DirectX 12 Agility SDK Downloads

Summary of features and other fixes in 1.618.*:

• Runtime regression fix: Creating resources in state D3D12_RESOURCE_STATE_RAYTRACING_ACCELERATION_STRUCTURE with custom heap properties validated L0 and L1 in backwards order. Apps hitting this incorrect stricter validation even if they didn’t use new AgilitySDK just by using PIX which uses latest SDK, hence it looks like PIX regressed. At least one developer also hit this bug directly.
• Debug layer validation regression fix (no runtime impact): New compute and work graph dispatch grid size validation incorrectly validated against the mesh shader’s 2^22 limit everywhere. Compute dispatch has no spec limit, and the limit for work graphs is 2^24-1 groups.
• Fixes an issue where, only with the debug layer enabled, recording ProcessFrames command in ID3D12VideoCommandList fails in the Close() method with E_INVALIDARG

• Correctly enable SODB capture for advanced shader delivery
• Add ID3D12StateObjectDatabase for programmatic SODB generation
• Add D3D12StateObjectCompiler ‘merge’ command

• Advanced Shader Delivery
• Application Specific Driver States
• Video Encode features
  • D3D12_Video_Encoding_HEVC_ReferenceListExtension
  • D3D12_Video_Encoding_LowerResolution_2Pass
  • PSNR stats added to: D3D12_Video_Encoding_Stats_Metadata
  • D3D12 Video encode subregion
  • D3D12 Video encode output stats (per block QP, per block SATD, per block bit allocation)
  • D3D12 Video encode GPU texture input QP map
  • D3D12 Video encode GPU texture/CPU buffer Dirty maps/rects
  • D3D12 Video encode GPU texture/CPU buffer motion vector hints
  • D3D12 Video Encode Spatial Adaptive Rate Control Quantization
• Recreate at GPUVA
• Tight Alignment of Resources

• Everything in 1.616.0 plus:
• Shader Model 6.9 preview, adding:
  • Cooperative Vectors
  • Shader Execution Reordering
  • Support for RayQuery using Opacity Micromaps
    • Normal TraceRay doesn’t require SM 6.9, so the 1.616.0 release works for it
• D3D Video Encoding updates

See the blog post for more information about the release.

.1 point release adds:

• Debug Layer fixes:
  • Enhanced Barrier placed resource validation fix for false errors
    • For D3D12_BARRIER_SUBRESOURCE_RANGE, the Debug Layer was not handling the NumMipLevels=0 special case (treat IndexOrFirstMipLevel as a subresource index)
    • Debug Layer now understands that RT/DS initialization can be accomplished by RenderPass Clear/Discard
  • GPU based validation (GBV):
    • Fix for incorrect texture state/layout validation in cases where GBV needed to insert LOD calculation

• Runtime regression fix: Creating resources in state D3D12_RESOURCE_STATE_RAYTRACING_ACCELERATION_STRUCTURE with custom heap properties validated L0 and L1 in backwards order. Apps hitting this incorrect stricter validation even if they didn’t use new AgilitySDK just by using PIX which uses latest SDK, hence it looks like PIX regressed. At least one developer also hit this bug directly.
• Debug layer validation regression fix (no runtime impact): New compute and work graph dispatch grid size validation incorrectly validated against the mesh shader’s 2^22 limit everywhere. Compute dispatch has no spec limit, and the limit for work graphs is 2^24-1 groups.
• Fixes an issue where, only with the debug layer enabled, recording ProcessFrames command in ID3D12VideoCommandList fails in the Close() method with E_INVALIDARG

• Correctly enable SODB capture for advanced shader delivery
• Add ID3D12StateObjectDatabase for programmatic SODB generation
• Add D3D12StateObjectCompiler ‘merge’ command

• Advanced Shader Delivery
• Application Specific Driver States
• Video Encode features
  • D3D12_Video_Encoding_HEVC_ReferenceListExtension
  • D3D12_Video_Encoding_LowerResolution_2Pass
  • PSNR stats added to: D3D12_Video_Encoding_Stats_Metadata
  • D3D12 Video encode subregion
  • D3D12 Video encode output stats (per block QP, per block SATD, per block bit allocation)
  • D3D12 Video encode GPU texture input QP map
  • D3D12 Video encode GPU texture/CPU buffer Dirty maps/rects
  • D3D12 Video encode GPU texture/CPU buffer motion vector hints
  • D3D12 Video Encode Spatial Adaptive Rate Control Quantization
• Recreate at GPUVA
• Tight Alignment of Resources

• Opacity Micromap support
  • RayQuery+OMM support needs Shader Model 6.9, in preview status (AgilitySDK 1.717.0) as of this release
• Tier 4 Tiled Resource

See the blog post for more information about the release.

