feat: Asynchronous pipeline creation by aleksanderkatan · Pull Request #2604 · software-mansion/TypeGPU

aleksanderkatan · 2026-06-12T12:29:40Z

Changes:

refactor ComputePipelineCore,
add new API for initAsync and initSync,
update some examples to use initAsync,
add tests, update docs.

The PERF is set to true only in browser tests, which currently do not work.
Also, I think that the measurements are just incorrect.
Still, I opted to leave them as is in a separate class. I tested manually the results before and after on Stable Fluids (7 resolves), and they were similar, so I don't think any logic changed.

github-actions · 2026-06-12T12:30:45Z

pkg.pr.new

packages
Ready to be installed by your favorite package manager ⬇️

https://pkg.pr.new/software-mansion/TypeGPU/typegpu@9b0adbbe2e93a34cd9fd7856e7c6e6a9fa79ba97

https://pkg.pr.new/software-mansion/TypeGPU/@typegpu/cli@9b0adbbe2e93a34cd9fd7856e7c6e6a9fa79ba97

https://pkg.pr.new/software-mansion/TypeGPU/@typegpu/noise@9b0adbbe2e93a34cd9fd7856e7c6e6a9fa79ba97

https://pkg.pr.new/software-mansion/TypeGPU/@typegpu/react@9b0adbbe2e93a34cd9fd7856e7c6e6a9fa79ba97

https://pkg.pr.new/software-mansion/TypeGPU/unplugin-typegpu@9b0adbbe2e93a34cd9fd7856e7c6e6a9fa79ba97

benchmark
view benchmark

commit
view commit

github-actions · 2026-06-12T12:31:24Z

Bundle size comparison (`import * as ...` in PR vs `import * as ...` in target):

🟢 Decreased	➖ Unchanged	🔴 Increased	❔ Unknown
0	299	21	0

`import * as ...` in PR vs `import * as ...` in target (did bundle size increase?):

Test	tsdown
STATIC_tgpu.ts	263.15 kB ($${\color{red}+0.6\%}$$)
tgpu_fn.ts	263.15 kB ($${\color{red}+0.6\%}$$)
tgpu_init.ts	263.15 kB ($${\color{red}+0.6\%}$$)
tgpu_lazy.ts	263.15 kB ($${\color{red}+0.6\%}$$)
tgpu_slot.ts	263.15 kB ($${\color{red}+0.6\%}$$)
tgpu_const.ts	263.15 kB ($${\color{red}+0.6\%}$$)
tgpu_unroll.ts	263.15 kB ($${\color{red}+0.6\%}$$)
tgpu_resolve.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_accessor.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_comptime.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_vertexFn.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_computeFn.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_fragmentFn.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_privateVar.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_vertexLayout.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_workgroupVar.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_initFromDevice.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_bindGroupLayout.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_mutableAccessor.ts	263.16 kB ($${\color{red}+0.6\%}$$)
tgpu_resolveWithContext.ts	263.17 kB ($${\color{red}+0.6\%}$$)
STATIC_allImports.ts	293.02 kB ($${\color{red}+0.6\%}$$)

`import { ... }` in PR vs `import * as ...` in PR (is the library tree-Shakeable?):

Click to reveal the results table (297 entries).

