#define NNUE_LAYERS_AFFINE_TRANSFORM_H_INCLUDED
#include <iostream>
+#include <algorithm>
+#include <type_traits>
#include "../nnue_common.h"
+#include "../../simd.h"
+
+/*
+ This file contains the definition for a fully connected layer (aka affine transform).
+ Two approaches are employed, depending on the sizes of the transform.
+
+ Approach 1:
+ - used when the PaddedInputDimensions >= 128
+ - uses AVX512 if possible
+ - processes inputs in batches of 2*InputSimdWidth
+ - so in batches of 128 for AVX512
+ - the weight blocks of size InputSimdWidth are transposed such that
+ access is sequential
+ - N columns of the weight matrix are processed a time, where N
+ depends on the architecture (the amount of registers)
+ - accumulate + hadd is used
+
+ Approach 2:
+ - used when the PaddedInputDimensions < 128
+ - does not use AVX512
+ - expected use-case is for when PaddedInputDimensions == 32 and InputDimensions <= 32.
+ - that's why AVX512 is hard to implement
+ - expected use-case is small layers
+ - not optimized as well as the approach 1
+ - inputs are processed in chunks of 4, weights are respectively transposed
+ - accumulation happens directly to int32s
+*/
namespace Stockfish::Eval::NNUE::Layers {
- // Affine transformation layer
+// Fallback implementation for older/other architectures.
+// Identical for both approaches. Requires the input to be padded to at least 16 values.
+#if !defined(USE_SSSE3)
+ template <IndexType InputDimensions, IndexType PaddedInputDimensions, IndexType OutputDimensions>
+ static void affine_transform_non_ssse3(std::int32_t* output, const std::int8_t* weights, const std::int32_t* biases, const std::uint8_t* input)
+ {
+# if defined(USE_SSE2)
+ // At least a multiple of 16, with SSE2.
+ static_assert(PaddedInputDimensions % 16 == 0);
+ constexpr IndexType NumChunks = PaddedInputDimensions / 16;
+ const __m128i Zeros = _mm_setzero_si128();
+ const auto inputVector = reinterpret_cast<const __m128i*>(input);
+
+# elif defined(USE_MMX)
+ static_assert(InputDimensions % 8 == 0);
+ constexpr IndexType NumChunks = InputDimensions / 8;
+ const __m64 Zeros = _mm_setzero_si64();
+ const auto inputVector = reinterpret_cast<const __m64*>(input);
+
+# elif defined(USE_NEON)
+ static_assert(PaddedInputDimensions % 16 == 0);
+ constexpr IndexType NumChunks = PaddedInputDimensions / 16;
+ const auto inputVector = reinterpret_cast<const int8x8_t*>(input);
+# endif
+
+ for (IndexType i = 0; i < OutputDimensions; ++i) {
+ const IndexType offset = i * PaddedInputDimensions;
+
+# if defined(USE_SSE2)
+ __m128i sumLo = _mm_cvtsi32_si128(biases[i]);
+ __m128i sumHi = Zeros;
+ const auto row = reinterpret_cast<const __m128i*>(&weights[offset]);
+ for (IndexType j = 0; j < NumChunks; ++j) {
+ __m128i row_j = _mm_load_si128(&row[j]);
+ __m128i input_j = _mm_load_si128(&inputVector[j]);
+ __m128i extendedRowLo = _mm_srai_epi16(_mm_unpacklo_epi8(row_j, row_j), 8);
+ __m128i extendedRowHi = _mm_srai_epi16(_mm_unpackhi_epi8(row_j, row_j), 8);
+ __m128i extendedInputLo = _mm_unpacklo_epi8(input_j, Zeros);
+ __m128i extendedInputHi = _mm_unpackhi_epi8(input_j, Zeros);
+ __m128i productLo = _mm_madd_epi16(extendedRowLo, extendedInputLo);
+ __m128i productHi = _mm_madd_epi16(extendedRowHi, extendedInputHi);
+ sumLo = _mm_add_epi32(sumLo, productLo);
+ sumHi = _mm_add_epi32(sumHi, productHi);
+ }
+ __m128i sum = _mm_add_epi32(sumLo, sumHi);
+ __m128i sumHigh_64 = _mm_shuffle_epi32(sum, _MM_SHUFFLE(1, 0, 3, 2));
+ sum = _mm_add_epi32(sum, sumHigh_64);
+ __m128i sum_second_32 = _mm_shufflelo_epi16(sum, _MM_SHUFFLE(1, 0, 3, 2));
+ sum = _mm_add_epi32(sum, sum_second_32);
+ output[i] = _mm_cvtsi128_si32(sum);
+
+# elif defined(USE_MMX)
+ __m64 sumLo = _mm_cvtsi32_si64(biases[i]);
+ __m64 sumHi = Zeros;
+ const auto row = reinterpret_cast<const __m64*>(&weights[offset]);
+ for (IndexType j = 0; j < NumChunks; ++j) {
+ __m64 row_j = row[j];
+ __m64 input_j = inputVector[j];
+ __m64 extendedRowLo = _mm_srai_pi16(_mm_unpacklo_pi8(row_j, row_j), 8);
+ __m64 extendedRowHi = _mm_srai_pi16(_mm_unpackhi_pi8(row_j, row_j), 8);
+ __m64 extendedInputLo = _mm_unpacklo_pi8(input_j, Zeros);
+ __m64 extendedInputHi = _mm_unpackhi_pi8(input_j, Zeros);
+ __m64 productLo = _mm_madd_pi16(extendedRowLo, extendedInputLo);
+ __m64 productHi = _mm_madd_pi16(extendedRowHi, extendedInputHi);
+ sumLo = _mm_add_pi32(sumLo, productLo);
+ sumHi = _mm_add_pi32(sumHi, productHi);
+ }
+ __m64 sum = _mm_add_pi32(sumLo, sumHi);
+ sum = _mm_add_pi32(sum, _mm_unpackhi_pi32(sum, sum));
+ output[i] = _mm_cvtsi64_si32(sum);
+
+# elif defined(USE_NEON)
+ int32x4_t sum = {biases[i]};
+ const auto row = reinterpret_cast<const int8x8_t*>(&weights[offset]);
+ for (IndexType j = 0; j < NumChunks; ++j) {
+ int16x8_t product = vmull_s8(inputVector[j * 2], row[j * 2]);
+ product = vmlal_s8(product, inputVector[j * 2 + 1], row[j * 2 + 1]);
+ sum = vpadalq_s16(sum, product);
+ }
+ output[i] = sum[0] + sum[1] + sum[2] + sum[3];
+
+# else
+ std::int32_t sum = biases[i];
+ for (IndexType j = 0; j < InputDimensions; ++j) {
+ sum += weights[offset + j] * input[j];
+ }
+ output[i] = sum;
+# endif
+ }
+
+# if defined(USE_MMX)
+ _mm_empty();
+# endif
+ }
+#endif
+
+ template <typename PreviousLayer, IndexType OutDims, typename Enabled = void>
+ class AffineTransform;
+
+ // A specialization for large inputs.
