19,18,17,16, 27,26,25,24};
+    const uint8x16_p m2 = {7,6,5,4, 15,14,13,12, 23,22,21,20, 31,30,29,28};
+#endif
+
+    // [A1 A2 A3 A4][B1 B2 B3 B4] ... => [A1 A3 B1 B3][A2 A4 B2 B4] ...
+    uint32x4_p x1 = (uint32x4_p)vec_perm(block0, block1, m1);
+    uint32x4_p y1 = (uint32x4_p)vec_perm(block0, block1, m2);
+    uint32x4_p x2 = (uint32x4_p)vec_perm(block2, block3, m1);
+    uint32x4_p y2 = (uint32x4_p)vec_perm(block2, block3, m2);
+    uint32x4_p x3 = (uint32x4_p)vec_perm(block4, block5, m1);
+    uint32x4_p y3 = (uint32x4_p)vec_perm(block4, block5, m2);
+
+    for (int i = static_cast<int>(rounds-1); i >= 0; --i)
+    {
+        const uint32x4_p rk = vec_splats(subkeys[i]);
+
+        y1 = VectorXor(y1, x1);
+        y2 = VectorXor(y2, x2);
+        y3 = VectorXor(y3, x3);
+
+        y1 = RotateRight32<3>(y1);
+        y2 = RotateRight32<3>(y2);
+        y3 = RotateRight32<3>(y3);
+
+        x1 = VectorXor(x1, rk);
+        x2 = VectorXor(x2, rk);
+        x3 = VectorXor(x3, rk);
+
+        x1 = VectorSub(x1, y1);
+        x2 = VectorSub(x2, y2);
+        x3 = VectorSub(x3, y3);
+
+        x1 = RotateLeft32<8>(x1);
+        x2 = RotateLeft32<8>(x2);
+        x3 = RotateLeft32<8>(x3);
+    }
+
+#if defined(CRYPTOPP_BIG_ENDIAN)
+    const uint8x16_p m3 = {19,18,17,16, 3,2,1,0, 23,22,21,20, 7,6,5,4};
+    const uint8x16_p m4 = {27,26,25,24, 11,10,9,8, 31,30,29,28, 15,14,13,12};
+#else
+    const uint8x16_p m3 = {3,2,1,0, 19,18,17,16, 7,6,5,4, 23,22,21,20};
+    const uint8x16_p m4 = {11,10,9,8, 27,26,25,24, 15,14,13,12, 31,30,29,28};
+#endif
+
+    // [A1 A3 B1 B3][A2 A4 B2 B4] => [A1 A2 A3 A4][B1 B2 B3 B4]
+    block0 = (uint32x4_p)vec_perm(x1, y1, m3);
+    block1 = (uint32x4_p)vec_perm(x1, y1, m4);
+    block2 = (uint32x4_p)vec_perm(x2, y2, m3);
+    block3 = (uint32x4_p)vec_perm(x2, y2, m4);
+    block4 = (uint32x4_p)vec_perm(x3, y3, m3);
+    block5 = (uint32x4_p)vec_perm(x3, y3, m4);
+}
+
+#endif  // POWER8
+
+// ***************************** Power8 ***************************** //
+
 #if defined(CRYPTOPP_POWER8_AVAILABLE)
 
 using CryptoPP::uint8x16_p;
@@ -853,7 +1086,6 @@ using CryptoPP::uint64x2_p;
 using CryptoPP::VectorAdd;
 using CryptoPP::VectorSub;
 using CryptoPP::VectorXor;
-using CryptoPP::VectorSwapWords;
 
 // Rotate left by bit count
 template<unsigned int C>
@@ -1143,6 +1375,20 @@ size_t SPECK128_Dec_AdvancedProcessBlocks_SSSE3(const word64* subKeys, size_t ro
 // ***************************** Power8 ***************************** //
 
 #if defined(CRYPTOPP_POWER8_AVAILABLE)
+size_t SPECK64_Enc_AdvancedProcessBlocks_POWER8(const word32* subKeys, size_t rounds,
+    const byte *inBlocks, const byte *xorBlocks, byte *outBlocks, size_t length, word32 flags)
+{
+    return AdvancedProcessBlocks64_6x2_ALTIVEC(SPECK64_Enc_Block, SPECK64_Enc_6_Blocks,
+        subKeys, rounds, inBlocks, xorBlocks, outBlocks, length, flags);
+}
+
+size_t SPECK64_Dec_AdvancedProcessBlocks_POWER8(const word32* subKeys, size_t rounds,
+    const byte *inBlocks, const byte *xorBlocks, byte *outBlocks, size_t length, word32 flags)
+{
+    return AdvancedProcessBlocks64_6x2_ALTIVEC(SPECK64_Dec_Block, SPECK64_Dec_6_Blocks,
+        subKeys, rounds, inBlocks, xorBlocks, outBlocks, length, flags);
+}
+
 size_t SPECK128_Enc_AdvancedProcessBlocks_POWER8(const word64* subKeys, size_t rounds,
     const byte *inBlocks, const byte *xorBlocks, byte *outBlocks, size_t length, word32 flags)
 {
diff --git a/speck.cpp b/speck.cpp
index f0bfb8ac..da5d9afd 100644
--- a/speck.cpp
+++ b/speck.cpp
@@ -201,6 +201,12 @@ extern size_t SPECK128_Dec_AdvancedProcessBlocks_SSSE3(const word64* subKeys, si
 #endif
 