Test	tsdown
d_Void.ts	87.36 kB ($${\color{red}+12087.6\%}$$)
d_sampler.ts	87.36 kB ($${\color{red}+11956.6\%}$$)
d_isPtr.ts	87.36 kB ($${\color{red}+11796.0\%}$$)
d_comparisonSampler.ts	87.37 kB ($${\color{red}+11781.8\%}$$)
d_isSizeAttrib.ts	87.37 kB ($${\color{red}+11765.4\%}$$)
d_isWgslArray.ts	87.37 kB ($${\color{red}+11765.3\%}$$)
d_isAlignAttrib.ts	87.37 kB ($${\color{red}+11749.8\%}$$)
d_isWgslStruct.ts	87.37 kB ($${\color{red}+11749.7\%}$$)
d_isAtomic.ts	87.36 kB ($${\color{red}+11749.1\%}$$)
d_isBuiltinAttrib.ts	87.37 kB ($${\color{red}+11718.8\%}$$)
d_isLocationAttrib.ts	87.37 kB ($${\color{red}+11703.3\%}$$)
d_isDecorated.ts	87.37 kB ($${\color{red}+11702.6\%}$$)
d_isInterpolateAttrib.ts	87.38 kB ($${\color{red}+11657.2\%}$$)
d_atomic.ts	87.36 kB ($${\color{red}+11383.8\%}$$)
d_ptrHandle.ts	87.37 kB ($${\color{red}+10412.6\%}$$)
d_ptrUniform.ts	87.37 kB ($${\color{red}+10400.4\%}$$)
d_ptrStorage.ts	87.37 kB ($${\color{red}+10351.3\%}$$)
d_ptrPrivate.ts	87.37 kB ($${\color{red}+10326.9\%}$$)
d_ptrFn.ts	87.36 kB ($${\color{red}+10314.2\%}$$)
d_ptrWorkgroup.ts	87.37 kB ($${\color{red}+10302.9\%}$$)
d_textureExternal.ts	87.37 kB ($${\color{red}+10148.3\%}$$)
d_textureStorage1d.ts	87.37 kB ($${\color{red}+8527.7\%}$$)
d_textureStorage2d.ts	87.37 kB ($${\color{red}+8527.7\%}$$)
d_textureStorage3d.ts	87.37 kB ($${\color{red}+8527.7\%}$$)
d_textureStorage2dArray.ts	87.38 kB ($${\color{red}+8380.9\%}$$)
d_isDisarray.ts	87.37 kB ($${\color{red}+7754.5\%}$$)
d_isUnstruct.ts	87.37 kB ($${\color{red}+7754.5\%}$$)
d_isLooseDecorated.ts	87.37 kB ($${\color{red}+7707.1\%}$$)
d_isLooseData.ts	87.37 kB ($${\color{red}+7443.3\%}$$)
d_isWgslData.ts	87.37 kB ($${\color{red}+6566.4\%}$$)
d_unstruct.ts	87.36 kB ($${\color{red}+5544.2\%}$$)
d_isData.ts	87.36 kB ($${\color{red}+4733.0\%}$$)
d_deepEqual.ts	87.37 kB ($${\color{red}+3884.9\%}$$)
d_struct.ts	87.36 kB ($${\color{red}+2323.7\%}$$)
d_ref.ts	87.36 kB ($${\color{red}+1724.5\%}$$)
std_discard.ts	103.09 kB ($${\color{red}+626.6\%}$$)
std_copy.ts	103.09 kB ($${\color{red}+614.4\%}$$)
std_arrayLength.ts	103.10 kB ($${\color{red}+614.2\%}$$)
std_isBeingTranspiled.ts	103.10 kB ($${\color{red}+612.7\%}$$)
std_getTargetShaderLanguage.ts	103.11 kB ($${\color{red}+609.5\%}$$)
std_extensionEnabled.ts	103.10 kB ($${\color{red}+607.0\%}$$)
std_range.ts	103.09 kB ($${\color{red}+594.2\%}$$)
std_dpdx.ts	103.09 kB ($${\color{red}+585.8\%}$$)
std_fwidthCoarse.ts	103.10 kB ($${\color{red}+585.6\%}$$)
std_dpdxCoarse.ts	103.10 kB ($${\color{red}+585.6\%}$$)
std_dpdyCoarse.ts	103.10 kB ($${\color{red}+585.6\%}$$)
std_fwidthFine.ts	103.10 kB ($${\color{red}+585.6\%}$$)
std_dpdxFine.ts	103.09 kB ($${\color{red}+585.6\%}$$)
std_dpdyFine.ts	103.09 kB ($${\color{red}+585.6\%}$$)
std_fwidth.ts	103.09 kB ($${\color{red}+585.6\%}$$)
std_dpdy.ts	103.09 kB ($${\color{red}+585.6\%}$$)
d_bool.ts	87.36 kB ($${\color{red}+573.3\%}$$)
d_f16.ts	87.36 kB ($${\color{red}+573.3\%}$$)
d_f32.ts	87.36 kB ($${\color{red}+573.3\%}$$)
d_i32.ts	87.36 kB ($${\color{red}+573.3\%}$$)
d_u32.ts	87.36 kB ($${\color{red}+573.3\%}$$)
d_u16.ts	87.36 kB ($${\color{red}+572.1\%}$$)
std_atomicLoad.ts	103.10 kB ($${\color{red}+553.0\%}$$)
std_textureBarrier.ts	103.10 kB ($${\color{red}+552.8\%}$$)
std_atomicStore.ts	103.10 kB ($${\color{red}+552.8\%}$$)
std_workgroupBarrier.ts	103.10 kB ($${\color{red}+552.5\%}$$)
std_storageBarrier.ts	103.10 kB ($${\color{red}+552.5\%}$$)
std_atomicAdd.ts	103.09 kB ($${\color{red}+552.5\%}$$)
std_atomicAnd.ts	103.09 kB ($${\color{red}+552.5\%}$$)
std_atomicMax.ts	103.09 kB ($${\color{red}+552.5\%}$$)
std_atomicMin.ts	103.09 kB ($${\color{red}+552.5\%}$$)
std_atomicSub.ts	103.09 kB ($${\color{red}+552.5\%}$$)
std_atomicXor.ts	103.09 kB ($${\color{red}+552.5\%}$$)
std_atomicOr.ts	103.09 kB ($${\color{red}+552.5\%}$$)
d_textureDepth2d.ts	87.37 kB ($${\color{red}+551.1\%}$$)
d_textureDepthCube.ts	87.37 kB ($${\color{red}+550.9\%}$$)
d_texture1d.ts	87.37 kB ($${\color{red}+550.6\%}$$)
d_texture2d.ts	87.37 kB ($${\color{red}+550.6\%}$$)
d_texture3d.ts	87.37 kB ($${\color{red}+550.6\%}$$)
d_textureDepth2dArray.ts	87.38 kB ($${\color{red}+550.3\%}$$)
d_textureCube.ts	87.37 kB ($${\color{red}+550.3\%}$$)
d_textureDepthCubeArray.ts	87.38 kB ($${\color{red}+550.1\%}$$)
d_textureDepthMultisampled2d.ts	87.38 kB ($${\color{red}+550.0\%}$$)