template <typename PreviousLayer, IndexType OutDims>
- class AffineTransform {
+ class AffineTransform<PreviousLayer, OutDims, std::enable_if_t<(PreviousLayer::OutputDimensions >= 2*64-1)>> {
public:
// Input/output type
using InputType = typename PreviousLayer::OutputType;
static_assert(std::is_same<InputType, std::uint8_t>::value, "");
// Number of input/output dimensions
- static constexpr IndexType InputDimensions =
- PreviousLayer::OutputDimensions;
+ static constexpr IndexType InputDimensions = PreviousLayer::OutputDimensions;
static constexpr IndexType OutputDimensions = OutDims;
+
static constexpr IndexType PaddedInputDimensions =
- ceil_to_multiple<IndexType>(InputDimensions, MaxSimdWidth);
-#if defined (USE_AVX512)
- static constexpr const IndexType OutputSimdWidth = SimdWidth / 2;
-#elif defined (USE_SSSE3)
- static constexpr const IndexType OutputSimdWidth = SimdWidth / 4;
-#endif
+ ceil_to_multiple<IndexType>(InputDimensions, MaxSimdWidth);
+
+ static_assert(PaddedInputDimensions >= 128, "Something went wrong. This specialization should not have been chosen.");
+
#if defined (USE_AVX512)
- static constexpr const IndexType InputSimdWidth = SimdWidth * 2;
+ static constexpr const IndexType InputSimdWidth = 64;
+ static constexpr const IndexType MaxNumOutputRegs = 16;
+#elif defined (USE_AVX2)
+ static constexpr const IndexType InputSimdWidth = 32;
+ static constexpr const IndexType MaxNumOutputRegs = 8;
#elif defined (USE_SSSE3)
- static constexpr const IndexType InputSimdWidth = SimdWidth;
+ static constexpr const IndexType InputSimdWidth = 16;
+ static constexpr const IndexType MaxNumOutputRegs = 8;
+#else
+ // The fallback implementation will not have permuted weights.
+ // We define these to avoid a lot of ifdefs later.
+ static constexpr const IndexType InputSimdWidth = 1;
+ static constexpr const IndexType MaxNumOutputRegs = 1;
#endif
+ // A big block is a region in the weight matrix of the size [PaddedInputDimensions, NumOutputRegs].
+ // A small block is a region of size [InputSimdWidth, 1]
+
+ static constexpr const IndexType NumOutputRegs = std::min(MaxNumOutputRegs, OutputDimensions);
+ static constexpr const IndexType SmallBlockSize = InputSimdWidth;
+ static constexpr const IndexType BigBlockSize = NumOutputRegs * PaddedInputDimensions;
+ static constexpr const IndexType NumSmallBlocksInBigBlock = BigBlockSize / SmallBlockSize;
+ static constexpr const IndexType NumSmallBlocksPerOutput = PaddedInputDimensions / SmallBlockSize;
+ static constexpr const IndexType NumBigBlocks = OutputDimensions / NumOutputRegs;
+
+ static_assert(OutputDimensions % NumOutputRegs == 0);
+
// Size of forward propagation buffer used in this layer
static constexpr std::size_t SelfBufferSize =
- ceil_to_multiple(OutputDimensions * sizeof(OutputType), CacheLineSize);
+ ceil_to_multiple(OutputDimensions * sizeof(OutputType), CacheLineSize);
// Size of the forward propagation buffer used from the input layer to this layer
static constexpr std::size_t BufferSize =
- PreviousLayer::BufferSize + SelfBufferSize;
+ PreviousLayer::BufferSize + SelfBufferSize;
// Hash value embedded in the evaluation file
static constexpr std::uint32_t get_hash_value() {
return hashValue;
}
+ /*
+ Transposes the small blocks within a block.
+ Effectively means that weights can be traversed sequentially during inference.