 #if defined(CRYPTOPP_POWER8_AVAILABLE)
+extern size_t SPECK64_Enc_AdvancedProcessBlocks_POWER8(const word32* subKeys, size_t rounds,
+    const byte *inBlocks, const byte *xorBlocks, byte *outBlocks, size_t length, word32 flags);
+
+extern size_t SPECK64_Dec_AdvancedProcessBlocks_POWER8(const word32* subKeys, size_t rounds,
+    const byte *inBlocks, const byte *xorBlocks, byte *outBlocks, size_t length, word32 flags);
+
 extern size_t SPECK128_Enc_AdvancedProcessBlocks_POWER8(const word64* subKeys, size_t rounds,
     const byte *inBlocks, const byte *xorBlocks, byte *outBlocks, size_t length, word32 flags);
 
@@ -217,6 +223,10 @@ std::string SPECK64::Base::AlgorithmProvider() const
 #if (CRYPTOPP_ARM_NEON_AVAILABLE)
     if (HasNEON())
         return "NEON";
+#endif
+#if (CRYPTOPP_POWER8_AVAILABLE)
+    if (HasPower8())
+        return "Power8";
 #endif
     return "C++";
 }
@@ -418,6 +428,11 @@ size_t SPECK64::Enc::AdvancedProcessBlocks(const byte *inBlocks, const byte *xor
     if (HasNEON())
         return SPECK64_Enc_AdvancedProcessBlocks_NEON(m_rkeys, (size_t)m_rounds,
             inBlocks, xorBlocks, outBlocks, length, flags);
+#endif
+#if (CRYPTOPP_POWER8_AVAILABLE)
+    if (HasPower8())
+        return SPECK64_Enc_AdvancedProcessBlocks_POWER8(m_rkeys, (size_t)m_rounds,
+            inBlocks, xorBlocks, outBlocks, length, flags);
 #endif
     return BlockTransformation::AdvancedProcessBlocks(inBlocks, xorBlocks, outBlocks, length, flags);
 }
@@ -434,6 +449,11 @@ size_t SPECK64::Dec::AdvancedProcessBlocks(const byte *inBlocks, const byte *xor
     if (HasNEON())
         return SPECK64_Dec_AdvancedProcessBlocks_NEON(m_rkeys, (size_t)m_rounds,
             inBlocks, xorBlocks, outBlocks, length, flags);
+#endif
+#if (CRYPTOPP_POWER8_AVAILABLE)
+    if (HasPower8())
+        return SPECK64_Dec_AdvancedProcessBlocks_POWER8(m_rkeys, (size_t)m_rounds,
+            inBlocks, xorBlocks, outBlocks, length, flags);
 #endif
     return BlockTransformation::AdvancedProcessBlocks(inBlocks, xorBlocks, outBlocks, length, flags);
 }
diff --git a/speck.h b/speck.h
index fbfc0b43..c0b67315 100644
--- a/speck.h
+++ b/speck.h
@@ -17,11 +17,15 @@
 #include "seckey.h"
 #include "secblock.h"
 
-#if CRYPTOPP_BOOL_X64 || CRYPTOPP_BOOL_X32 || CRYPTOPP_BOOL_X86 || CRYPTOPP_BOOL_ARM32 || CRYPTOPP_BOOL_ARM64
+#if CRYPTOPP_BOOL_X64 || CRYPTOPP_BOOL_X32 || CRYPTOPP_BOOL_X86 || \
+    CRYPTOPP_BOOL_ARM32 || CRYPTOPP_BOOL_ARM64 || \
+    CRYPTOPP_BOOL_PPC32 || CRYPTOPP_BOOL_PPC64
 # define CRYPTOPP_SPECK64_ADVANCED_PROCESS_BLOCKS 1
 #endif
 
-#if CRYPTOPP_BOOL_X64 || CRYPTOPP_BOOL_X32 || CRYPTOPP_BOOL_X86 || CRYPTOPP_BOOL_ARM32 || CRYPTOPP_BOOL_ARM64 || CRYPTOPP_BOOL_PPC32 || CRYPTOPP_BOOL_PPC64
+#if CRYPTOPP_BOOL_X64 || CRYPTOPP_BOOL_X32 || CRYPTOPP_BOOL_X86 || \
+    CRYPTOPP_BOOL_ARM32 || CRYPTOPP_BOOL_ARM64 || \
+    CRYPTOPP_BOOL_PPC32 || CRYPTOPP_BOOL_PPC64
 # define CRYPTOPP_SPECK128_ADVANCED_PROCESS_BLOCKS 1
 #endif