diff --git a/mlir/examples/toy/Ch2/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch2/mlir/MLIRGen.cpp
index 39ae6a016eb41..a9592bcd7814a 100644
--- a/mlir/examples/toy/Ch2/mlir/MLIRGen.cpp
+++ b/mlir/examples/toy/Ch2/mlir/MLIRGen.cpp
@@ -264,8 +264,7 @@ class MLIRGenImpl {
     // The attribute is a vector with a floating point value per element
     // (number) in the array, see `collectData()` below for more details.
     std::vector<double> data;
-    data.reserve(std::accumulate(lit.getDims().begin(), lit.getDims().end(), 1,
-                                 std::multiplies<int>()));
+    data.reserve(llvm::product_of(lit.getDims()));
     collectData(lit, data);
 
     // The type of this attribute is tensor of 64-bit floating-point with the
diff --git a/mlir/examples/toy/Ch3/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch3/mlir/MLIRGen.cpp
index 0573af699c1f4..8c21951948496 100644
--- a/mlir/examples/toy/Ch3/mlir/MLIRGen.cpp
+++ b/mlir/examples/toy/Ch3/mlir/MLIRGen.cpp
@@ -264,8 +264,7 @@ class MLIRGenImpl {
     // The attribute is a vector with a floating point value per element
     // (number) in the array, see `collectData()` below for more details.
     std::vector<double> data;
-    data.reserve(std::accumulate(lit.getDims().begin(), lit.getDims().end(), 1,
-                                 std::multiplies<int>()));
+    data.reserve(llvm::product_of(lit.getDims()));
     collectData(lit, data);
 
     // The type of this attribute is tensor of 64-bit floating-point with the
diff --git a/mlir/examples/toy/Ch4/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch4/mlir/MLIRGen.cpp
index 7d676f1b39200..6b7ab40299be5 100644
--- a/mlir/examples/toy/Ch4/mlir/MLIRGen.cpp
+++ b/mlir/examples/toy/Ch4/mlir/MLIRGen.cpp
@@ -268,8 +268,7 @@ class MLIRGenImpl {
     // The attribute is a vector with a floating point value per element
     // (number) in the array, see `collectData()` below for more details.
     std::vector<double> data;
-    data.reserve(std::accumulate(lit.getDims().begin(), lit.getDims().end(), 1,
-                                 std::multiplies<int>()));
+    data.reserve(llvm::product_of(lit.getDims()));
     collectData(lit, data);
 
     // The type of this attribute is tensor of 64-bit floating-point with the
diff --git a/mlir/examples/toy/Ch5/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch5/mlir/MLIRGen.cpp
index 7d676f1b39200..6b7ab40299be5 100644
--- a/mlir/examples/toy/Ch5/mlir/MLIRGen.cpp
+++ b/mlir/examples/toy/Ch5/mlir/MLIRGen.cpp
@@ -268,8 +268,7 @@ class MLIRGenImpl {
     // The attribute is a vector with a floating point value per element
     // (number) in the array, see `collectData()` below for more details.
     std::vector<double> data;
-    data.reserve(std::accumulate(lit.getDims().begin(), lit.getDims().end(), 1,
-                                 std::multiplies<int>()));
+    data.reserve(llvm::product_of(lit.getDims()));
     collectData(lit, data);
 
     // The type of this attribute is tensor of 64-bit floating-point with the
diff --git a/mlir/examples/toy/Ch6/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch6/mlir/MLIRGen.cpp
index 7d676f1b39200..6b7ab40299be5 100644
--- a/mlir/examples/toy/Ch6/mlir/MLIRGen.cpp
+++ b/mlir/examples/toy/Ch6/mlir/MLIRGen.cpp
@@ -268,8 +268,7 @@ class MLIRGenImpl {
     // The attribute is a vector with a floating point value per element
     // (number) in the array, see `collectData()` below for more details.
     std::vector<double> data;
-    data.reserve(std::accumulate(lit.getDims().begin(), lit.getDims().end(), 1,
-                                 std::multiplies<int>()));
+    data.reserve(llvm::product_of(lit.getDims()));
     collectData(lit, data);
 
     // The type of this attribute is tensor of 64-bit floating-point with the
diff --git a/mlir/examples/toy/Ch7/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch7/mlir/MLIRGen.cpp
index 75dbc9104a992..73133247335dc 100644
--- a/mlir/examples/toy/Ch7/mlir/MLIRGen.cpp
+++ b/mlir/examples/toy/Ch7/mlir/MLIRGen.cpp
@@ -405,8 +405,7 @@ class MLIRGenImpl {
     // The attribute is a vector with a floating point value per element
     // (number) in the array, see `collectData()` below for more details.
     std::vector<double> data;
-    data.reserve(std::accumulate(lit.getDims().begin(), lit.getDims().end(), 1,
-                                 std::multiplies<int>()));
+    data.reserve(llvm::product_of(lit.getDims()));
     collectData(lit, data);
 
