Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Test][Don't review] load B twice which means output 16 data in each workgroup #23334

Draft
wants to merge 2 commits into
base: main
Choose a base branch
from
+93 −21
Draft

[Test][Don't review] load B twice which means output 16 data in each workgroup #23334

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
602000a
loadB twice which means output 16 data in each workgroup
qjia7 Jan 13, 2025
91ab40a
reduce the writing times to workgroup memory
qjia7 Jan 13, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
114 changes: 93 additions & 21 deletions onnxruntime/contrib_ops/webgpu/quantization/matmul_nbits.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -74,12 +74,10 @@ Status MatMulNBitsProgram::GenerateShaderCode(ShaderHelper& shader) const {
" }\n"
"}\n"
<< "var<workgroup> sub_a: array<input_a_value_t, " << a_length_per_tile << ">;\n"
<< "var<workgroup> inter_results: array<array<output_value_t, " << WorkgroupSizeX() << ">, " << WorkgroupSizeY() << ">;\n";
std::string offset = "workgroup_idx * " + std::to_string(WorkgroupSizeY());
shader.MainFunctionBody() << " let output_indices = " << y.OffsetToIndices(offset) << ";\n"
<< " let col = output_indices[2];\n"
" let row = output_indices[1];\n"
" let batch = output_indices[0];\n";
<< "var<workgroup> inter_results: array<array<output_value_t, " << WorkgroupSizeX() << ">, " << WorkgroupSizeY() * 2 << ">;\n";
shader.MainFunctionBody() << " let col = workgroup_id.x * " << WorkgroupSizeY() * 2 << ";\n"
<< " let row = workgroup_id.y * " << tile_m_ << ";\n"
<< " let batch = workgroup_id.z;\n";
} else {
ORT_ENFORCE(tile_m_ < WorkgroupSizeY(), "tile_m must be less than or equal to WorkgroupSizeY.");
ORT_ENFORCE(WorkgroupSizeX() == WorkgroupSizeY(), "WorkgroupSizeX must be equal to WorkgroupSizeY.");
Expand All @@ -99,6 +97,8 @@ Status MatMulNBitsProgram::GenerateShaderCode(ShaderHelper& shader) const {
}
shader.MainFunctionBody() << " let n_blocks_per_col = uniforms.input_b_shape[1];\n"
<< " let num_tiles = (n_blocks_per_col - 1) / " << blocks_per_tile << " + 1;\n"
<< " var inter_res0 = output_element_t(0);\n"
<< " var inter_res1 = output_element_t(0);\n"
// Loop over shared dimension.
<< " for (var tile: u32 = 0; tile < num_tiles; tile += 1) {\n"
<< " let a_col_start = tile * " << a_length_per_tile << ";\n"
Expand All @@ -113,10 +113,16 @@ Status MatMulNBitsProgram::GenerateShaderCode(ShaderHelper& shader) const {
}
}
shader.MainFunctionBody() << " }\n"
" workgroupBarrier();\n"
// Each thread processes one block.
" let b_row = col + local_id.y;\n"
<< " let block = tile * " << blocks_per_tile << " + local_id.x;\n";
" workgroupBarrier();\n";
if (tile_m_ == 1) {
shader.MainFunctionBody() << " var b_row = col + local_id.y;\n"
<< " let block = tile * " << blocks_per_tile << " + local_id.x;\n";
} else {
// Each thread processes one block.
shader.MainFunctionBody() << " let b_row = col + local_id.y;\n"
<< " let block = tile * " << blocks_per_tile << " + local_id.x;\n";
}