+ */
+ static IndexType get_weight_index(IndexType i)
+ {
+ const IndexType smallBlock = (i / SmallBlockSize) % NumSmallBlocksInBigBlock;
+ const IndexType smallBlockCol = smallBlock / NumSmallBlocksPerOutput;
+ const IndexType smallBlockRow = smallBlock % NumSmallBlocksPerOutput;
+ const IndexType bigBlock = i / BigBlockSize;
+ const IndexType rest = i % SmallBlockSize;
+
+ const IndexType idx =
+ bigBlock * BigBlockSize
+ + smallBlockRow * SmallBlockSize * NumOutputRegs
+ + smallBlockCol * SmallBlockSize
+ + rest;
+
+ return idx;
+ }
+
// Read network parameters
bool read_parameters(std::istream& stream) {
if (!previousLayer.read_parameters(stream)) return false;
for (std::size_t i = 0; i < OutputDimensions; ++i)
biases[i] = read_little_endian<BiasType>(stream);
+
for (std::size_t i = 0; i < OutputDimensions * PaddedInputDimensions; ++i)
-#if !defined (USE_SSSE3)
- weights[i] = read_little_endian<WeightType>(stream);
-#elif defined (USE_VNNI) || defined (USE_AVX512)
- if constexpr (OutputDimensions <= 8 && OutputDimensions != 1)
- weights[i] = read_little_endian<WeightType>(stream);
- else
- weights[
- (i / 4) % (PaddedInputDimensions / 4) * OutputDimensions * 4 +
- i / PaddedInputDimensions * 4 +
- i % 4
- ] = read_little_endian<WeightType>(stream);
-#else
- weights[
- (i / 4) % (PaddedInputDimensions / 4) * OutputDimensions * 4 +
- i / PaddedInputDimensions * 4 +
- i % 4
- ] = read_little_endian<WeightType>(stream);
-#endif
+ weights[get_weight_index(i)] = read_little_endian<WeightType>(stream);
return !stream.fail();
}
if (!previousLayer.write_parameters(stream)) return false;
for (std::size_t i = 0; i < OutputDimensions; ++i)
write_little_endian<BiasType>(stream, biases[i]);
-#if !defined (USE_SSSE3)
- for (std::size_t i = 0; i < OutputDimensions * PaddedInputDimensions; ++i)
- write_little_endian<WeightType>(stream, weights[i]);
-#else
- std::unique_ptr<WeightType[]> unscrambledWeights = std::make_unique<WeightType[]>(OutputDimensions * PaddedInputDimensions);
- for (std::size_t i = 0; i < OutputDimensions * PaddedInputDimensions; ++i) {
- unscrambledWeights[i] =
- weights[
- (i / 4) % (PaddedInputDimensions / 4) * OutputDimensions * 4 +
- i / PaddedInputDimensions * 4 +
- i % 4
- ];
- }
for (std::size_t i = 0; i < OutputDimensions * PaddedInputDimensions; ++i)
- write_little_endian<WeightType>(stream, unscrambledWeights[i]);
-#endif
+ write_little_endian<WeightType>(stream, weights[get_weight_index(i)]);
return !stream.fail();
}
+
// Forward propagation
const OutputType* propagate(
const TransformedFeatureType* transformedFeatures, char* buffer) const {
const auto input = previousLayer.propagate(
- transformedFeatures, buffer + SelfBufferSize);
+ transformedFeatures, buffer + SelfBufferSize);
+ OutputType* output = reinterpret_cast<OutputType*>(buffer);
#if defined (USE_AVX512)
+ using vec_t = __m512i;
+ #define vec_setzero _mm512_setzero_si512
+ #define vec_set_32 _mm512_set1_epi32
+ #define vec_add_dpbusd_32 Simd::m512_add_dpbusd_epi32
+ #define vec_add_dpbusd_32x2 Simd::m512_add_dpbusd_epi32x2
+ #define vec_hadd Simd::m512_hadd
+ #define vec_haddx4 Simd::m512_haddx4
+#elif defined (USE_AVX2)
+ using vec_t = __m256i;
+ #define vec_setzero _mm256_setzero_si256
+ #define vec_set_32 _mm256_set1_epi32
+ #define vec_add_dpbusd_32 Simd::m256_add_dpbusd_epi32
+ #define vec_add_dpbusd_32x2 Simd::m256_add_dpbusd_epi32x2
+ #define vec_hadd Simd::m256_hadd
+ #define vec_haddx4 Simd::m256_haddx4
+#elif defined (USE_SSSE3)
+ using vec_t = __m128i;
+ #define vec_setzero _mm_setzero_si128
+ #define vec_set_32 _mm_set1_epi32
+ #define vec_add_dpbusd_32 Simd::m128_add_dpbusd_epi32
+ #define vec_add_dpbusd_32x2 Simd::m128_add_dpbusd_epi32x2
+ #define vec_hadd Simd::m128_hadd
+ #define vec_haddx4 Simd::m128_haddx4
+#endif
- [[maybe_unused]] const __m512i Ones512 = _mm512_set1_epi16(1);
-
- [[maybe_unused]] auto m512_hadd = [](__m512i sum, int bias) -> int {
- return _mm512_reduce_add_epi32(sum) + bias;
- };
-
- [[maybe_unused]] auto m512_hadd128x16_interleave = [](
- __m512i sum0, __m512i sum1, __m512i sum2, __m512i sum3) -> __m512i {
+#if defined (USE_SSSE3)
+ const vec_t* invec = reinterpret_cast<const vec_t*>(input);
- __m512i sum01a = _mm512_unpacklo_epi32(sum0, sum1);
- __m512i sum01b = _mm512_unpackhi_epi32(sum0, sum1);
- __m512i sum23a = _mm512_unpacklo_epi32(sum2, sum3);
- __m512i sum23b = _mm512_unpackhi_epi32(sum2, sum3);
+ // Perform accumulation to registers for each big block
+ for (IndexType bigBlock = 0; bigBlock < NumBigBlocks; ++bigBlock)
+ {
+ vec_t acc[NumOutputRegs] = { vec_setzero() };
+
+ // Each big block has NumOutputRegs small blocks in each "row", one per register.
+ // We process two small blocks at a time to save on one addition without VNNI.
+ for (IndexType smallBlock = 0; smallBlock < NumSmallBlocksPerOutput; smallBlock += 2)
+ {
+ const vec_t* weightvec =
+ reinterpret_cast<const vec_t*>(
+ weights
+ + bigBlock * BigBlockSize
+ + smallBlock * SmallBlockSize * NumOutputRegs);
+
+ const vec_t in0 = invec[smallBlock + 0];
+ const vec_t in1 = invec[smallBlock + 1];
+
+ for (IndexType k = 0; k < NumOutputRegs; ++k)
+ vec_add_dpbusd_32x2(acc[k], in0, weightvec[k], in1, weightvec[k + NumOutputRegs]);
+ }
- __m512i sum01 = _mm512_add_epi32(sum01a, sum01b);
- __m512i sum23 = _mm512_add_epi32(sum23a, sum23b);
+ // Horizontally add all accumulators.
+ if constexpr (NumOutputRegs % 4 == 0)
+ {
+ __m128i* outputvec = reinterpret_cast<__m128i*>(output);
+ const __m128i* biasvec = reinterpret_cast<const __m128i*>(biases);
- __m512i sum0123a = _mm512_unpacklo_epi64(sum01, sum23);
- __m512i sum0123b = _mm512_unpackhi_epi64(sum01, sum23);
+ for (IndexType k = 0; k < NumOutputRegs; k += 4)
+ {
+ const IndexType idx = (bigBlock * NumOutputRegs + k) / 4;
+ outputvec[idx] = vec_haddx4(acc[k+0], acc[k+1], acc[k+2], acc[k+3], biasvec[idx]);
+ }
+ }
+ else
+ {
+ for (IndexType k = 0; k < NumOutputRegs; ++k)
+ {
+ const IndexType idx = (bigBlock * NumOutputRegs + k);
+ output[idx] = vec_hadd(acc[k], biases[idx]);
+ }
+ }
+ }
- return _mm512_add_epi32(sum0123a, sum0123b);
- };
+# undef vec_setzero
+# undef vec_set_32
+# undef vec_add_dpbusd_32
+# undef vec_add_dpbusd_32x2
+# undef vec_hadd
+# undef vec_haddx4
+#else
+ // Use old implementation for the other architectures.