d_texture2dArray.ts	87.37 kB ($${\color{red}+549.8\%}$$)
d_textureCubeArray.ts	87.37 kB ($${\color{red}+549.5\%}$$)
d_textureMultisampled2d.ts	87.38 kB ($${\color{red}+549.4\%}$$)
d_disarrayOf.ts	87.37 kB ($${\color{red}+481.7\%}$$)
d_vec2b.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec2f.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec2h.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec2i.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec2u.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec3b.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec3f.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec3h.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec3i.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec3u.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec4b.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec4f.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec4h.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec4i.ts	87.36 kB ($${\color{red}+350.2\%}$$)
d_vec4u.ts	87.36 kB ($${\color{red}+350.2\%}$$)
std_subgroupAdd.ts	103.10 kB ($${\color{red}+329.9\%}$$)
std_subgroupBroadcastFirst.ts	103.11 kB ($${\color{red}+329.8\%}$$)
std_subgroupExclusiveAdd.ts	103.11 kB ($${\color{red}+329.8\%}$$)
std_subgroupExclusiveMul.ts	103.11 kB ($${\color{red}+329.8\%}$$)
std_subgroupInclusiveAdd.ts	103.11 kB ($${\color{red}+329.8\%}$$)
std_subgroupInclusiveMul.ts	103.11 kB ($${\color{red}+329.8\%}$$)
std_subgroupShuffleDown.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupShuffleXor.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupBroadcast.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupShuffleUp.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupShuffle.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupBallot.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupElect.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupAll.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupAnd.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupAny.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupMax.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupMin.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupMul.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupXor.ts	103.10 kB ($${\color{red}+329.8\%}$$)
std_subgroupOr.ts	103.10 kB ($${\color{red}+329.8\%}$$)
d_formatToWGSLType.ts	87.37 kB ($${\color{red}+318.1\%}$$)
d_uint8.ts	87.36 kB ($${\color{red}+318.0\%}$$)
d_unorm10_10_10_2.ts	87.37 kB ($${\color{red}+317.9\%}$$)
d_unorm8x4_bgra.ts	87.37 kB ($${\color{red}+317.9\%}$$)
d_float16x2.ts	87.37 kB ($${\color{red}+317.9\%}$$)
d_float16x4.ts	87.37 kB ($${\color{red}+317.9\%}$$)
d_float32x2.ts	87.37 kB ($${\color{red}+317.9\%}$$)
d_float32x3.ts	87.37 kB ($${\color{red}+317.9\%}$$)
d_float32x4.ts	87.37 kB ($${\color{red}+317.9\%}$$)
d_snorm16x2.ts	87.37 kB ($${\color{red}+317.9\%}$$)
d_snorm16x4.ts	87.37 kB ($${\color{red}+317.9\%}$$)
d_unorm16x2.ts	87.37 kB ($${\color{red}+317.9\%}$$)
d_unorm16x4.ts	87.37 kB ($${\color{red}+317.9\%}$$)
d_sint16x2.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_sint16x4.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_sint32x2.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_sint32x3.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_sint32x4.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_snorm8x2.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_snorm8x4.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_uint16x2.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_uint16x4.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_uint32x2.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_uint32x3.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_uint32x4.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_unorm8x2.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_unorm8x4.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_float16.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_float32.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_sint8x2.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_sint8x4.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_snorm16.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_uint8x2.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_uint8x4.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_unorm16.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_sint16.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_sint32.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_snorm8.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_uint16.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_uint32.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_unorm8.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_sint8.ts	87.36 kB ($${\color{red}+317.9\%}$$)