if (has_zero_points_) {
const auto& zero_points = shader.AddInput("zero_points", ShaderUsage::UseUniform);
shader.MainFunctionBody() << " let zero_point_bytes_per_col = (n_blocks_per_col + 1) / 2;\n"
Expand All @@ -133,7 +139,7 @@ Status MatMulNBitsProgram::GenerateShaderCode(ShaderHelper& shader) const {
}
shader.MainFunctionBody() << " var scale = output_element_t(0);\n"
" var b_data = input_b_value_t(0);\n"
<< " if (block < n_blocks_per_col) {\n"
<< " if (b_row < uniforms.input_b_shape[0] && block < n_blocks_per_col) {\n"
<< " scale = " << scales.GetByOffset("b_row * n_blocks_per_col + block") << ";\n"
<< " b_data = " << b.GetByIndices("input_b_indices_t(b_row, block, 0)") << ";\n"
<< " }\n"
Expand All @@ -158,13 +164,13 @@ Status MatMulNBitsProgram::GenerateShaderCode(ShaderHelper& shader) const {
if (tile_m_ == 1) {
switch (a.NumComponents()) {
case 1:
shader.MainFunctionBody() << " inter_results[local_id.y][local_id.x] += dot(vec4<output_element_t>(sub_a[word_offset], sub_a[word_offset + 1], sub_a[word_offset + 2], sub_a[word_offset + 3]), b_dequantized_values[0]) + dot(vec4<output_element_t>(sub_a[word_offset + 4], sub_a[word_offset + 5], sub_a[word_offset + 6], sub_a[word_offset + 7]), b_dequantized_values[1]);\n";
shader.MainFunctionBody() << " inter_res0 += dot(vec4<output_element_t>(sub_a[word_offset], sub_a[word_offset + 1], sub_a[word_offset + 2], sub_a[word_offset + 3]), b_dequantized_values[0]) + dot(vec4<output_element_t>(sub_a[word_offset + 4], sub_a[word_offset + 5], sub_a[word_offset + 6], sub_a[word_offset + 7]), b_dequantized_values[1]);\n";
break;
case 2:
shader.MainFunctionBody() << " inter_results[local_id.y][local_id.x] += dot(vec4<output_element_t>(sub_a[word_offset], sub_a[word_offset + 1]), b_dequantized_values[0]) + dot(vec4<output_element_t>(sub_a[word_offset + 2], sub_a[word_offset + 3]), b_dequantized_values[1]);\n";
shader.MainFunctionBody() << " inter_res0 += dot(vec4<output_element_t>(sub_a[word_offset], sub_a[word_offset + 1]), b_dequantized_values[0]) + dot(vec4<output_element_t>(sub_a[word_offset + 2], sub_a[word_offset + 3]), b_dequantized_values[1]);\n";
break;
case 4:
shader.MainFunctionBody() << " inter_results[local_id.y][local_id.x] += dot(sub_a[word_offset], b_dequantized_values[0]) + dot(sub_a[word_offset + 1], b_dequantized_values[1]);\n";
shader.MainFunctionBody() << " inter_res0 += dot(sub_a[word_offset], b_dequantized_values[0]) + dot(sub_a[word_offset + 1], b_dequantized_values[1]);\n";
break;
default:
break;
Expand All @@ -187,11 +193,76 @@ Status MatMulNBitsProgram::GenerateShaderCode(ShaderHelper& shader) const {
}
}
shader.MainFunctionBody() << " word_offset += " << 8 / a.NumComponents() << ";\n"
<< " }\n"
" workgroupBarrier();\n"
<< " }\n";
if (tile_m_ == 1) {
shader.MainFunctionBody() << " b_row = col + local_id.y + 8;\n"
" if (b_row < uniforms.input_b_shape[0] && block < n_blocks_per_col) {\n"
<< " scale = " << scales.GetByOffset("b_row * n_blocks_per_col + block") << ";\n"
<< " b_data = " << b.GetByIndices("input_b_indices_t(b_row, block, 0)") << ";\n"
<< " } else {\n"
" scale = output_element_t(0);\n"
" b_data = input_b_value_t(0);\n"
" }\n"
<< " word_offset = local_id.x * " << block_size_ / a.NumComponents() << ";\n"
<< " for (var i: u32 = 0; i < " << components_b_ << "; i++) {\n";
shader.MainFunctionBody() << " let b_value = b_data";
if (components_b_ > 1) {
shader.MainFunctionBody() << "[i]";
}
shader.MainFunctionBody() << ";\n"
" let b_value_lower = unpack4xU8(b_value & 0x0F0F0F0Fu);\n"
" let b_value_upper = unpack4xU8((b_value >> 4) & 0x0F0F0F0Fu);\n"
" let b_quantized_values = mat2x4<output_element_t>(output_element_t(b_value_lower[0]), output_element_t(b_value_upper[0]), output_element_t(b_value_lower[1]), output_element_t(b_value_upper[1]), output_element_t(b_value_lower[2]), output_element_t(b_value_upper[2]), output_element_t(b_value_lower[3]), output_element_t(b_value_upper[3]));\n"
" let b_dequantized_values = (b_quantized_values - mat2x4<output_element_t>(";
for (int i = 0; i < 8; i++) {
shader.MainFunctionBody() << "zero_point";
if (i < 7) {
shader.MainFunctionBody() << ", ";
}
}
shader.MainFunctionBody() << ")) * scale;\n";
if (tile_m_ == 1) {
switch (a.NumComponents()) {
case 1:
shader.MainFunctionBody() << " inter_res1 += dot(vec4<output_element_t>(sub_a[word_offset], sub_a[word_offset + 1], sub_a[word_offset + 2], sub_a[word_offset + 3]), b_dequantized_values[0]) + dot(vec4<output_element_t>(sub_a[word_offset + 4], sub_a[word_offset + 5], sub_a[word_offset + 6], sub_a[word_offset + 7]), b_dequantized_values[1]);\n";
break;
case 2:
shader.MainFunctionBody() << " inter_res1 += dot(vec4<output_element_t>(sub_a[word_offset], sub_a[word_offset + 1]), b_dequantized_values[0]) + dot(vec4<output_element_t>(sub_a[word_offset + 2], sub_a[word_offset + 3]), b_dequantized_values[1]);\n";
break;
case 4:
shader.MainFunctionBody() << " inter_res1 += dot(sub_a[word_offset], b_dequantized_values[0]) + dot(sub_a[word_offset + 1], b_dequantized_values[1]);\n";
break;
default:
break;
}
} else {
for (uint32_t i = 0; i < tile_m_; i++) {
switch (a.NumComponents()) {
case 1:
shader.MainFunctionBody() << " inter_results[" << i << "][local_id.y][local_id.x] += dot(vec4<output_element_t>(sub_a[" << i << "][word_offset], sub_a[" << i << "][word_offset + 1], sub_a[" << i << "][word_offset + 2], sub_a[" << i << "][word_offset + 3]), b_dequantized_values[0]) + dot(vec4<output_element_t>(sub_a[" << i << "][word_offset + 4], sub_a[" << i << "][word_offset + 5], sub_a[" << i << "][word_offset + 6], sub_a[" << i << "][word_offset + 7]), b_dequantized_values[1]);\n";
break;
case 2:
shader.MainFunctionBody() << " inter_results[" << i << "][local_id.y][local_id.x] += dot(vec4<output_element_t>(sub_a[" << i << "][word_offset], sub_a[" << i << "][word_offset + 1]), b_dequantized_values[0]) + dot(vec4<output_element_t>(sub_a[" << i << "][word_offset + 2], sub_a[" << i << "][word_offset + 3]), b_dequantized_values[1]);\n";
break;
case 4:
shader.MainFunctionBody() << " inter_results[" << i << "][local_id.y][local_id.x] += dot(sub_a[" << i << "][word_offset], b_dequantized_values[0]) + dot(sub_a[" << i << "][word_offset + 1], b_dequantized_values[1]);\n";
break;
default:
break;
}
}
}
shader.MainFunctionBody() << " word_offset += " << 8 / a.NumComponents() << ";\n"
<< " }\n";
}

