Add bytes_per_second to transpose benchmark (#14170)

This patch relates to #13735.

Benchmark: [transpose_benchmark.txt](https://github.com/rapidsai/cudf/files/12699834/transpose_benchmark.txt)

Authors:
  - Martin Marenz (https://github.com/Blonck)
  - Mark Harris (https://github.com/harrism)

Approvers:
  - Mark Harris (https://github.com/harrism)
  - Yunsong Wang (https://github.com/PointKernel)

URL: #14170

cpp/benchmarks/transpose/transpose.cpp

@@ -20,17 +20,19 @@
 #include <cudf/column/column_factories.hpp>
 #include <cudf/table/table.hpp>
 #include <cudf/transpose.hpp>
+#include <cudf/types.hpp>
 
 #include <thrust/iterator/counting_iterator.h>
 #include <thrust/iterator/transform_iterator.h>
 
 static void BM_transpose(benchmark::State& state)
 {
-  auto count = state.range(0);
+  auto count                    = state.range(0);
+  constexpr auto column_type_id = cudf::type_id::INT32;
   auto int_column_generator =
     thrust::make_transform_iterator(thrust::counting_iterator(0), [count](int i) {
       return cudf::make_numeric_column(
-        cudf::data_type{cudf::type_id::INT32}, count, cudf::mask_state::ALL_VALID);
+        cudf::data_type{column_type_id}, count, cudf::mask_state::ALL_VALID);
     });
 
   auto input_table = cudf::table(std::vector(int_column_generator, int_column_generator + count));
@@ -40,6 +42,17 @@ static void BM_transpose(benchmark::State& state)
     cuda_event_timer raii(state, true);
     auto output = cudf::transpose(input);
   }
+
+  // Collect memory statistics.
+  auto const bytes_read = static_cast<uint64_t>(input.num_columns()) * input.num_rows() *
+                          sizeof(cudf::id_to_type<column_type_id>);
+  auto const bytes_written = bytes_read;
+  // Account for nullability in input and output.
+  auto const null_bytes = 2 * static_cast<uint64_t>(input.num_columns()) *
+                          cudf::bitmask_allocation_size_bytes(input.num_rows());
+
+  state.SetBytesProcessed(static_cast<int64_t>(state.iterations()) *
+                          (bytes_read + bytes_written + null_bytes));
 }
 
 class Transpose : public cudf::benchmark {};