+ affine_transform_non_ssse3<
+ InputDimensions,
+ PaddedInputDimensions,
+ OutputDimensions>(output, weights, biases, input);
- [[maybe_unused]] auto m512_haddx4 = [m512_hadd128x16_interleave](
- __m512i sum0, __m512i sum1, __m512i sum2, __m512i sum3, __m128i bias) -> __m128i {
+#endif
- __m512i sum = m512_hadd128x16_interleave(sum0, sum1, sum2, sum3);
+ return output;
+ }
- __m256i sum256lo = _mm512_castsi512_si256(sum);
- __m256i sum256hi = _mm512_extracti64x4_epi64(sum, 1);
+ private:
+ using BiasType = OutputType;
+ using WeightType = std::int8_t;
- sum256lo = _mm256_add_epi32(sum256lo, sum256hi);
+ PreviousLayer previousLayer;
- __m128i sum128lo = _mm256_castsi256_si128(sum256lo);
- __m128i sum128hi = _mm256_extracti128_si256(sum256lo, 1);
+ alignas(CacheLineSize) BiasType biases[OutputDimensions];
+ alignas(CacheLineSize) WeightType weights[OutputDimensions * PaddedInputDimensions];
+ };
- return _mm_add_epi32(_mm_add_epi32(sum128lo, sum128hi), bias);
- };
+ template <typename PreviousLayer, IndexType OutDims>
+ class AffineTransform<PreviousLayer, OutDims, std::enable_if_t<(PreviousLayer::OutputDimensions < 2*64-1)>> {
+ public:
+ // Input/output type
+ using InputType = typename PreviousLayer::OutputType;
+ using OutputType = std::int32_t;
+ static_assert(std::is_same<InputType, std::uint8_t>::value, "");
- [[maybe_unused]] auto m512_add_dpbusd_epi32 = [=](__m512i& acc, __m512i a, __m512i b) {
-#if defined (USE_VNNI)
- acc = _mm512_dpbusd_epi32(acc, a, b);
-#else
- __m512i product0 = _mm512_maddubs_epi16(a, b);
- product0 = _mm512_madd_epi16(product0, Ones512);
- acc = _mm512_add_epi32(acc, product0);
-#endif
- };
+ // Number of input/output dimensions
+ static constexpr IndexType InputDimensions =
+ PreviousLayer::OutputDimensions;
+ static constexpr IndexType OutputDimensions = OutDims;
+ static constexpr IndexType PaddedInputDimensions =
+ ceil_to_multiple<IndexType>(InputDimensions, MaxSimdWidth);
- [[maybe_unused]] auto m512_add_dpbusd_epi32x2 = [=](__m512i& acc, __m512i a0, __m512i b0, __m512i a1, __m512i b1) {
-#if defined (USE_VNNI)
- acc = _mm512_dpbusd_epi32(acc, a0, b0);
- acc = _mm512_dpbusd_epi32(acc, a1, b1);
-#else
- __m512i product0 = _mm512_maddubs_epi16(a0, b0);
- __m512i product1 = _mm512_maddubs_epi16(a1, b1);
- product0 = _mm512_adds_epi16(product0, product1);
- product0 = _mm512_madd_epi16(product0, Ones512);
- acc = _mm512_add_epi32(acc, product0);
-#endif
- };
-
- [[maybe_unused]] auto m512_add_dpbusd_epi32x4 = [=](__m512i& acc, __m512i a0, __m512i b0, __m512i a1, __m512i b1,
- __m512i a2, __m512i b2, __m512i a3, __m512i b3) {
-#if defined (USE_VNNI)
- acc = _mm512_dpbusd_epi32(acc, a0, b0);
- acc = _mm512_dpbusd_epi32(acc, a1, b1);
- acc = _mm512_dpbusd_epi32(acc, a2, b2);
- acc = _mm512_dpbusd_epi32(acc, a3, b3);
-#else
- __m512i product0 = _mm512_maddubs_epi16(a0, b0);
- __m512i product1 = _mm512_maddubs_epi16(a1, b1);
- __m512i product2 = _mm512_maddubs_epi16(a2, b2);
- __m512i product3 = _mm512_maddubs_epi16(a3, b3);
- product0 = _mm512_adds_epi16(product0, product1);
- product0 = _mm512_madd_epi16(product0, Ones512);
- product2 = _mm512_adds_epi16(product2, product3);
- product2 = _mm512_madd_epi16(product2, Ones512);
- acc = _mm512_add_epi32(acc, _mm512_add_epi32(product0, product2));
-#endif
- };
+ static_assert(PaddedInputDimensions < 128, "Something went wrong. This specialization should not have been chosen.");
+#if defined (USE_SSSE3)
+ static constexpr const IndexType OutputSimdWidth = SimdWidth / 4;
+ static constexpr const IndexType InputSimdWidth = SimdWidth;
#endif
-#if defined (USE_AVX2)
- [[maybe_unused]] const __m256i Ones256 = _mm256_set1_epi16(1);
-
- [[maybe_unused]] auto m256_hadd = [](__m256i sum, int bias) -> int {
- __m128i sum128 = _mm_add_epi32(_mm256_castsi256_si128(sum), _mm256_extracti128_si256(sum, 1));
- sum128 = _mm_add_epi32(sum128, _mm_shuffle_epi32(sum128, _MM_PERM_BADC));
- sum128 = _mm_add_epi32(sum128, _mm_shuffle_epi32(sum128, _MM_PERM_CDAB));
- return _mm_cvtsi128_si32(sum128) + bias;
- };
-
- [[maybe_unused]] auto m256_haddx4 = [](__m256i sum0, __m256i sum1, __m256i sum2, __m256i sum3, __m128i bias) -> __m128i {