d_packedFormats.ts	87.37 kB ($${\color{red}+317.5\%}$$)
d_isPackedData.ts	87.37 kB ($${\color{red}+316.7\%}$$)
std_textureDimensions.ts	103.10 kB ($${\color{red}+300.6\%}$$)
std_textureGather.ts	103.10 kB ($${\color{red}+300.6\%}$$)
std_textureSampleBaseClampToEdge.ts	103.11 kB ($${\color{red}+300.5\%}$$)
std_textureSampleCompareLevel.ts	103.11 kB ($${\color{red}+300.5\%}$$)
std_textureSampleCompare.ts	103.11 kB ($${\color{red}+300.5\%}$$)
std_textureSampleLevel.ts	103.10 kB ($${\color{red}+300.5\%}$$)
std_textureSampleBias.ts	103.10 kB ($${\color{red}+300.5\%}$$)
std_textureSampleGrad.ts	103.10 kB ($${\color{red}+300.5\%}$$)
std_textureSample.ts	103.10 kB ($${\color{red}+300.5\%}$$)
std_textureStore.ts	103.10 kB ($${\color{red}+300.5\%}$$)
std_textureLoad.ts	103.10 kB ($${\color{red}+300.5\%}$$)
d_alignmentOf.ts	87.37 kB ($${\color{red}+299.8\%}$$)
std_translation4.ts	103.10 kB ($${\color{red}+282.0\%}$$)
std_rotationX4.ts	103.10 kB ($${\color{red}+282.0\%}$$)
std_rotationY4.ts	103.10 kB ($${\color{red}+282.0\%}$$)
std_rotationZ4.ts	103.10 kB ($${\color{red}+282.0\%}$$)
std_identity2.ts	103.09 kB ($${\color{red}+282.0\%}$$)
std_identity3.ts	103.09 kB ($${\color{red}+282.0\%}$$)
std_identity4.ts	103.09 kB ($${\color{red}+282.0\%}$$)
std_scaling4.ts	103.09 kB ($${\color{red}+282.0\%}$$)
d_isBuiltin.ts	87.37 kB ($${\color{red}+255.5\%}$$)
d_sizeOf.ts	87.36 kB ($${\color{red}+254.8\%}$$)
d_isContiguous.ts	87.37 kB ($${\color{red}+254.7\%}$$)
d_getLongestContiguousPrefix.ts	87.38 kB ($${\color{red}+254.6\%}$$)
d_arrayOf.ts	87.36 kB ($${\color{red}+233.8\%}$$)
d_size.ts	87.36 kB ($${\color{red}+232.0\%}$$)
d_align.ts	87.36 kB ($${\color{red}+232.0\%}$$)
d_location.ts	87.36 kB ($${\color{red}+232.0\%}$$)
d_interpolate.ts	87.37 kB ($${\color{red}+232.0\%}$$)
d_invariant.ts	87.37 kB ($${\color{red}+227.4\%}$$)
d_mat4x4f.ts	87.36 kB ($${\color{red}+223.7\%}$$)
d_mat2x2f.ts	87.36 kB ($${\color{red}+223.7\%}$$)
d_mat3x3f.ts	87.36 kB ($${\color{red}+223.7\%}$$)
d_matToArray.ts	87.37 kB ($${\color{red}+222.1\%}$$)
d_builtin.ts	87.36 kB ($${\color{red}+205.1\%}$$)
std_unpack2x16float.ts	103.10 kB ($${\color{red}+186.8\%}$$)
std_unpack4x8unorm.ts	103.10 kB ($${\color{red}+186.8\%}$$)
std_pack2x16float.ts	103.10 kB ($${\color{red}+186.8\%}$$)
std_pack4x8unorm.ts	103.10 kB ($${\color{red}+186.8\%}$$)
std_bitcastU32toF32.ts	103.10 kB ($${\color{red}+133.3\%}$$)
std_bitcastU32toI32.ts	103.10 kB ($${\color{red}+133.3\%}$$)
std_neg.ts	103.09 kB ($${\color{red}+113.2\%}$$)
std_bitShiftRight.ts	103.10 kB ($${\color{red}+113.2\%}$$)
std_bitShiftLeft.ts	103.10 kB ($${\color{red}+113.2\%}$$)
std_add.ts	103.09 kB ($${\color{red}+113.1\%}$$)
std_div.ts	103.09 kB ($${\color{red}+113.1\%}$$)
std_mod.ts	103.09 kB ($${\color{red}+113.1\%}$$)
std_mul.ts	103.09 kB ($${\color{red}+113.1\%}$$)
std_sub.ts	103.09 kB ($${\color{red}+113.1\%}$$)
std_translate4.ts	103.10 kB ($${\color{red}+109.9\%}$$)
std_rotateX4.ts	103.09 kB ($${\color{red}+109.9\%}$$)
std_scale4.ts	103.09 kB ($${\color{red}+109.9\%}$$)
std_rotateY4.ts	103.09 kB ($${\color{red}+109.8\%}$$)
std_rotateZ4.ts	103.09 kB ($${\color{red}+109.8\%}$$)
d_memoryLayoutOf.ts	87.37 kB ($${\color{red}+108.3\%}$$)
std_and.ts	103.09 kB ($${\color{red}+99.3\%}$$)
std_or.ts	103.09 kB ($${\color{red}+99.3\%}$$)
std_select.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_allEq.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_all.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_not.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_eq.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_lt.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_ne.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_isCloseTo.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_any.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_ge.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_gt.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_le.ts	103.09 kB ($${\color{red}+99.2\%}$$)