     // The type of this attribute is tensor of 64-bit floating-point with the
diff --git a/mlir/lib/Conversion/MemRefToEmitC/MemRefToEmitC.cpp b/mlir/lib/Conversion/MemRefToEmitC/MemRefToEmitC.cpp
index 2b7bdc9a7b7f8..11f866c103639 100644
--- a/mlir/lib/Conversion/MemRefToEmitC/MemRefToEmitC.cpp
+++ b/mlir/lib/Conversion/MemRefToEmitC/MemRefToEmitC.cpp
@@ -22,6 +22,7 @@
 #include "mlir/IR/TypeRange.h"
 #include "mlir/IR/Value.h"
 #include "mlir/Transforms/DialectConversion.h"
+#include "llvm/ADT/STLExtras.h"
 #include <cstdint>
 #include <numeric>
 
@@ -110,9 +111,7 @@ static Value calculateMemrefTotalSizeBytes(Location loc, MemRefType memrefType,
                      {TypeAttr::get(memrefType.getElementType())}));
 
   IndexType indexType = builder.getIndexType();
-  int64_t numElements = std::accumulate(memrefType.getShape().begin(),
-                                        memrefType.getShape().end(), int64_t{1},
-                                        std::multiplies<int64_t>());
+  int64_t numElements = llvm::product_of(memrefType.getShape());
   emitc::ConstantOp numElementsValue = emitc::ConstantOp::create(
       builder, loc, indexType, builder.getIndexAttr(numElements));
 
diff --git a/mlir/lib/Conversion/TosaToTensor/TosaToTensor.cpp b/mlir/lib/Conversion/TosaToTensor/TosaToTensor.cpp
index 802691c1f7d76..9bf9ca3ae7a89 100644
--- a/mlir/lib/Conversion/TosaToTensor/TosaToTensor.cpp
+++ b/mlir/lib/Conversion/TosaToTensor/TosaToTensor.cpp
@@ -18,6 +18,7 @@
 #include "mlir/Dialect/Tosa/Utils/ConversionUtils.h"
 #include "mlir/IR/PatternMatch.h"
 #include "mlir/Transforms/DialectConversion.h"
+#include "llvm/ADT/STLExtras.h"
 
 #include <numeric>
 
@@ -70,8 +71,7 @@ TensorType inferReshapeExpandedType(TensorType inputType,
 
         // Calculate the product of all elements in 'newShape' except for the -1
         // placeholder, which we discard by negating the result.
-        int64_t totalSizeNoPlaceholder = -std::accumulate(
-            newShape.begin(), newShape.end(), 1, std::multiplies<int64_t>());
+        int64_t totalSizeNoPlaceholder = -llvm::product_of(newShape);
 
         // If there is a 0 component in 'newShape', resolve the placeholder as
         // 0.
diff --git a/mlir/lib/Conversion/VectorToAMX/VectorToAMX.cpp b/mlir/lib/Conversion/VectorToAMX/VectorToAMX.cpp
index 79c2f23c8e7f3..245a3efe98ecc 100644
--- a/mlir/lib/Conversion/VectorToAMX/VectorToAMX.cpp
+++ b/mlir/lib/Conversion/VectorToAMX/VectorToAMX.cpp
@@ -20,6 +20,7 @@
 #include "mlir/Pass/Pass.h"
 #include "mlir/Transforms/GreedyPatternRewriteDriver.h"
 
+#include "llvm/ADT/STLExtras.h"
 #include "llvm/Support/DebugLog.h"
 
 #include <numeric>
@@ -265,8 +266,7 @@ loadStoreFromTransfer(PatternRewriter &rewriter,
   if (isPacked)
     src = collapseLastDim(rewriter, src);
   int64_t rows = vecShape[0];
-  int64_t cols = std::accumulate(vecShape.begin() + 1, vecShape.end(), 1,
-                                 std::multiplies<int64_t>());
+  int64_t cols = llvm::product_of(vecShape.drop_front());
   auto tileType = amx::TileType::get({rows, cols}, vecTy.getElementType());
 
   Value zeroIndex = rewriter.createOrFold<arith::ConstantIndexOp>(loc, 0);
@@ -336,8 +336,7 @@ static TypedValue<amx::TileType> loadTile(PatternRewriter &rewriter,
 