- sum0 = _mm256_hadd_epi32(sum0, sum1);
- sum2 = _mm256_hadd_epi32(sum2, sum3);
+ // Size of forward propagation buffer used in this layer
+ static constexpr std::size_t SelfBufferSize =
+ ceil_to_multiple(OutputDimensions * sizeof(OutputType), CacheLineSize);
- sum0 = _mm256_hadd_epi32(sum0, sum2);
+ // Size of the forward propagation buffer used from the input layer to this layer
+ static constexpr std::size_t BufferSize =
+ PreviousLayer::BufferSize + SelfBufferSize;
- __m128i sum128lo = _mm256_castsi256_si128(sum0);
- __m128i sum128hi = _mm256_extracti128_si256(sum0, 1);
+ // Hash value embedded in the evaluation file
+ static constexpr std::uint32_t get_hash_value() {
+ std::uint32_t hashValue = 0xCC03DAE4u;
+ hashValue += OutputDimensions;
+ hashValue ^= PreviousLayer::get_hash_value() >> 1;
+ hashValue ^= PreviousLayer::get_hash_value() << 31;
+ return hashValue;
+ }
- return _mm_add_epi32(_mm_add_epi32(sum128lo, sum128hi), bias);
- };
+ static IndexType get_weight_index_scrambled(IndexType i)
+ {
+ return
+ (i / 4) % (PaddedInputDimensions / 4) * OutputDimensions * 4 +
+ i / PaddedInputDimensions * 4 +
+ i % 4;
+ }
- [[maybe_unused]] auto m256_add_dpbusd_epi32 = [=](__m256i& acc, __m256i a, __m256i b) {
-#if defined (USE_VNNI)
- acc = _mm256_dpbusd_epi32(acc, a, b);
+ static IndexType get_weight_index(IndexType i)
+ {
+#if defined (USE_SSSE3)
+ return get_weight_index_scrambled(i);
#else
- __m256i product0 = _mm256_maddubs_epi16(a, b);
- product0 = _mm256_madd_epi16(product0, Ones256);
- acc = _mm256_add_epi32(acc, product0);
+ return i;
#endif
- };
+ }
- [[maybe_unused]] auto m256_add_dpbusd_epi32x2 = [=](__m256i& acc, __m256i a0, __m256i b0, __m256i a1, __m256i b1) {
-#if defined (USE_VNNI)
- acc = _mm256_dpbusd_epi32(acc, a0, b0);
- acc = _mm256_dpbusd_epi32(acc, a1, b1);
-#else
- __m256i product0 = _mm256_maddubs_epi16(a0, b0);
- __m256i product1 = _mm256_maddubs_epi16(a1, b1);
- product0 = _mm256_adds_epi16(product0, product1);
- product0 = _mm256_madd_epi16(product0, Ones256);
- acc = _mm256_add_epi32(acc, product0);
-#endif
- };
-
- [[maybe_unused]] auto m256_add_dpbusd_epi32x4 = [=](__m256i& acc, __m256i a0, __m256i b0, __m256i a1, __m256i b1,
- __m256i a2, __m256i b2, __m256i a3, __m256i b3) {
-#if defined (USE_VNNI)
- acc = _mm256_dpbusd_epi32(acc, a0, b0);
- acc = _mm256_dpbusd_epi32(acc, a1, b1);
- acc = _mm256_dpbusd_epi32(acc, a2, b2);
- acc = _mm256_dpbusd_epi32(acc, a3, b3);
-#else
- __m256i product0 = _mm256_maddubs_epi16(a0, b0);
- __m256i product1 = _mm256_maddubs_epi16(a1, b1);
- __m256i product2 = _mm256_maddubs_epi16(a2, b2);
- __m256i product3 = _mm256_maddubs_epi16(a3, b3);
- product0 = _mm256_adds_epi16(product0, product1);
- product0 = _mm256_madd_epi16(product0, Ones256);
- product2 = _mm256_adds_epi16(product2, product3);
- product2 = _mm256_madd_epi16(product2, Ones256);
- acc = _mm256_add_epi32(acc, _mm256_add_epi32(product0, product2));
-#endif
- };
+ // Read network parameters
+ bool read_parameters(std::istream& stream) {
+ if (!previousLayer.read_parameters(stream)) return false;
+ for (std::size_t i = 0; i < OutputDimensions; ++i)
+ biases[i] = read_little_endian<BiasType>(stream);
+ for (std::size_t i = 0; i < OutputDimensions * PaddedInputDimensions; ++i)
+ weights[get_weight_index(i)] = read_little_endian<WeightType>(stream);
-#endif
-#if defined (USE_SSSE3)
+ return !stream.fail();
+ }
- [[maybe_unused]] const __m128i Ones128 = _mm_set1_epi16(1);
-
- [[maybe_unused]] auto m128_hadd = [](__m128i sum, int bias) -> int {
- sum = _mm_add_epi32(sum, _mm_shuffle_epi32(sum, 0x4E)); //_MM_PERM_BADC
- sum = _mm_add_epi32(sum, _mm_shuffle_epi32(sum, 0xB1)); //_MM_PERM_CDAB
- return _mm_cvtsi128_si32(sum) + bias;
- };
-
- [[maybe_unused]] auto m128_haddx4 = [](__m128i sum0, __m128i sum1, __m128i sum2, __m128i sum3, __m128i bias) -> __m128i {
- sum0 = _mm_hadd_epi32(sum0, sum1);
- sum2 = _mm_hadd_epi32(sum2, sum3);
- sum0 = _mm_hadd_epi32(sum0, sum2);
- return _mm_add_epi32(sum0, bias);
- };
-
- [[maybe_unused]] auto m128_add_dpbusd_epi32 = [=](__m128i& acc, __m128i a, __m128i b) {