std_dot4U8Packed.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_length.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_countTrailingZeros.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_distance.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_dot.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_countLeadingZeros.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_firstTrailingBit.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_firstLeadingBit.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_fract.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_frexp.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_modf.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_quantizeToF16.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_mix.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_countOneBits.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_dot4I8Packed.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_determinant.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_extractBits.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_faceForward.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_inverseSqrt.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_reverseBits.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_insertBits.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_smoothstep.ts	103.10 kB ($${\color{red}+56.2\%}$$)
std_normalize.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_transpose.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_saturate.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_degrees.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_radians.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_reflect.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_refract.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_acosh.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_asinh.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_atan2.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_atanh.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_clamp.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_cross.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_floor.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_ldexp.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_round.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_trunc.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_acos.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_asin.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_atan.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_ceil.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_cosh.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_exp2.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_log2.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_sign.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_sinh.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_sqrt.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_step.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_tanh.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_abs.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_cos.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_exp.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_fma.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_log.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_max.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_min.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_pow.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_sin.ts	103.09 kB ($${\color{red}+56.2\%}$$)
std_tan.ts	103.09 kB ($${\color{red}+56.2\%}$$)
common_fullScreenTriangle.ts	56.66 kB ($${\color{red}+42.1\%}$$)
common_writeSoA.ts	56.65 kB ($${\color{red}+37.5\%}$$)
STATIC_allImports.ts	288.00 kB ($${\color{green}-1.7\%}$$)
STATIC_d.ts	87.36 kB ($${\color{green}-2.6\%}$$)
STATIC_std.ts	103.08 kB ($${\color{green}-2.7\%}$$)

