#blocked = #triton_gpu.blocked<{sizePerThread = [2], threadsPerWarp = [32], warpsPerCTA = [8], order = [0], CTAsPerCGA = [1], CTASplitNum = [1], CTAOrder = [0]}>
module attributes {"triton_gpu.compute-capability" = 89 : i32, "triton_gpu.num-ctas" = 1 : i32, "triton_gpu.num-warps" = 8 : i32, "triton_gpu.threads-per-warp" = 32 : i32} {
  tt.func public @triton__0d1d2de(%arg0: !tt.ptr<f32, 1> {tt.divisibility = 16 : i32}, %arg1: !tt.ptr<bf16, 1> {tt.divisibility = 16 : i32}, %arg2: i32 {tt.divisibility = 16 : i32, tt.max_divisibility = 16 : i32}) attributes {noinline = false} {
    %c512_i32 = arith.constant 512 : i32
    %0 = tt.get_program_id x : i32
    %1 = arith.muli %0, %c512_i32 : i32
    %2 = tt.make_range {end = 512 : i32, start = 0 : i32} : tensor<512xi32, #blocked>
    %3 = tt.splat %1 : (i32) -> tensor<512xi32, #blocked>
    %4 = arith.addi %3, %2 : tensor<512xi32, #blocked>
    %5 = tt.splat %arg0 : (!tt.ptr<f32, 1>) -> tensor<512x!tt.ptr<f32, 1>, #blocked>
    %6 = tt.addptr %5, %4 : tensor<512x!tt.ptr<f32, 1>, #blocked>, tensor<512xi32, #blocked>
    %7 = tt.load %6 {cache = 1 : i32, evict = 1 : i32, isVolatile = false} : tensor<512xf32, #blocked>
    %8 = tt.splat %arg1 : (!tt.ptr<bf16, 1>) -> tensor<512x!tt.ptr<bf16, 1>, #blocked>
    %9 = tt.addptr %8, %4 : tensor<512x!tt.ptr<bf16, 1>, #blocked>, tensor<512xi32, #blocked>
    %10 = arith.truncf %7 : tensor<512xf32, #blocked> to tensor<512xbf16, #blocked>
    tt.store %9, %10 {cache = 1 : i32, evict = 1 : i32} : tensor<512xbf16, #blocked>
    tt.return
  }
}