- __m128i product0 = _mm_maddubs_epi16(a, b);
- product0 = _mm_madd_epi16(product0, Ones128);
- acc = _mm_add_epi32(acc, product0);
- };
-
- [[maybe_unused]] auto m128_add_dpbusd_epi32x2 = [=](__m128i& acc, __m128i a0, __m128i b0, __m128i a1, __m128i b1) {
- __m128i product0 = _mm_maddubs_epi16(a0, b0);
- __m128i product1 = _mm_maddubs_epi16(a1, b1);
- product0 = _mm_adds_epi16(product0, product1);
- product0 = _mm_madd_epi16(product0, Ones128);
- acc = _mm_add_epi32(acc, product0);
- };
-
- [[maybe_unused]] auto m128_add_dpbusd_epi32x4 = [=](__m128i& acc, __m128i a0, __m128i b0, __m128i a1, __m128i b1,
- __m128i a2, __m128i b2, __m128i a3, __m128i b3) {
- __m128i product0 = _mm_maddubs_epi16(a0, b0);
- __m128i product1 = _mm_maddubs_epi16(a1, b1);
- __m128i product2 = _mm_maddubs_epi16(a2, b2);
- __m128i product3 = _mm_maddubs_epi16(a3, b3);
- product0 = _mm_adds_epi16(product0, product1);
- product0 = _mm_madd_epi16(product0, Ones128);
- product2 = _mm_adds_epi16(product2, product3);
- product2 = _mm_madd_epi16(product2, Ones128);
- acc = _mm_add_epi32(acc, _mm_add_epi32(product0, product2));
- };
+ // Write network parameters
+ bool write_parameters(std::ostream& stream) const {
+ if (!previousLayer.write_parameters(stream)) return false;
+ for (std::size_t i = 0; i < OutputDimensions; ++i)
+ write_little_endian<BiasType>(stream, biases[i]);
-#endif
+ for (std::size_t i = 0; i < OutputDimensions * PaddedInputDimensions; ++i)
+ write_little_endian<WeightType>(stream, weights[get_weight_index(i)]);
-#if defined (USE_AVX512)
- using vec_t = __m512i;
- #define vec_setzero _mm512_setzero_si512
- #define vec_set_32 _mm512_set1_epi32
- [[maybe_unused]] auto& vec_add_dpbusd_32 = m512_add_dpbusd_epi32;
- [[maybe_unused]] auto& vec_add_dpbusd_32x2 = m512_add_dpbusd_epi32x2;
- [[maybe_unused]] auto& vec_add_dpbusd_32x4 = m512_add_dpbusd_epi32x4;
- [[maybe_unused]] auto& vec_hadd = m512_hadd;
- [[maybe_unused]] auto& vec_haddx4 = m512_haddx4;
-#elif defined (USE_AVX2)
+ return !stream.fail();
+ }
+ // Forward propagation
+ const OutputType* propagate(
+ const TransformedFeatureType* transformedFeatures, char* buffer) const {
+ const auto input = previousLayer.propagate(
+ transformedFeatures, buffer + SelfBufferSize);
+ const auto output = reinterpret_cast<OutputType*>(buffer);
+
+#if defined (USE_AVX2)
using vec_t = __m256i;
#define vec_setzero _mm256_setzero_si256
#define vec_set_32 _mm256_set1_epi32
- [[maybe_unused]] auto& vec_add_dpbusd_32 = m256_add_dpbusd_epi32;
- [[maybe_unused]] auto& vec_add_dpbusd_32x2 = m256_add_dpbusd_epi32x2;
- [[maybe_unused]] auto& vec_add_dpbusd_32x4 = m256_add_dpbusd_epi32x4;
- [[maybe_unused]] auto& vec_hadd = m256_hadd;
- [[maybe_unused]] auto& vec_haddx4 = m256_haddx4;
+ #define vec_add_dpbusd_32 Simd::m256_add_dpbusd_epi32
+ #define vec_add_dpbusd_32x2 Simd::m256_add_dpbusd_epi32x2
+ #define vec_add_dpbusd_32x4 Simd::m256_add_dpbusd_epi32x4
+ #define vec_hadd Simd::m256_hadd
+ #define vec_haddx4 Simd::m256_haddx4
#elif defined (USE_SSSE3)
using vec_t = __m128i;
#define vec_setzero _mm_setzero_si128
#define vec_set_32 _mm_set1_epi32
- [[maybe_unused]] auto& vec_add_dpbusd_32 = m128_add_dpbusd_epi32;
- [[maybe_unused]] auto& vec_add_dpbusd_32x2 = m128_add_dpbusd_epi32x2;
- [[maybe_unused]] auto& vec_add_dpbusd_32x4 = m128_add_dpbusd_epi32x4;
- [[maybe_unused]] auto& vec_hadd = m128_hadd;
- [[maybe_unused]] auto& vec_haddx4 = m128_haddx4;
+ #define vec_add_dpbusd_32 Simd::m128_add_dpbusd_epi32
+ #define vec_add_dpbusd_32x2 Simd::m128_add_dpbusd_epi32x2
+ #define vec_add_dpbusd_32x4 Simd::m128_add_dpbusd_epi32x4
+ #define vec_hadd Simd::m128_hadd
+ #define vec_haddx4 Simd::m128_haddx4
#endif
#if defined (USE_SSSE3)
- const auto output = reinterpret_cast<OutputType*>(buffer);
const auto inputVector = reinterpret_cast<const vec_t*>(input);
-#endif
-
-#if defined (USE_VNNI) || defined (USE_AVX512)
-
- static_assert(OutputDimensions == 1 || OutputDimensions % 4 == 0);
-
- // OutputDimensions is either 1 or a multiple of SimdWidth
- // because then it is also an input dimension.