If you wish to run a comparison for other, slower bundlers, run the 'Tree-shake test' from the GitHub Actions menu.

github-actions · 2026-06-12T12:32:32Z

Resolution Time Benchmark

---
config:
  themeVariables:
    xyChart:
      plotColorPalette: "#E63946, #3B82F6, #059669"
---
xychart
  title "Random Branching (🔴 PR | 🔵 main | 🟢 release)"
  x-axis "max depth" [1, 2, 3, 4, 5, 6, 7, 8]
  y-axis "time (ms)"
  line [0.95, 2.04, 4.81, 6.83, 8.07, 11.46, 23.36, 23.34]
  line [1.02, 2.03, 4.23, 6.46, 8.39, 11.59, 23.16, 25.90]
  line [0.91, 1.86, 4.31, 6.60, 7.47, 10.76, 22.08, 24.70]

---
config:
  themeVariables:
    xyChart:
      plotColorPalette: "#E63946, #3B82F6, #059669"
---
xychart
  title "Linear Recursion (🔴 PR | 🔵 main | 🟢 release)"
  x-axis "max depth" [1, 2, 3, 4, 5, 6, 7, 8]
  y-axis "time (ms)"
  line [0.34, 0.52, 0.72, 0.83, 1.16, 1.27, 1.52, 1.63]
  line [0.28, 0.54, 0.73, 0.91, 1.22, 1.31, 1.57, 1.69]
  line [0.27, 0.52, 0.70, 0.82, 1.18, 1.21, 1.44, 1.53]

---
config:
  themeVariables:
    xyChart:
      plotColorPalette: "#E63946, #3B82F6, #059669"
---
xychart
  title "Full Tree (🔴 PR | 🔵 main | 🟢 release)"
  x-axis "max depth" [1, 2, 3, 4, 5, 6, 7, 8]
  y-axis "time (ms)"
  line [0.90, 2.01, 4.68, 6.38, 12.65, 26.64, 57.38, 112.97]
  line [1.01, 2.01, 4.20, 6.49, 12.54, 26.82, 58.05, 118.37]
  line [0.84, 2.09, 4.22, 6.35, 12.65, 25.27, 54.47, 113.92]

Copilot

Pull request overview

This PR introduces explicit pipeline initialization APIs for compute pipelines, enabling asynchronous compilation to reduce first-dispatch stalls, and updates docs/examples/tests to exercise the new behavior.

Changes:

Refactors ComputePipelineCore and adds initAsync() / initSync() to compute pipelines.
Updates multiple docs examples to eagerly initialize compute pipelines (primarily via initAsync()).
Adds unit tests and updates docs/example snapshots to reflect the new initialization flow.

Reviewed changes

Copilot reviewed 13 out of 13 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
packages/typegpu/src/core/pipeline/computePipeline.ts	Adds `initAsync`/`initSync`, refactors memoization/initialization flow, and moves perf tracking into a helper.
packages/typegpu/tests/pipelineInit.test.ts	Adds tests covering sync/async initialization, caching, and unwrap behavior during async init.
packages/typegpu-testing-utility/src/extendedIt.ts	Extends the GPU device mock with `createComputePipelineAsync` for new tests.
apps/typegpu-docs/src/examples/simulation/stable-fluid/index.ts	Eagerly initializes all compute pipelines with `initAsync()` to avoid first-frame stalls.
apps/typegpu-docs/src/examples/simulation/gravity/index.ts	Eagerly initializes compute pipelines with `initAsync()`.
apps/typegpu-docs/src/examples/rendering/function-visualizer/index.ts	Eagerly initializes dynamically-created guarded compute pipelines with `initAsync()`.
apps/typegpu-docs/src/examples/image-processing/background-segmentation/index.ts	Eagerly initializes guarded compute pipelines via `initSync()` before use.
apps/typegpu-docs/src/examples/algorithms/mnist-inference/index.ts	Eagerly initializes compute pipelines via `initAsync()` (including optional subgroup pipeline).
apps/typegpu-docs/src/examples/algorithms/matrix-next/index.ts	Eagerly initializes both compute strategies via `initAsync()`.
apps/typegpu-docs/tests/individual-example-tests/stable-fluid.test.ts	Updates expected shader-module call count and snapshot output.
apps/typegpu-docs/tests/individual-example-tests/mnist-inference.test.ts	Updates shader snapshot output ordering/content.
apps/typegpu-docs/tests/individual-example-tests/matrix-next.test.ts	Updates expected shader-module call count and snapshot output.
apps/typegpu-docs/src/content/docs/apis/pipelines.mdx	Documents pipeline initialization and shows `initSync`/`initAsync` usage.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

iwoplaza · 2026-06-15T15:58:21Z

+Pipeline initialization involves resolving the pipeline code, creating the shader module, and creating the underlying WebGPU pipeline.
+This happens automatically the first time the pipeline is executed via `.draw`, `.dispatchWorkgroups`, or a similar method.
+The `initSync` and `initAsync` methods let you perform this work ahead of time, avoiding a stall on first execution.