.1 point release adds:

• Debug Layer fixes:
  • Enhanced Barrier placed resource validation fix for false errors
    • For D3D12_BARRIER_SUBRESOURCE_RANGE, the Debug Layer was not handling the NumMipLevels=0 special case (treat IndexOrFirstMipLevel as a subresource index)
    • Debug Layer now understands that RT/DS initialization can be accomplished by RenderPass Clear/Discard
  • GPU based validation (GBV):
    • Fix for incorrect texture state/layout validation in cases where GBV needed to insert LOD calculation
• Runtime fix: D3D12EnableExperimentalFeatures() was allowing D3D12ExperimentalShaderModels
  • This made it look like Shader Model 6.9 could work even though it’s only supported in the 1.717.x preview
  • Now asking for D3D12ExperimentalShaderModels returns E_NOINTERFACE.

• Opacity Micromap support
  • RayQuery+OMM support needs Shader Model 6.9, in preview status (AgilitySDK 1.717.0) as of this release
• Tier 4 Tiled Resource

See the blog post for more information about the release.

• Fix placed resource initialization validation bug related to plane validation
• Fix a bug where GBV patching will crash when initializing if the caller didn’t request a debug info parser

See the blog post for more info about the release.

• Shader hash bypass officially supports applications opting to bypass shader hash validation

• Enabled R9G9B9E5_SHAREDEXP format for Render Target and Unordered Access Views (device support still needs to be queried)
• Minor state object validation fixes related to dxil defined subobjects and collections

See here for more info about the release.

• Work Graphs
• Shader Model 6.8
  • Work Graphs support
  • Start Vertex/Instance Location
  • Wave Size Range
  • Expanded Comparison Sampling
• GPU Upload Heaps (requires future OS)
• Incrementing Constant in ExecuteIndirect
• Generic Programs in State objects

• Minor updates for .3:
  • Honor root signature associations added to export in collections when the export is imported into generic program in an executable state object.
  • Propagate root signatures to exports that have bindings and no root signatures within the scope of a generic program, like what happens with PSOs
  • Other generic programs related bug fixes, additional validation and optimizations.
  • For state objects, minor fixes to subobject association logic in edge cases involving collections to more precisely match spec wording.

See here for more info about the main release, 1.613.

• Work Graphs
• Shader Model 6.8
  • Work Graphs support
  • Start Vertex/Instance Location
  • Wave Size Range
  • Expanded Comparison Sampling
• GPU Upload Heaps (requires future OS)
• Incrementing Constant in ExecuteIndirect
• Generic Programs in State objects

• Minor updates for .2:
  • Close validation hole: Prevent use of features in incompatible state object types, such as defining generic program components (like rast state) in a raytracing pipeline.
  • Debug layer fixes around its lifetime management of program identifiers (e.g. the handles for the new generic programs)
  • ExecuteIndirect validation was too strict: buffer size only needs to be big enough for the number of commands * stride, minus the unused space at end of last command

See here for more info about the main release, 1.613.

• Work Graphs
• Shader Model 6.8
  • Work Graphs support
  • Start Vertex/Instance Location
  • Wave Size Range
  • Expanded Comparison Sampling
• GPU Upload Heaps (requires future OS)
• Incrementing Constant in ExecuteIndirect
• Generic Programs in State objects

• Minor updates for .1:
  • In d3d12.h: missing const on input pointer members: D3D12_NODE_CPU_INPUT.pRecords and D3D12_MULTI_NODE_CPU_INPUT.pNodeInputs
  • Debug layer validation fixes for D3D12_DISPATCH_MODE_MULTI_NODE_CPU_INPUT
  • Some missing validation for generic programs (to line up with existing PSO path)
  • Allow generic programs or work graphs only in executable state objects, not collections

See here for more info about the main release, 1.613.