   ArrayRef<int64_t> shape = vecTy.getShape();
   int64_t rows = shape[0];
-  int64_t cols = std::accumulate(shape.begin() + 1, shape.end(), 1,
-                                 std::multiplies<int64_t>());
+  int64_t cols = llvm::product_of(shape.drop_front());
   auto tileType = amx::TileType::get({rows, cols}, vecTy.getElementType());
 
   return amx::TileLoadOp::create(rewriter, loc, tileType, buf,
diff --git a/mlir/lib/Conversion/VectorToSCF/VectorToSCF.cpp b/mlir/lib/Conversion/VectorToSCF/VectorToSCF.cpp
index c45c45e4712f3..c9eba6962e6a4 100644
--- a/mlir/lib/Conversion/VectorToSCF/VectorToSCF.cpp
+++ b/mlir/lib/Conversion/VectorToSCF/VectorToSCF.cpp
@@ -26,6 +26,7 @@
 #include "mlir/IR/Builders.h"
 #include "mlir/Pass/Pass.h"
 #include "mlir/Transforms/GreedyPatternRewriteDriver.h"
+#include "llvm/ADT/STLExtras.h"
 
 namespace mlir {
 #define GEN_PASS_DEF_CONVERTVECTORTOSCF
@@ -760,8 +761,7 @@ struct DecomposePrintOpConversion : public VectorToSCFPattern<vector::PrintOp> {
     if (vectorType.getRank() != 1) {
       // Flatten n-D vectors to 1D. This is done to allow indexing with a
       // non-constant value.
-      auto flatLength = std::accumulate(shape.begin(), shape.end(), 1,
-                                        std::multiplies<int64_t>());
+      int64_t flatLength = llvm::product_of(shape);
       auto flatVectorType =
           VectorType::get({flatLength}, vectorType.getElementType());
       value = vector::ShapeCastOp::create(rewriter, loc, flatVectorType, value);
diff --git a/mlir/lib/Conversion/XeGPUToXeVM/XeGPUToXeVM.cpp b/mlir/lib/Conversion/XeGPUToXeVM/XeGPUToXeVM.cpp
index 9ead1d89069d6..71687b1479a7d 100644
--- a/mlir/lib/Conversion/XeGPUToXeVM/XeGPUToXeVM.cpp
+++ b/mlir/lib/Conversion/XeGPUToXeVM/XeGPUToXeVM.cpp
@@ -23,6 +23,7 @@
 #include "mlir/Dialect/XeGPU/IR/XeGPU.h"
 #include "mlir/Pass/Pass.h"
 #include "mlir/Support/LLVM.h"
+#include "llvm/ADT/STLExtras.h"
 #include "llvm/Support/FormatVariadic.h"
 
 #include "mlir/IR/BuiltinTypes.h"
@@ -774,9 +775,7 @@ struct ConvertXeGPUToXeVMPass
       if (rank < 1 || type.getNumElements() == 1)
         return elemType;
       // Otherwise, convert the vector to a flat vector type.
-      int64_t sum =
-          std::accumulate(type.getShape().begin(), type.getShape().end(),
-                          int64_t{1}, std::multiplies<int64_t>());
+      int64_t sum = llvm::product_of(type.getShape());
       return VectorType::get(sum, elemType);
     });
     typeConverter.addConversion([&](xegpu::TensorDescType type) -> Type {
diff --git a/mlir/lib/Dialect/Arith/Utils/Utils.cpp b/mlir/lib/Dialect/Arith/Utils/Utils.cpp
index b1fc9aa57c3ba..f54bafff58322 100644
--- a/mlir/lib/Dialect/Arith/Utils/Utils.cpp
+++ b/mlir/lib/Dialect/Arith/Utils/Utils.cpp
@@ -351,9 +351,9 @@ Value createProduct(OpBuilder &builder, Location loc, ArrayRef<Value> values,
   Value one = ConstantOp::create(builder, loc, resultType,
                                  builder.getOneAttr(resultType));
   ArithBuilder arithBuilder(builder, loc);
-  return std::accumulate(
-      values.begin(), values.end(), one,
-      [&arithBuilder](Value acc, Value v) { return arithBuilder.mul(acc, v); });
+  return llvm::accumulate(values, one, [&arithBuilder](Value acc, Value v) {
+    return arithBuilder.mul(acc, v);
+  });
 }
 
 /// Map strings to float types.
diff --git a/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp b/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp
index 19eba6beacd86..b5f8ddaadacdf 100644
--- a/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp
+++ b/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp
@@ -2460,8 +2460,7 @@ static LogicalResult verifyDistributedType(Type expanded, Type distributed,
              << dDim << ")";
     scales[i] = eDim / dDim;
   }
-  if (std::accumulate(scales.begin(), scales.end(), 1,
-                      std::multiplies<int64_t>()) != warpSize)
+  if (llvm::product_of(scales) != warpSize)
     return op->emitOpError()
            << "incompatible distribution dimensions from " << expandedVecType
            << " to " << distributedVecType << " with warp size = " << warpSize;
diff --git a/mlir/lib/Dialect/GPU/Utils/DistributionUtils.cpp b/mlir/lib/Dialect/GPU/Utils/DistributionUtils.cpp
index 88f531f394765..572b746358379 100644
--- a/mlir/lib/Dialect/GPU/Utils/DistributionUtils.cpp
+++ b/mlir/lib/Dialect/GPU/Utils/DistributionUtils.cpp
@@ -15,6 +15,7 @@
 #include "mlir/Dialect/Arith/IR/Arith.h"
 #include "mlir/IR/Value.h"
 #include "llvm/ADT/DenseMap.h"
+#include "llvm/ADT/STLExtras.h"
 