- if constexpr (OutputDimensions <= 8 && OutputDimensions != 1)
- {
- constexpr IndexType NumChunks = PaddedInputDimensions / InputSimdWidth;
-
- static_assert(NumChunks % 2 == 0);
-
- const auto input_vec = reinterpret_cast<const vec_t*>(input);
- const auto bias_vec = reinterpret_cast<const __m128i*>(biases);
- auto out_vec = reinterpret_cast<__m128i*>(output);
-
- vec_t regs[OutputDimensions];
- for (IndexType k = 0; k < OutputDimensions; ++k)
- regs[k] = vec_setzero();
-
- for (IndexType i = 0; i < NumChunks / 2; ++i)
- {
- const vec_t in0 = input_vec[i * 2 + 0];
- const vec_t in1 = input_vec[i * 2 + 1];
- for (IndexType k = 0; k < OutputDimensions; ++k)
- {
- const vec_t w0 = reinterpret_cast<const vec_t*>(&weights[k * PaddedInputDimensions])[i * 2 + 0];
- const vec_t w1 = reinterpret_cast<const vec_t*>(&weights[k * PaddedInputDimensions])[i * 2 + 1];
- vec_add_dpbusd_32(regs[k], in0, w0);
- vec_add_dpbusd_32(regs[k], in1, w1);
- }
- }
-
- for (IndexType k = 0; k < OutputDimensions / 4; ++k)
- {
- out_vec[k] = vec_haddx4(
- regs[k * 4 + 0],
- regs[k * 4 + 1],
- regs[k * 4 + 2],
- regs[k * 4 + 3],
- bias_vec[k]
- );
- }
- }
- else if constexpr (InputDimensions == 8)
- {
- const auto input32 = reinterpret_cast<const std::int32_t*>(input);
- __m256i* outptr = reinterpret_cast<__m256i*>(output);
- std::memcpy(output, biases, OutputDimensions * sizeof(OutputType));
-
- const __m256i in0 = _mm256_set1_epi32(input32[0]);
- const __m256i in1 = _mm256_set1_epi32(input32[1]);
- const auto col0 = reinterpret_cast<const __m256i*>(&weights[0]);
- const auto col1 = reinterpret_cast<const __m256i*>(&weights[OutputDimensions * 4]);
- for (IndexType j = 0; j * 8 < OutputDimensions; ++j)
- m256_add_dpbusd_epi32x2(outptr[j], in0, col0[j], in1, col1[j]);
- }
- else
-#elif defined (USE_SSSE3)
-
- if constexpr (OutputDimensions % OutputSimdWidth == 0 && InputDimensions == 8)
- {
- const auto input32 = reinterpret_cast<const std::int32_t*>(input);
- vec_t* outptr = reinterpret_cast<vec_t*>(output);
- std::memcpy(output, biases, OutputDimensions * sizeof(OutputType));
-
- const vec_t in0 = vec_set_32(input32[0]);
- const vec_t in1 = vec_set_32(input32[1]);
- const auto col0 = reinterpret_cast<const vec_t*>(&weights[0]);
- const auto col1 = reinterpret_cast<const vec_t*>(&weights[OutputDimensions * 4]);
- for (IndexType j = 0; j * OutputSimdWidth < OutputDimensions; ++j)
- vec_add_dpbusd_32x2(outptr[j], in0, col0[j], in1, col1[j]);
- }
- else
-
-#endif
-
-#if defined (USE_SSSE3)
+ static_assert(InputDimensions % 8 == 0);
+ static_assert(OutputDimensions % OutputSimdWidth == 0 || OutputDimensions == 1);
if constexpr (OutputDimensions % OutputSimdWidth == 0)
{
- static_assert(InputDimensions % 16 == 0);
-
- constexpr IndexType NumChunks = InputDimensions / 4;
- constexpr IndexType NumRegs = OutputDimensions / OutputSimdWidth;
-
- const auto input32 = reinterpret_cast<const std::int32_t*>(input);
- const vec_t* biasvec = reinterpret_cast<const vec_t*>(biases);
- vec_t outs[NumRegs];
+ constexpr IndexType NumChunks = InputDimensions / 4;
+ constexpr IndexType NumRegs = OutputDimensions / OutputSimdWidth;
+
+ const auto input32 = reinterpret_cast<const std::int32_t*>(input);
+ const vec_t* biasvec = reinterpret_cast<const vec_t*>(biases);
+ vec_t acc[NumRegs];
+ for (IndexType k = 0; k < NumRegs; ++k)
+ acc[k] = biasvec[k];
+
+ for (IndexType i = 0; i < NumChunks; i += 2)
+ {
+ const vec_t in0 = vec_set_32(input32[i + 0]);
+ const vec_t in1 = vec_set_32(input32[i + 1]);
+ const auto col0 = reinterpret_cast<const vec_t*>(&weights[(i + 0) * OutputDimensions * 4]);
+ const auto col1 = reinterpret_cast<const vec_t*>(&weights[(i + 1) * OutputDimensions * 4]);
for (IndexType k = 0; k < NumRegs; ++k)
- outs[k] = biasvec[k];
-
- for (IndexType i = 0; i < NumChunks; i += 4)
- {
- const vec_t in0 = vec_set_32(input32[i + 0]);
- const vec_t in1 = vec_set_32(input32[i + 1]);
- const vec_t in2 = vec_set_32(input32[i + 2]);
- const vec_t in3 = vec_set_32(input32[i + 3]);
- const auto col0 = reinterpret_cast<const vec_t*>(&weights[(i + 0) * OutputDimensions * 4]);
- const auto col1 = reinterpret_cast<const vec_t*>(&weights[(i + 1) * OutputDimensions * 4]);
- const auto col2 = reinterpret_cast<const vec_t*>(&weights[(i + 2) * OutputDimensions * 4]);
- const auto col3 = reinterpret_cast<const vec_t*>(&weights[(i + 3) * OutputDimensions * 4]);
- for (IndexType k = 0; k < NumRegs; ++k)
- vec_add_dpbusd_32x4(outs[k], in0, col0[k], in1, col1[k], in2, col2[k], in3, col3[k]);
- }
+ vec_add_dpbusd_32x2(acc[k], in0, col0[k], in1, col1[k]);
+ }
- vec_t* outptr = reinterpret_cast<vec_t*>(output);
- for (IndexType k = 0; k < NumRegs; ++k)
- outptr[k] = outs[k];
+ vec_t* outptr = reinterpret_cast<vec_t*>(output);
+ for (IndexType k = 0; k < NumRegs; ++k)
+ outptr[k] = acc[k];
}
else if constexpr (OutputDimensions == 1)
{
- static_assert(InputDimensions % 4 == 0);
-
-#if defined (USE_AVX512)
- if constexpr (PaddedInputDimensions % (SimdWidth * 2) != 0)
- {
- constexpr IndexType NumChunks = PaddedInputDimensions / SimdWidth;
- const auto inputVector256 = reinterpret_cast<const __m256i*>(input);
-
- __m256i sum0 = _mm256_setzero_si256();
- const auto row0 = reinterpret_cast<const __m256i*>(&weights[0]);
-
- for (int j = 0; j < (int)NumChunks; ++j)
- {
- const __m256i in = inputVector256[j];
- m256_add_dpbusd_epi32(sum0, in, row0[j]);
- }
- output[0] = m256_hadd(sum0, biases[0]);
- }
- else
-#endif
- {
-#if defined (USE_AVX512)
- constexpr IndexType NumChunks = PaddedInputDimensions / (SimdWidth * 2);
-#else
- constexpr IndexType NumChunks = PaddedInputDimensions / SimdWidth;
-#endif
- vec_t sum0 = vec_setzero();
- const auto row0 = reinterpret_cast<const vec_t*>(&weights[0]);
-
- for (int j = 0; j < (int)NumChunks; ++j)
- {
- const vec_t in = inputVector[j];
- vec_add_dpbusd_32(sum0, in, row0[j]);
- }
- output[0] = vec_hadd(sum0, biases[0]);
- }
- }
-
-#else
-
-// Use old implementation for the other architectures.
-
- auto output = reinterpret_cast<OutputType*>(buffer);
-
-#if defined(USE_SSE2)
- // At least a multiple of 16, with SSE2.