shader.MainFunctionBody() << " workgroupBarrier();\n"
" }\n";
if (tile_m_ == 1) {
shader.MainFunctionBody() << " if (local_idx < " << WorkgroupSizeY() << ") {\n"
shader.MainFunctionBody() << " inter_results[local_id.y][local_id.x] = inter_res0;\n"
<< " inter_results[local_id.y + 8][local_id.x] = inter_res1;\n"
<< " workgroupBarrier();\n"
<< " if (local_idx < " << WorkgroupSizeY() * 2 << ") {\n"
<< " var output_value = output_value_t(0);\n"
<< " for (var b = 0u; b < " << WorkgroupSizeX() << "; b++) {\n"
<< " output_value += inter_results[local_idx][b];\n"
Expand Down Expand Up @@ -426,12 +497,13 @@ Status MatMulNBits::ComputeInternal(onnxruntime::webgpu::ComputeContext& context
program.CacheHint("T_M" + std::to_string(tile_m));
} else if (block_size == 32) {
components = 1;
constexpr uint32_t workgroup_size = 128;
const uint32_t workgroup_y = N % 8 == 0 ? 8 : N % 4 == 0 ? 4
: 1;
constexpr uint32_t workgroup_size = 64;
constexpr uint32_t workgroup_y = 8;
const uint32_t workgroup_x = workgroup_size / workgroup_y;
program.SetWorkgroupSize(workgroup_x, workgroup_y, 1);
program.SetDispatchGroupSize(data_size / components / workgroup_y);
program.SetDispatchGroupSize((N + 16 - 1) / 16,
(M + tile_m - 1) / tile_m,
batch_count);
program.CacheHint("T_M" + std::to_string(tile_m));
} else {
program.SetDispatchGroupSize(data_size / components / output_number);
Expand Down