 #include <numeric>
 
@@ -118,8 +119,7 @@ bool WarpDistributionPattern::delinearizeLaneId(
       return false;
     sizes.push_back(large / small);
   }
-  if (std::accumulate(sizes.begin(), sizes.end(), 1,
-                      std::multiplies<int64_t>()) != warpSize)
+  if (llvm::product_of(sizes) != warpSize)
     return false;
 
   AffineExpr s0, s1;
diff --git a/mlir/lib/Dialect/Linalg/Transforms/ShardingInterfaceImpl.cpp b/mlir/lib/Dialect/Linalg/Transforms/ShardingInterfaceImpl.cpp
index f277c5f5be5fc..0ae2a9cc0318c 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/ShardingInterfaceImpl.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/ShardingInterfaceImpl.cpp
@@ -266,9 +266,8 @@ struct StructuredOpShardingInterface
     LinalgOp linalgOp = llvm::cast<LinalgOp>(op);
     SmallVector<utils::IteratorType> iteratorTypes =
         linalgOp.getIteratorTypesArray();
-    unsigned reductionItersCount = std::accumulate(
-        iteratorTypes.begin(), iteratorTypes.end(), 0,
-        [](unsigned count, utils::IteratorType iter) {
+    unsigned reductionItersCount = llvm::accumulate(
+        iteratorTypes, 0u, [](unsigned count, utils::IteratorType iter) {
           return count + (iter == utils::IteratorType::reduction);
         });
     shard::ReductionKind reductionKind = getReductionKindOfLinalgOp(linalgOp);
diff --git a/mlir/lib/Dialect/Quant/Utils/UniformSupport.cpp b/mlir/lib/Dialect/Quant/Utils/UniformSupport.cpp
index b66390819103e..8c4f80f13e2f6 100644
--- a/mlir/lib/Dialect/Quant/Utils/UniformSupport.cpp
+++ b/mlir/lib/Dialect/Quant/Utils/UniformSupport.cpp
@@ -8,6 +8,7 @@
 
 #include "mlir/Dialect/Quant/Utils/UniformSupport.h"
 #include "mlir/IR/BuiltinTypes.h"
+#include "llvm/ADT/STLExtras.h"
 #include <numeric>
 
 using namespace mlir;
@@ -76,9 +77,7 @@ UniformQuantizedPerAxisValueConverter::convert(DenseFPElementsAttr attr) {
   // using the right quantization parameters.
   int64_t flattenIndex = 0;
   auto shape = type.getShape();
-  int64_t chunkSize =
-      std::accumulate(std::next(shape.begin(), quantizationDim + 1),
-                      shape.end(), 1, std::multiplies<int64_t>());
+  int64_t chunkSize = llvm::product_of(shape.drop_front(quantizationDim + 1));
   Type newElementType = IntegerType::get(attr.getContext(), storageBitWidth);
   return attr.mapValues(newElementType, [&](const APFloat &old) {
     int chunkIndex = (flattenIndex++) / chunkSize;
diff --git a/mlir/lib/Dialect/SPIRV/IR/SPIRVOps.cpp b/mlir/lib/Dialect/SPIRV/IR/SPIRVOps.cpp
index 55119984d9cc1..fe50865bb7c49 100644
--- a/mlir/lib/Dialect/SPIRV/IR/SPIRVOps.cpp
+++ b/mlir/lib/Dialect/SPIRV/IR/SPIRVOps.cpp
@@ -400,7 +400,7 @@ LogicalResult spirv::CompositeConstructOp::verify() {
       return emitOpError("operand element type mismatch: expected to be ")
              << resultType.getElementType() << ", but provided " << elementType;
   }
-  unsigned totalCount = std::accumulate(sizes.begin(), sizes.end(), 0);
+  unsigned totalCount = llvm::sum_of(sizes);
   if (totalCount != cType.getNumElements())
     return emitOpError("has incorrect number of operands: expected ")
            << cType.getNumElements() << ", but provided " << totalCount;
diff --git a/mlir/lib/Dialect/Shard/IR/ShardOps.cpp b/mlir/lib/Dialect/Shard/IR/ShardOps.cpp
index 08fccfa25c0c7..135c03311ce0e 100644
--- a/mlir/lib/Dialect/Shard/IR/ShardOps.cpp
+++ b/mlir/lib/Dialect/Shard/IR/ShardOps.cpp
@@ -1010,18 +1010,6 @@ static LogicalResult verifyInGroupDevice(Location loc, StringRef deviceName,
   return success();
 }
 