- static_assert(PaddedInputDimensions % SimdWidth == 0);
- constexpr IndexType NumChunks = PaddedInputDimensions / SimdWidth;
- const __m128i Zeros = _mm_setzero_si128();
- const auto inputVector = reinterpret_cast<const __m128i*>(input);
-
-#elif defined(USE_MMX)
- static_assert(InputDimensions % SimdWidth == 0);
- constexpr IndexType NumChunks = InputDimensions / SimdWidth;
- const __m64 Zeros = _mm_setzero_si64();
- const auto inputVector = reinterpret_cast<const __m64*>(input);
-
-#elif defined(USE_NEON)
- static_assert(PaddedInputDimensions % SimdWidth == 0);
- constexpr IndexType NumChunks = PaddedInputDimensions / SimdWidth;
- const auto inputVector = reinterpret_cast<const int8x8_t*>(input);
-#endif
-
- for (IndexType i = 0; i < OutputDimensions; ++i) {
- const IndexType offset = i * PaddedInputDimensions;
-
-#if defined(USE_SSE2)
- __m128i sumLo = _mm_cvtsi32_si128(biases[i]);
- __m128i sumHi = Zeros;
- const auto row = reinterpret_cast<const __m128i*>(&weights[offset]);
- for (IndexType j = 0; j < NumChunks; ++j) {
- __m128i row_j = _mm_load_si128(&row[j]);
- __m128i input_j = _mm_load_si128(&inputVector[j]);
- __m128i extendedRowLo = _mm_srai_epi16(_mm_unpacklo_epi8(row_j, row_j), 8);
- __m128i extendedRowHi = _mm_srai_epi16(_mm_unpackhi_epi8(row_j, row_j), 8);
- __m128i extendedInputLo = _mm_unpacklo_epi8(input_j, Zeros);
- __m128i extendedInputHi = _mm_unpackhi_epi8(input_j, Zeros);
- __m128i productLo = _mm_madd_epi16(extendedRowLo, extendedInputLo);
- __m128i productHi = _mm_madd_epi16(extendedRowHi, extendedInputHi);
- sumLo = _mm_add_epi32(sumLo, productLo);
- sumHi = _mm_add_epi32(sumHi, productHi);
- }
- __m128i sum = _mm_add_epi32(sumLo, sumHi);
- __m128i sumHigh_64 = _mm_shuffle_epi32(sum, _MM_SHUFFLE(1, 0, 3, 2));
- sum = _mm_add_epi32(sum, sumHigh_64);
- __m128i sum_second_32 = _mm_shufflelo_epi16(sum, _MM_SHUFFLE(1, 0, 3, 2));
- sum = _mm_add_epi32(sum, sum_second_32);
- output[i] = _mm_cvtsi128_si32(sum);
-
-#elif defined(USE_MMX)
- __m64 sumLo = _mm_cvtsi32_si64(biases[i]);
- __m64 sumHi = Zeros;
- const auto row = reinterpret_cast<const __m64*>(&weights[offset]);
- for (IndexType j = 0; j < NumChunks; ++j) {
- __m64 row_j = row[j];
- __m64 input_j = inputVector[j];
- __m64 extendedRowLo = _mm_srai_pi16(_mm_unpacklo_pi8(row_j, row_j), 8);
- __m64 extendedRowHi = _mm_srai_pi16(_mm_unpackhi_pi8(row_j, row_j), 8);
- __m64 extendedInputLo = _mm_unpacklo_pi8(input_j, Zeros);
- __m64 extendedInputHi = _mm_unpackhi_pi8(input_j, Zeros);
- __m64 productLo = _mm_madd_pi16(extendedRowLo, extendedInputLo);
- __m64 productHi = _mm_madd_pi16(extendedRowHi, extendedInputHi);
- sumLo = _mm_add_pi32(sumLo, productLo);
- sumHi = _mm_add_pi32(sumHi, productHi);
- }
- __m64 sum = _mm_add_pi32(sumLo, sumHi);
- sum = _mm_add_pi32(sum, _mm_unpackhi_pi32(sum, sum));
- output[i] = _mm_cvtsi64_si32(sum);
-
-#elif defined(USE_NEON)
- int32x4_t sum = {biases[i]};
- const auto row = reinterpret_cast<const int8x8_t*>(&weights[offset]);
- for (IndexType j = 0; j < NumChunks; ++j) {
- int16x8_t product = vmull_s8(inputVector[j * 2], row[j * 2]);
- product = vmlal_s8(product, inputVector[j * 2 + 1], row[j * 2 + 1]);
- sum = vpadalq_s16(sum, product);
- }
- output[i] = sum[0] + sum[1] + sum[2] + sum[3];
-
-#else
- OutputType sum = biases[i];
- for (IndexType j = 0; j < InputDimensions; ++j) {
- sum += weights[offset + j] * input[j];
+ constexpr IndexType NumChunks = PaddedInputDimensions / SimdWidth;
+ vec_t sum0 = vec_setzero();
+ const auto row0 = reinterpret_cast<const vec_t*>(&weights[0]);
+
+ for (int j = 0; j < (int)NumChunks; ++j)
+ {
+ const vec_t in = inputVector[j];
+ vec_add_dpbusd_32(sum0, in, row0[j]);
}
- output[i] = sum;
-#endif
-
+ output[0] = vec_hadd(sum0, biases[0]);
}
-#if defined(USE_MMX)
- _mm_empty();
-#endif
-#endif
-
-#if (!defined (USE_SSSE3) && defined (USE_SSE2)) || defined (USE_NEON)
- static_assert(SimdWidth <= 16, "Otherwise we run outside of the padding for the output.");
- if constexpr (SimdWidth > OutputDimensions && OutputDimensions != 1)
- for (IndexType i = OutputDimensions; i < SimdWidth; ++i)
- output[i] = 0;
+# undef vec_setzero
+# undef vec_set_32
+# undef vec_add_dpbusd_32
+# undef vec_add_dpbusd_32x2
+# undef vec_add_dpbusd_32x4
+# undef vec_hadd
+# undef vec_haddx4
+#else
+ // Use old implementation for the other architectures.
+ affine_transform_non_ssse3<
+ InputDimensions,
+ PaddedInputDimensions,
+ OutputDimensions>(output, weights, biases, input);
#endif
return output;
projects / stockfish / blobdiff