I'm missing the explanation of when one should use initSync as opposed to initAsync and vice versa. On the other hand, it's non-trivial to explain it briefly 🤔

Maybe:

Suggested change

Pipeline initialization involves resolving the pipeline code, creating the shader module, and creating the underlying WebGPU pipeline.

This happens automatically the first time the pipeline is executed via `.draw`, `.dispatchWorkgroups`, or a similar method.

The `initSync` and `initAsync` methods let you perform this work ahead of time, avoiding a stall on first execution.

Pipeline initialization involves resolving the pipeline code, creating the shader module, and creating the underlying WebGPU pipeline.

This happens automatically the first time the pipeline is executed via `.draw`, `.dispatchWorkgroups`, or a similar method.

The `initSync` function lets you jump-start that initialization on the WebGPU process early. To await until the initialization actually finishes and avoid a stall on first execution, you can use the `initAsync` function instead.

I updated this paragraph and added commented snippets, tell me what you think

iwoplaza · 2026-06-15T16:04:04Z

    },
  });

+await Promise.all([computeCollisionsPipeline.initAsync(), computeGravityPipeline.initAsync()]);


I'd prewarm these promises early:

Suggested change

await Promise.all([computeCollisionsPipeline.initAsync(), computeGravityPipeline.initAsync()]);

// Initializing all used pipelines

const pipelinePromises = [

computeCollisionsPipeline.initAsync(),

computeGravityPipeline.initAsync(),

skyBoxPipeline.initAsync(),

renderPipeline.initAsync(),

];

And await them right before requesting the first frame:

// Only requesting the first frame after all pipelines have been initialized await Promise.all(pipelinePromises); requestAnimationFrame(frame);

On another note, why aren't we preinitializing the render pipelines?

we aren't because they do not yethave initAsync/initSync methods - as I mentioned, I wanted to wait with them for an initial review so as not to do the same fixes twice

reczkok

Looks nice, I like the performance tracking approach

reczkok · 2026-06-16T11:10:03Z

+## Initialization
+
+Pipeline initialization involves resolving the pipeline code, creating the shader module, and creating the underlying WebGPU pipeline.
+This happens automatically the first time the pipeline is executed via `.draw`, `.dispatchWorkgroups`, or a similar method.
+The `initSync` method lets you start the initialization early.
+To wait until initialization actually finishes on the device (fully avoiding a stall on first execution), use `initAsync` instead.


I would stress here (and especially in the AI Skill) that this is not that important for small/mid sized shaders - only makes sense when you have a bigger application or shaders that are very big or have very many of them) - AI is prone to overusing solutions like those.

Wouldn't it be better to just not mention it in the skill at all? Can we discuss it in #2650 ?

…eline-creation

aleksanderkatan · 2026-06-25T09:03:50Z

@pullfrog review

pullfrog

ℹ️ No critical issues — one docstring typo inline.

Reviewed changes — added explicit initAsync/initSync pipeline initialization for compute and render pipelines, extracted a PerformanceTracker, updated docs/examples, and realigned test snapshots and expected call counts.

Added initAsync() and initSync() to TgpuComputePipeline, TgpuRenderPipeline, and TgpuGuardedComputePipeline.
Refactored ComputePipelineCore and RenderPipelineCore to share async-init logic through resolveAndCreateShaderModule() and PerformanceTracker.
Updated the pipelines docs page and example code to pre-initialize pipelines.
Added packages/typegpu/tests/pipelineInit.test.ts and updated example test snapshots/call counts.

^{｜ Fix all ➔ ｜ Fix 👍s ➔ ｜ View workflow run ｜ Using Kimi K2 (free via Pullfrog for OSS) ｜ 𝕏}

aleksanderkatan added 5 commits June 12, 2026 13:06

Implement initAsync

43b4724

Add tests

e5168a6

Abstract PerformanceCollector from unwrap

11607d8

Abstract resolveAndCreateShaderModule

c352876

Clean up initAsync

8852e65

aleksanderkatan linked an issue Jun 12, 2026 that may be closed by this pull request

feat: Add support for asynchronous pipeline creation #2501

Open

aleksanderkatan added 8 commits June 12, 2026 14:53

Add jsDocs

fcbed03

Extract unwrap to initSync

fa5c706

Add initSync to the interface

3f4379d

Update examples

674c6b2

Update docs

32c9d16

Update examples

e3ee140

Rename PerformanceCollector to PerformanceTracker

af270ae

Use true private props

de777f1

aleksanderkatan force-pushed the feat/asynchronous-pipeline-creation branch from 9fe9b72 to de777f1 Compare June 12, 2026 15:49

aleksanderkatan marked this pull request as ready for review June 12, 2026 15:50

Copilot AI review requested due to automatic review settings June 12, 2026 15:50

Copilot started reviewing on behalf of aleksanderkatan June 12, 2026 15:50 View session

aleksanderkatan requested review from cieplypolar, iwoplaza and reczkok June 12, 2026 15:50