-template <typename It>
-static auto product(It begin, It end) {
-  using ElementType = std::decay_t<decltype(*begin)>;
-  return std::accumulate(begin, end, static_cast<ElementType>(1),
-                         std::multiplies<ElementType>());
-}
-
-template <typename R>
-static auto product(R &&range) {
-  return product(adl_begin(range), adl_end(range));
-}
-
 static LogicalResult verifyDimensionCompatibility(Location loc,
                                                   int64_t expectedDimSize,
                                                   int64_t resultDimSize,
diff --git a/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp b/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp
index c51b5e9cbfc78..00f84bc43f444 100644
--- a/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp
+++ b/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp
@@ -2368,9 +2368,10 @@ llvm::LogicalResult tosa::ReshapeOp::verify() {
       }
     }
 
-    int64_t newShapeElementsNum = std::accumulate(
-        shapeValues.begin(), shapeValues.end(), 1LL,
-        [](int64_t acc, int64_t dim) { return (dim > 0) ? acc * dim : acc; });
+    int64_t newShapeElementsNum =
+        llvm::accumulate(shapeValues, int64_t(1), [](int64_t acc, int64_t dim) {
+          return (dim > 0) ? acc * dim : acc;
+        });
     bool isStaticNewShape =
         llvm::all_of(shapeValues, [](int64_t s) { return s > 0; });
     if ((isStaticNewShape && inputElementsNum != newShapeElementsNum) ||
diff --git a/mlir/lib/Dialect/Tosa/Transforms/TosaFolders.cpp b/mlir/lib/Dialect/Tosa/Transforms/TosaFolders.cpp
index d33ebe397cd35..5786f53b8133d 100644
--- a/mlir/lib/Dialect/Tosa/Transforms/TosaFolders.cpp
+++ b/mlir/lib/Dialect/Tosa/Transforms/TosaFolders.cpp
@@ -20,6 +20,7 @@
 #include "mlir/IR/BuiltinTypes.h"
 #include "mlir/IR/DialectResourceBlobManager.h"
 #include "mlir/IR/Matchers.h"
+#include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SmallVector.h"
 
 using namespace mlir;
@@ -375,8 +376,7 @@ llvm::APInt calculateReducedValue(const mlir::ElementsAttr &oldTensorAttr,
   for (int64_t reductionAxisVal = 1; reductionAxisVal < oldShape[reductionAxis];
        ++reductionAxisVal) {
 
-    int64_t stride = std::accumulate(oldShape.begin() + reductionAxis + 1,
-                                     oldShape.end(), 1, std::multiplies<int>());
+    int64_t stride = llvm::product_of(oldShape.drop_front(reductionAxis + 1));
     int64_t index = indexAtOldTensor + stride * reductionAxisVal;
     reducedValue =
         OperationType::calcOneElement(reducedValue, oldTensor[index]);
@@ -424,8 +424,7 @@ struct ReduceConstantOptimization : public OpRewritePattern<OperationType> {
     auto oldShape = shapedOldElementsValues.getShape();
     auto newShape = resultType.getShape();
 
-    auto newNumOfElements = std::accumulate(newShape.begin(), newShape.end(), 1,
-                                            std::multiplies<int>());
+    int64_t newNumOfElements = llvm::product_of(newShape);
     llvm::SmallVector<APInt> newReducedTensor(newNumOfElements);
 
     for (int64_t reductionIndex = 0; reductionIndex < newNumOfElements;
diff --git a/mlir/lib/Dialect/Utils/IndexingUtils.cpp b/mlir/lib/Dialect/Utils/IndexingUtils.cpp
index e1648ab99ff25..305b06eb38fdf 100644
--- a/mlir/lib/Dialect/Utils/IndexingUtils.cpp
+++ b/mlir/lib/Dialect/Utils/IndexingUtils.cpp
@@ -81,21 +81,10 @@ SmallVector<int64_t> mlir::computeElementwiseMul(ArrayRef<int64_t> v1,
   return computeElementwiseMulImpl(v1, v2);
 }
 
-int64_t mlir::computeSum(ArrayRef<int64_t> basis) {
-  assert(llvm::all_of(basis, [](int64_t s) { return s > 0; }) &&
-         "basis must be nonnegative");
-  if (basis.empty())
-    return 0;
-  return std::accumulate(basis.begin(), basis.end(), 1, std::plus<int64_t>());
-}
-
 int64_t mlir::computeProduct(ArrayRef<int64_t> basis) {
   assert(llvm::all_of(basis, [](int64_t s) { return s > 0; }) &&
          "basis must be nonnegative");
-  if (basis.empty())
-    return 1;
-  return std::accumulate(basis.begin(), basis.end(), 1,
-                         std::multiplies<int64_t>());
+  return llvm::product_of(basis);
 }
 
 int64_t mlir::linearize(ArrayRef<int64_t> offsets, ArrayRef<int64_t> basis) {
@@ -158,19 +147,11 @@ SmallVector<AffineExpr> mlir::computeElementwiseMul(ArrayRef<AffineExpr> v1,
 }
 