• Work Graphs
• Shader Model 6.8
  • Work Graphs support
  • Start Vertex/Instance Location
  • Wave Size Range
  • Expanded Comparison Sampling
• GPU Upload Heaps (requires future OS)
• Incrementing Constant in ExecuteIndirect
• Generic Programs in State objects

See here for more info about this release

• Fix crash in D3D12SDKLayers creating resource with initial enhanced layout when legacy state validation forced on.
• Fix std::bad_optional_access exception in D3D12SDKLayers and fixup partially promoted state.

• Added support for AV1 Encode
• Make pContext mutable in ID3D12InfoQueue1::RegisterMessageCallback.
• Legacy transitions into/out-of UAV state includes: D3D12_BARRIER_SYNC_BUILD_RAYTRACING_ACCELERATION_STRUCTURE D3D12_BARRIER_SYNC_EMIT_RAYTRACING_ACCELERATION_STRUCTURE_POSTBUILD_INFO D3D12_BARRIER_SYNC_COPY_RAYTRACING_ACCELERATION_STRUCTURE This only applies to buffer barriers.
• Fix enhanced barrier SYNC_DRAW being incompatible with ACCESS_RENDER_TARGET
• Improve enhanced barrier validation for raytracing acceleration data structures

• Fix some aliasing barrier regressions on drivers supporting enhanced barriers.
  • Full cache flush supporting aliasing barrier on stencil resources
  • Use a global barrier rather than a buffer or texture barrier when translating non-null aliasing barriers. The global barrier guarantees a cache flush, which is needed when aliased resources share memory but do not share the same VA space.

• Fixed state corruption causing incorrect programmable sample positions validation during ExecuteCommandList
• Address incorrect Enhanced Barriers simultaneous-access validation
• Fix incorrect Enhanced Barriers sync/access validation
• Add missing string conversion for D3D_BARRIER_SYNC_CLEAR_UNORDERED_ACCESS_VIEW

• Header build fix
• Fixes bug preventing enhanced barriers enablement

New minor features:

• Independent Front/Back Stencil Refs and Masks
• Triangle fan
• Dynamic Pipeline State: Depth Bias, IB Strip Cut

See here for more information

See here for more information

• Fixes bug in GBV to skip bounds checks on static descriptors already bounds-checked on the CPU timeline
• Fix false Object Deleted While Still in Use errors from debug layer
• Fix deadlock in debug layer while validating placed resource virtual addresses
• Fix crash in ResolveQueryData due to race condition

New features:

• Relaxed buffer/texture copy alignment
• Support for copying between different dimensions of textures
• Delayed input layout and vertex buffer alignment validation
• Negative height viewports flip y-axis intepretation
• Alpha/InvAlpha blend factors

Bugfixes

• Fixes a crash using GBV with shader patch mode TRACKING_ONLY. 
• Fixes false debug validation output resulting from depth slice state being confused with depth slice state. 
• Fixes a bug causing promoted COPY_DEST to not decay back to COMMON. 
• Report live objects when encountering device removed from a kernel memory failure.  

Shader Model 6.6

• Everything in 1.616.0 plus:
• Shader Model 6.9 preview, adding:
  • Cooperative Vectors
  • Shader Execution Reordering
  • Support for RayQuery using Opacity Micromaps
    • Normal TraceRay doesn’t require SM 6.9, so the 1.616.0 release works for it
• D3D Video Encoding updates

See the blog post for more information about the release.

• Fix meta command validation
• Add Raw UAV/SRV validation for tight aligned buffers
• Fix a bug where AlignmentRestriction hint wasn’t being passed to drivers
• Fix placed resource initialization validation bug related to plane validation
• Fix a bug where GBV patching will crash when initializing if the caller didn’t request a debug info parser

See the blog post for more info about the release.

• See blog for details: AgilitySDK 1.716.0-preview and 1.615-retail – DirectX Developer Blog

• Mesh nodes in work graphs. See the blog for more info.
• Continues to include DirectSR from the recent 1.714.0 preview.

See here for more info about this preview release.

Wave Matrix

AV1 Encode (officially released in 1.611.1)

See here for more info about this preview release

Non-normalized Sampling

See here for more info about this preview release

Enhanced Barriers Preview 2

New minor features:

• Independent Front/Back Stencil Refs and Masks
• Triangle fan

See here for more information

Video Encode Release