Copilot AI reviewed Jun 12, 2026

View reviewed changes

Comment thread packages/typegpu/src/core/pipeline/computePipeline.ts

iwoplaza reviewed Jun 15, 2026

View reviewed changes

aleksanderkatan changed the title ~~feat: Asynchronous pipeline creation~~ [WIP] feat: Asynchronous pipeline creation Jun 16, 2026

aleksanderkatan added 3 commits June 16, 2026 11:56

Add initAsync and initSync directly onto guarded compute pipeline

a5f7a10

Update docs

5dc41cb

Await promises in examples as late as possible

f090a7a

Move cleanup to finally

0aa4ef3

reczkok reviewed Jun 16, 2026

View reviewed changes

aleksanderkatan added 14 commits June 24, 2026 17:07

Update docs

8d56157

Update interfaces

22cffdd

Add guarded compute pipeline tests

9f6bee0

Add render pipeline tests

d6ca5ec

Move performance tracker to another file

f8b309d

Use performance tracker in render pipeline

51ec842

Separate unwrap logic to resolveAndCreateShaderModule

abda3fa

Add initSync and initAsync implementations

aa2ad94

Add render pipeline initialization to examples

507064d

Update expected calls and snapshots

2e3293a

Merge remote-tracking branch 'origin/main' into feat/asynchronous-pip…

1caa389

…eline-creation

Review fixes

aca8bfb

nr fix

858e21a

Fix lines combinations expected calls

4efe6ad

pullfrog Bot reviewed Jun 25, 2026

View reviewed changes

Comment thread packages/typegpu/src/core/pipeline/renderPipeline.ts Outdated

Fix js doc

9b0adbb

aleksanderkatan requested review from iwoplaza and reczkok June 25, 2026 09:53

aleksanderkatan changed the title ~~[WIP] feat: Asynchronous pipeline creation~~ feat: Asynchronous pipeline creation Jun 25, 2026

-await Promise.all([computeCollisionsPipeline.initAsync(), computeGravityPipeline.initAsync()]);
+// Initializing all used pipelines
+const pipelinePromises = [
+  computeCollisionsPipeline.initAsync(),
+  computeGravityPipeline.initAsync(),
+  skyBoxPipeline.initAsync(),
+  renderPipeline.initAsync(),
+];

Uh oh!

Uh oh!

Conversation

aleksanderkatan commented Jun 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot commented Jun 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot commented Jun 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Bundle size comparison (import * as ... in PR vs import * as ... in target):

import * as ... in PR vs import * as ... in target (did bundle size increase?):

import { ... } in PR vs import * as ... in PR (is the library tree-Shakeable?):

Uh oh!

github-actions Bot commented Jun 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Resolution Time Benchmark

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

iwoplaza Jun 15, 2026

Choose a reason for hiding this comment

Uh oh!

aleksanderkatan Jun 16, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

iwoplaza Jun 15, 2026

Choose a reason for hiding this comment

Uh oh!

aleksanderkatan Jun 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

reczkok left a comment

Choose a reason for hiding this comment

Uh oh!

reczkok Jun 16, 2026

Choose a reason for hiding this comment

Uh oh!

aleksanderkatan Jun 25, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

aleksanderkatan commented Jun 25, 2026

Uh oh!

pullfrog Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

aleksanderkatan commented Jun 12, 2026 •

edited

Loading

github-actions Bot commented Jun 12, 2026 •

edited

Loading

github-actions Bot commented Jun 12, 2026 •

edited

Loading

Bundle size comparison (`import * as ...` in PR vs `import * as ...` in target):

`import * as ...` in PR vs `import * as ...` in target (did bundle size increase?):

`import { ... }` in PR vs `import * as ...` in PR (is the library tree-Shakeable?):

github-actions Bot commented Jun 12, 2026 •

edited

Loading

aleksanderkatan Jun 16, 2026 •

edited

Loading