 AffineExpr mlir::computeSum(MLIRContext *ctx, ArrayRef<AffineExpr> basis) {
-  if (basis.empty())
-    return getAffineConstantExpr(0, ctx);
-  return std::accumulate(basis.begin(), basis.end(),
-                         getAffineConstantExpr(0, ctx),
-                         std::plus<AffineExpr>());
+  return llvm::sum_of(basis, getAffineConstantExpr(0, ctx));
 }
 
 AffineExpr mlir::computeProduct(MLIRContext *ctx, ArrayRef<AffineExpr> basis) {
-  if (basis.empty())
-    return getAffineConstantExpr(1, ctx);
-  return std::accumulate(basis.begin(), basis.end(),
-                         getAffineConstantExpr(1, ctx),
-                         std::multiplies<AffineExpr>());
+  return llvm::product_of(basis, getAffineConstantExpr(1, ctx));
 }
 
 AffineExpr mlir::linearize(MLIRContext *ctx, ArrayRef<AffineExpr> offsets,
diff --git a/mlir/lib/Dialect/Utils/ReshapeOpsUtils.cpp b/mlir/lib/Dialect/Utils/ReshapeOpsUtils.cpp
index 7b2734d8c22fc..6e9118e1f7b0b 100644
--- a/mlir/lib/Dialect/Utils/ReshapeOpsUtils.cpp
+++ b/mlir/lib/Dialect/Utils/ReshapeOpsUtils.cpp
@@ -374,11 +374,11 @@ mlir::composeReassociationIndices(
   if (consumerReassociations.empty())
     return composedIndices;
 
-  size_t consumerDims = std::accumulate(
-      consumerReassociations.begin(), consumerReassociations.end(), 0,
-      [](size_t all, ReassociationIndicesRef indices) {
-        return all + indices.size();
-      });
+  size_t consumerDims =
+      llvm::accumulate(consumerReassociations, size_t(0),
+                       [](size_t all, ReassociationIndicesRef indices) {
+                         return all + indices.size();
+                       });
   if (producerReassociations.size() != consumerDims)
     return std::nullopt;
 
diff --git a/mlir/lib/Dialect/Vector/IR/VectorOps.cpp b/mlir/lib/Dialect/Vector/IR/VectorOps.cpp
index a7e3ba8ca3285..58256b0ade9f6 100644
--- a/mlir/lib/Dialect/Vector/IR/VectorOps.cpp
+++ b/mlir/lib/Dialect/Vector/IR/VectorOps.cpp
@@ -2496,8 +2496,7 @@ struct ToElementsOfBroadcast final : OpRewritePattern<ToElementsOp> {
     auto srcElems = vector::ToElementsOp::create(
         rewriter, toElementsOp.getLoc(), bcastOp.getSource());
 
-    int64_t dstCount = std::accumulate(dstShape.begin(), dstShape.end(), 1,
-                                       std::multiplies<int64_t>());
+    int64_t dstCount = llvm::product_of(dstShape);
 
     SmallVector<Value> replacements;
     replacements.reserve(dstCount);
diff --git a/mlir/lib/Dialect/Vector/Transforms/LowerVectorShapeCast.cpp b/mlir/lib/Dialect/Vector/Transforms/LowerVectorShapeCast.cpp
index c5f22b2eafeb7..0eba0b12259ac 100644
--- a/mlir/lib/Dialect/Vector/Transforms/LowerVectorShapeCast.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/LowerVectorShapeCast.cpp
@@ -21,6 +21,7 @@
 #include "mlir/IR/Location.h"
 #include "mlir/IR/PatternMatch.h"
 #include "mlir/IR/TypeUtilities.h"
+#include "llvm/ADT/STLExtras.h"
 #include <numeric>
 
 #define DEBUG_TYPE "vector-shape-cast-lowering"
@@ -166,10 +167,7 @@ class ShapeCastOpRewritePattern : public OpRewritePattern<vector::ShapeCastOp> {
     const VectorType resultType = shapeCast.getResultVectorType();
     const ArrayRef<int64_t> resultShape = resultType.getShape();
 
