Skip to content

vulkan: support sqrt operation #15370

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

Merged
merged 1 commit into from
Aug 17, 2025
Merged

vulkan: support sqrt operation #15370

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
23 changes: 23 additions & 0 deletions ggml/src/ggml-vulkan/ggml-vulkan.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -461,6 +461,7 @@ struct vk_device_struct {
vk_pipeline pipeline_upscale_nearest_f32, pipeline_upscale_bilinear_f32, pipeline_upscale_bilinear_ac_f32;
vk_pipeline pipeline_scale_f32;
vk_pipeline pipeline_sqr_f32;
vk_pipeline pipeline_sqrt_f32;
vk_pipeline pipeline_sin_f32;
vk_pipeline pipeline_cos_f32;
vk_pipeline pipeline_clamp_f32;
Expand Down Expand Up @@ -3029,6 +3030,7 @@ static void ggml_vk_load_shaders(vk_device& device) {
ggml_vk_create_pipeline(device, device->pipeline_scale_f32, "scale_f32", scale_f32_len, scale_f32_data, "main", 2, sizeof(vk_op_unary_push_constants), {512, 1, 1}, {}, 1);

ggml_vk_create_pipeline(device, device->pipeline_sqr_f32, "sqr_f32", sqr_f32_len, sqr_f32_data, "main", 2, sizeof(vk_op_unary_push_constants), {512, 1, 1}, {}, 1);
ggml_vk_create_pipeline(device, device->pipeline_sqrt_f32, "sqrt_f32", sqrt_f32_len, sqrt_f32_data, "main", 2, sizeof(vk_op_unary_push_constants), {512, 1, 1}, {}, 1);
ggml_vk_create_pipeline(device, device->pipeline_sin_f32, "sin_f32", sin_f32_len, sin_f32_data, "main", 2, sizeof(vk_op_unary_push_constants), {512, 1, 1}, {}, 1);
ggml_vk_create_pipeline(device, device->pipeline_cos_f32, "cos_f32", cos_f32_len, cos_f32_data, "main", 2, sizeof(vk_op_unary_push_constants), {512, 1, 1}, {}, 1);

Expand Down Expand Up @@ -6977,6 +6979,11 @@ static vk_pipeline ggml_vk_op_get_pipeline(ggml_backend_vk_context * ctx, const
return ctx->device->pipeline_sqr_f32;
}
return nullptr;
case GGML_OP_SQRT:
if (src0->type == GGML_TYPE_F32 && dst->type == GGML_TYPE_F32) {
return ctx->device->pipeline_sqrt_f32;
}
return nullptr;
case GGML_OP_SIN:
if (src0->type == GGML_TYPE_F32 && dst->type == GGML_TYPE_F32) {
return ctx->device->pipeline_sin_f32;
Expand Down Expand Up @@ -7285,6 +7292,7 @@ static bool ggml_vk_op_supports_incontiguous(ggml_op op) {
case GGML_OP_CONCAT:
case GGML_OP_UPSCALE:
case GGML_OP_SQR:
case GGML_OP_SQRT:
case GGML_OP_SIN:
case GGML_OP_COS:
case GGML_OP_CLAMP:
Expand Down Expand Up @@ -7590,6 +7598,7 @@ static void ggml_vk_op_f32(ggml_backend_vk_context * ctx, vk_context& subctx, co
case GGML_OP_MUL:
case GGML_OP_SCALE:
case GGML_OP_SQR:
case GGML_OP_SQRT:
case GGML_OP_SIN:
case GGML_OP_COS:
case GGML_OP_CLAMP:
Expand Down Expand Up @@ -8237,6 +8246,10 @@ static void ggml_vk_sqr(ggml_backend_vk_context * ctx, vk_context& subctx, const
ggml_vk_op_f32(ctx, subctx, src0, nullptr, nullptr, dst, GGML_OP_SQR, vk_op_unary_push_constants_init(src0, dst), dryrun);
}

static void ggml_vk_sqrt(ggml_backend_vk_context * ctx, vk_context& subctx, const ggml_tensor * src0, ggml_tensor * dst, bool dryrun = false) {
ggml_vk_op_f32(ctx, subctx, src0, nullptr, nullptr, dst, GGML_OP_SQRT, vk_op_unary_push_constants_init(src0, dst), dryrun);
}