-    const int64_t nSlices =
-        std::accumulate(sourceShape.begin(), sourceShape.begin() + sourceDim, 1,
-                        std::multiplies<int64_t>());
-
+    const int64_t nSlices = llvm::product_of(sourceShape.take_front(sourceDim));
     SmallVector<int64_t> extractIndex(sourceDim, 0);
     SmallVector<int64_t> insertIndex(resultDim, 0);
     Value result = ub::PoisonOp::create(rewriter, loc, resultType);
diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorDropLeadUnitDim.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorDropLeadUnitDim.cpp
index 963b2c803bc5a..aa2dd89b182e1 100644
--- a/mlir/lib/Dialect/Vector/Transforms/VectorDropLeadUnitDim.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/VectorDropLeadUnitDim.cpp
@@ -15,6 +15,7 @@
 #include "mlir/Dialect/Vector/Utils/VectorUtils.h"
 #include "mlir/IR/Builders.h"
 #include "mlir/IR/TypeUtilities.h"
+#include "llvm/ADT/STLExtras.h"
 
 #define DEBUG_TYPE "vector-drop-unit-dim"
 
@@ -557,8 +558,7 @@ struct CastAwayConstantMaskLeadingOneDim
     // If any of the dropped unit dims has a size of `0`, the entire mask is a
     // zero mask, else the unit dim has no effect on the mask.
     int64_t flatLeadingSize =
-        std::accumulate(dimSizes.begin(), dimSizes.begin() + dropDim + 1,
-                        static_cast<int64_t>(1), std::multiplies<int64_t>());
+        llvm::product_of(dimSizes.take_front(dropDim + 1));
     SmallVector<int64_t> newDimSizes = {flatLeadingSize};
     newDimSizes.append(dimSizes.begin() + dropDim + 1, dimSizes.end());
 
diff --git a/mlir/lib/Dialect/XeGPU/Utils/XeGPUUtils.cpp b/mlir/lib/Dialect/XeGPU/Utils/XeGPUUtils.cpp
index b72d5648b29f9..2c56a438ea62c 100644
--- a/mlir/lib/Dialect/XeGPU/Utils/XeGPUUtils.cpp
+++ b/mlir/lib/Dialect/XeGPU/Utils/XeGPUUtils.cpp
@@ -52,8 +52,7 @@ mlir::xegpu::getDistributedVectorType(xegpu::TensorDescType tdescTy) {
   // compute sgSize by multiply elements of laneLayout
   // e.g. for 2D layout, sgSize = laneLayout[0] * laneLayout[1]
   // e.g. for 1D layout, sgSize = laneLayout[0]
-  auto sgSize = std::accumulate(laneLayout.begin(), laneLayout.end(), 1,
-                                std::multiplies<int64_t>());
+  int64_t sgSize = llvm::product_of(laneLayout);
 
   // Case 1: regular loads/stores
   auto scatterAttr = tdescTy.getEncodingOfType<ScatterTensorDescAttr>();
diff --git a/mlir/lib/IR/Operation.cpp b/mlir/lib/IR/Operation.cpp
index 8bcfa465e4a22..ce421f4bf7e0e 100644
--- a/mlir/lib/IR/Operation.cpp
+++ b/mlir/lib/IR/Operation.cpp
@@ -18,6 +18,7 @@
 #include "mlir/IR/PatternMatch.h"
 #include "mlir/IR/TypeUtilities.h"
 #include "mlir/Interfaces/FoldInterfaces.h"
+#include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/Support/ErrorHandling.h"
 #include <numeric>
@@ -1274,10 +1275,7 @@ LogicalResult OpTrait::impl::verifyValueSizeAttr(Operation *op,
     return op->emitOpError("'")
            << attrName << "' attribute cannot have negative elements";
 
-  size_t totalCount =
-      std::accumulate(sizes.begin(), sizes.end(), 0,
-                      [](unsigned all, int32_t one) { return all + one; });
-
+  size_t totalCount = llvm::sum_of(sizes, size_t(0));
   if (totalCount != expectedCount)
     return op->emitOpError()
            << valueGroupName << " count (" << expectedCount
diff --git a/mlir/lib/IR/OperationSupport.cpp b/mlir/lib/IR/OperationSupport.cpp
index 394ac7765ed61..2a37f3860fe00 100644
--- a/mlir/lib/IR/OperationSupport.cpp
+++ b/mlir/lib/IR/OperationSupport.cpp
@@ -406,15 +406,13 @@ OperandRangeRange::OperandRangeRange(OperandRange operands,
 OperandRange OperandRangeRange::join() const {
   const OwnerT &owner = getBase();
   ArrayRef<int32_t> sizeData = llvm::cast<DenseI32ArrayAttr>(owner.second);
-  return OperandRange(owner.first,
-                      std::accumulate(sizeData.begin(), sizeData.end(), 0));
+  return OperandRange(owner.first, llvm::sum_of(sizeData));
 }
 
 OperandRange OperandRangeRange::dereference(const OwnerT &object,
                                             ptrdiff_t index) {
   ArrayRef<int32_t> sizeData = llvm::cast<DenseI32ArrayAttr>(object.second);
-  uint32_t startIndex =
-      std::accumulate(sizeData.begin(), sizeData.begin() + index, 0);
+  uint32_t startIndex = llvm::sum_of(sizeData.take_front(index));
   return OperandRange(object.first + startIndex, *(sizeData.begin() + index));
 }
 
@@ -565,8 +563,7 @@ MutableOperandRange MutableOperandRangeRange::dereference(const OwnerT &object,
                                                           ptrdiff_t index) {
   ArrayRef<int32_t> sizeData =
       llvm::cast<DenseI32ArrayAttr>(object.second.getValue());
-  uint32_t startIndex =
-      std::accumulate(sizeData.begin(), sizeData.begin() + index, 0);
+  uint32_t startIndex = llvm::sum_of(sizeData.take_front(index));
   return object.first.slice(
       startIndex, *(sizeData.begin() + index),
       MutableOperandRange::OperandSegment(index, object.second));
diff --git a/mlir/lib/IR/TypeUtilities.cpp b/mlir/lib/IR/TypeUtilities.cpp
index d2d115ea39277..e438631ffe1f5 100644
--- a/mlir/lib/IR/TypeUtilities.cpp
+++ b/mlir/lib/IR/TypeUtilities.cpp
@@ -104,8 +104,8 @@ LogicalResult mlir::verifyCompatibleShapes(TypeRange types1, TypeRange types2) {
 LogicalResult mlir::verifyCompatibleDims(ArrayRef<int64_t> dims) {
   if (dims.empty())
     return success();
-  auto staticDim = std::accumulate(
-      dims.begin(), dims.end(), dims.front(), [](auto fold, auto dim) {
+  auto staticDim =
+      llvm::accumulate(dims, dims.front(), [](auto fold, auto dim) {
         return ShapedType::isDynamic(dim) ? fold : dim;
       });
   return success(llvm::all_of(dims, [&](auto dim) {
diff --git a/mlir/lib/Rewrite/ByteCode.cpp b/mlir/lib/Rewrite/ByteCode.cpp
index 33fbd2a9579f0..42843ea1780c4 100644
--- a/mlir/lib/Rewrite/ByteCode.cpp
+++ b/mlir/lib/Rewrite/ByteCode.cpp
@@ -1835,8 +1835,7 @@ executeGetOperandsResults(RangeT values, Operation *op, unsigned index,
       return nullptr;
 
     ArrayRef<int32_t> segments = segmentAttr;
-    unsigned startIndex =
-        std::accumulate(segments.begin(), segments.begin() + index, 0);
+    unsigned startIndex = llvm::sum_of(segments.take_front(index));
     values = values.slice(startIndex, *std::next(segments.begin(), index));
 
     LDBG() << "  * Extracting range[" << startIndex << ", "
diff --git a/mlir/lib/Target/LLVMIR/ModuleTranslation.cpp b/mlir/lib/Target/LLVMIR/ModuleTranslation.cpp
index 5a3eb209f0a92..845a14f34c016 100644
--- a/mlir/lib/Target/LLVMIR/ModuleTranslation.cpp
+++ b/mlir/lib/Target/LLVMIR/ModuleTranslation.cpp
@@ -922,8 +922,7 @@ llvm::CallInst *mlir::LLVM::detail::createIntrinsicCall(
     assert(opBundleSizes.size() == opBundleTagsAttr.size() &&
            "operand bundles and tags do not match");
 
-    numOpBundleOperands =
-        std::accumulate(opBundleSizes.begin(), opBundleSizes.end(), size_t(0));
+    numOpBundleOperands = llvm::sum_of(opBundleSizes);
     assert(numOpBundleOperands <= intrOp->getNumOperands() &&
            "operand bundle operands is more than the number of operands");
 
diff --git a/mlir/tools/mlir-tblgen/OpDefinitionsGen.cpp b/mlir/tools/mlir-tblgen/OpDefinitionsGen.cpp
index 969011546985b..daae3c79ffd43 100644
--- a/mlir/tools/mlir-tblgen/OpDefinitionsGen.cpp
+++ b/mlir/tools/mlir-tblgen/OpDefinitionsGen.cpp
@@ -3513,9 +3513,9 @@ void OpEmitter::genCodeForAddingArgAndRegionForBuilder(
         body << "(" << operandName << " ? 1 : 0)";
       } else if (operand.isVariadicOfVariadic()) {
         body << llvm::formatv(
-            "static_cast<int32_t>(std::accumulate({0}.begin(), {0}.end(), 0, "
+            "llvm::accumulate({0}, int32_t(0), "
             "[](int32_t curSum, ::mlir::ValueRange range) {{ return curSum + "
-            "static_cast<int32_t>(range.size()); }))",
+            "static_cast<int32_t>(range.size()); })",
             operandName);
       } else {
         body << "static_cast<int32_t>(" << getArgumentName(op, i) << ".size())";