static void ggml_vk_sin(ggml_backend_vk_context * ctx, vk_context& subctx, const ggml_tensor * src0, ggml_tensor * dst, bool dryrun = false) {
ggml_vk_op_f32(ctx, subctx, src0, nullptr, nullptr, dst, GGML_OP_SIN, vk_op_unary_push_constants_init(src0, dst), dryrun);
}
Expand Down Expand Up @@ -9700,6 +9713,7 @@ static bool ggml_vk_build_graph(ggml_backend_vk_context * ctx, ggml_cgraph * cgr
case GGML_OP_UPSCALE:
case GGML_OP_SCALE:
case GGML_OP_SQR:
case GGML_OP_SQRT:
case GGML_OP_SIN:
case GGML_OP_COS:
case GGML_OP_CLAMP:
Expand Down Expand Up @@ -9769,6 +9783,7 @@ static bool ggml_vk_build_graph(ggml_backend_vk_context * ctx, ggml_cgraph * cgr
case GGML_OP_UPSCALE:
case GGML_OP_SCALE:
case GGML_OP_SQR:
case GGML_OP_SQRT:
case GGML_OP_SIN:
case GGML_OP_COS:
case GGML_OP_CLAMP:
Expand Down Expand Up @@ -9870,6 +9885,10 @@ static bool ggml_vk_build_graph(ggml_backend_vk_context * ctx, ggml_cgraph * cgr
case GGML_OP_SQR:
ggml_vk_sqr(ctx, compute_ctx, src0, node, dryrun);

break;
case GGML_OP_SQRT:
ggml_vk_sqrt(ctx, compute_ctx, src0, node, dryrun);

break;
case GGML_OP_SIN:
ggml_vk_sin(ctx, compute_ctx, src0, node, dryrun);
Expand Down Expand Up @@ -10121,6 +10140,7 @@ static bool ggml_vk_compute_forward(ggml_backend_vk_context * ctx, ggml_cgraph *
case GGML_OP_UPSCALE:
case GGML_OP_SCALE:
case GGML_OP_SQR:
case GGML_OP_SQRT:
case GGML_OP_SIN:
case GGML_OP_COS:
case GGML_OP_CLAMP:
Expand Down Expand Up @@ -11360,6 +11380,7 @@ static bool ggml_backend_vk_device_supports_op(ggml_backend_dev_t dev, const ggm
case GGML_OP_SILU_BACK:
case GGML_OP_RMS_NORM_BACK:
case GGML_OP_SQR:
case GGML_OP_SQRT:
case GGML_OP_SIN:
case GGML_OP_COS:
case GGML_OP_CLAMP:
Expand Down Expand Up @@ -11803,6 +11824,8 @@ static void ggml_vk_check_results_0(ggml_backend_vk_context * ctx, ggml_cgraph *
tensor_clone = ggml_scale_bias(ggml_ctx, src_clone[0], params[0], params[1]);
} else if (tensor->op == GGML_OP_SQR) {
tensor_clone = ggml_sqr(ggml_ctx, src_clone[0]);
} else if (tensor->op == GGML_OP_SQRT) {
tensor_clone = ggml_sqrt(ggml_ctx, src_clone[0]);
} else if (tensor->op == GGML_OP_SIN) {
tensor_clone = ggml_sin(ggml_ctx, src_clone[0]);
} else if (tensor->op == GGML_OP_COS) {
Expand Down
17 changes: 17 additions & 0 deletions ggml/src/ggml-vulkan/vulkan-shaders/sqrt.comp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,17 @@
#version 450

#include "types.comp"
#include "generic_unary_head.comp"

layout(local_size_x = 512, local_size_y = 1, local_size_z = 1) in;

void main() {
const uint idx = get_idx();

if (idx >= p.ne) {
return;
}

const FLOAT_TYPE val = FLOAT_TYPE(data_a[get_aoffset() + src0_idx(idx)]);
data_d[get_doffset() + dst_idx(idx)] = D_TYPE(sqrt(val));
}
2 changes: 2 additions & 0 deletions ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -566,6 +566,8 @@ void process_shaders() {

string_to_spv("sqr_f32", "square.comp", {{"A_TYPE", "float"}, {"D_TYPE", "float"}, {"FLOAT_TYPE", "float"}});

string_to_spv("sqrt_f32", "sqrt.comp", {{"A_TYPE", "float"}, {"D_TYPE", "float"}, {"FLOAT_TYPE", "float"}});

string_to_spv("sin_f32", "sin.comp", {{"A_TYPE", "float"}, {"D_TYPE", "float"}, {"FLOAT_TYPE", "float"}});

string_to_spv("cos_f32", "cos.comp", {{"A_TYPE", "float"}, {"D_TYPE", "float"}, {"FLOAT_TYPE", "float"}});
Expand Down
Loading