Direct3D feature levels discussion

devblogs.microsoft.com

Announcing our summer Agility SDKs

We’re pleased to announce that we have two new Agility SDKs available today! Agility SDK 1.606.3 contains support for Shader Model 6.7, which we announced was in the works earlier this year. Today, Shader Model 6.7 is ready for you to try!
devblogs.microsoft.com devblogs.microsoft.com
devblogs.microsoft.com

Agility SDK 1.606.3: Shader Model 6.7 is now publicly available!

The DirectX Compiler Team and our partners are pleased to announce the release of Shader Model 6.7! Shader Model 6.7 expands texture fetching, quad querying, and wave capabilities to enable ever more complex (and compatible) new shader-driven features! Advanced Texture Operations SM 6.7 adds a...
devblogs.microsoft.com devblogs.microsoft.com
devblogs.microsoft.com

Preview Agility SDK 1.706.3-preview: SM 6.7, Enhanced Barriers and more

In addition to support for Shader Model 6.7, this preview Agility SDK contains improved support for Enhanced Barriers and some smaller features For more on the public release of Shader Model 6.7, please see our 1.606.3 announcement. With the addition of preview Enhanced Barriers and Relaxed...
devblogs.microsoft.com devblogs.microsoft.com
 
Hi,
I'm attaching new DmitryKo d3d12 caps checker updated to latest agility SDK 1.706.3 in case one wants to provide reports using this updated checker with up to D3D12_FEATURE_DATA_D3D12_OPTIONS15 info..

included is RX vega report using AMD preview driver: https://www.amd.com/en/support/kb/release-notes/rn-rad-ms-agility-sdk-2022-1-706

sadly I don't have an Intel ARC graphics or even a Intel Xe GPU to test with updated Intel drivers like Agility preview (3220):
www.intel.com

Intel® Graphics Windows* DCH Drivers - Preview Features

This download installs Intel® Graphics BETA Preview Feature Driver 31.0.101.3220 for Intel® Arc™ A-Series Graphics intended for developers and early access usage.
www.intel.com www.intel.com
or:
www.intel.com

Download Intel Drivers and Software

Download new and previously released drivers including support software, bios, utilities, firmware and patches for Intel products.
www.intel.com www.intel.com

Nvidia would be nice to see reports on GTX 20 or 30 series using new 525.26 drivers..
 

Attachments

  • d3d12checkjul22src.zip
    5.2 MB · Views: 7
seems 525.26 provide full Agility support..

ADAPTER 0
"NVIDIA GeForce GTX 970"
VEN_10DE, DEV_13C2, SUBSYS_31601462, REV_A1
Dedicated video memory : 4043.4 MB (4239785984 bytes)
Total video memory : 20405.8 MB (21397035008 bytes)
Video driver version : 31.0.15.2526
WDDM version : KMT_DRIVERVERSION_WDDM_3_1 (3100)
Virtual memory model : GPUMMU
Hardware-accelerated scheduler : Disabled, DXGK_FEATURE_SUPPORT_ALWAYS_OFF (0)
GraphicsPreemptionGranularity : DXGI_GRAPHICS_PREEMPTION_DMA_BUFFER_BOUNDARY (0)
ComputePreemptionGranularity : DXGI_COMPUTE_PREEMPTION_DMA_BUFFER_BOUNDARY (0)
Maximum feature level : D3D_FEATURE_LEVEL_12_1 (0xc100)
DoublePrecisionFloatShaderOps : 1
OutputMergerLogicOp : 1
MinPrecisionSupport : D3D12_SHADER_MIN_PRECISION_SUPPORT_NONE (0) (0b0000'0000)
TiledResourcesTier : D3D12_TILED_RESOURCES_TIER_4 (4)
ResourceBindingTier : D3D12_RESOURCE_BINDING_TIER_3 (3)
PSSpecifiedStencilRefSupported : 0
TypedUAVLoadAdditionalFormats : 1
ROVsSupported : 1
ConservativeRasterizationTier : D3D12_CONSERVATIVE_RASTERIZATION_TIER_1 (1)
StandardSwizzle64KBSupported : 0
CrossNodeSharingTier : D3D12_CROSS_NODE_SHARING_TIER_NOT_SUPPORTED (0)
CrossAdapterRowMajorTextureSupported : 0
VPAndRTArrayIndexFromAnyShaderFeedingRasterizerSupportedWithoutGSEmulation : 1
ResourceHeapTier : D3D12_RESOURCE_HEAP_TIER_1 (1)
MaxGPUVirtualAddressBitsPerResource : 40
MaxGPUVirtualAddressBitsPerProcess : 40
Adapter Node 0: TileBasedRenderer: 0, UMA: 0, CacheCoherentUMA: 0, IsolatedMMU: 1, HeapSerializationTier: 0, ProtectedResourceSession.Support: 0
HighestShaderModel : D3D12_SHADER_MODEL_6_7 (0x0067)
WaveOps : 1
WaveLaneCountMin : 32
WaveLaneCountMax : 32
TotalLaneCount : 1664
ExpandedComputeResourceStates : 1
Int64ShaderOps : 1
RootSignature.HighestVersion : D3D_ROOT_SIGNATURE_VERSION_1_1 (2)
DepthBoundsTestSupported : 1
ProgrammableSamplePositionsTier : D3D12_PROGRAMMABLE_SAMPLE_POSITIONS_TIER_2 (2)
ShaderCache.SupportFlags : D3D12_SHADER_CACHE_SUPPORT_SINGLE_PSO | LIBRARY | DRIVER_MANAGED_CACHE | SHADER_CONTROL_CLEAR | SHADER_SESSION_DELETE (115) (0b0111'0011)
CopyQueueTimestampQueriesSupported : 1
CastingFullyTypedFormatSupported : 1
WriteBufferImmediateSupportFlags : D3D12_COMMAND_LIST_SUPPORT_FLAG_DIRECT | BUNDLE | COMPUTE | COPY | VIDEO_DECODE | VIDEO_PROCESS | VIDEO_ENCODE (127) (0b0111'1111)
ViewInstancingTier : D3D12_VIEW_INSTANCING_TIER_2 (2)
BarycentricsSupported : 0
ExistingHeaps.Supported : 1
MSAA64KBAlignedTextureSupported : 1
SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_2 (2)
Native16BitShaderOpsSupported : 0
AtomicShaderInstructions : 0
SRVOnlyTiledResourceTier3 : 1
RenderPassesTier : D3D12_RENDER_PASS_TIER_0 (0)
RaytracingTier : D3D12_RAYTRACING_TIER_NOT_SUPPORTED (0)
AdditionalShadingRatesSupported : 0
PerPrimitiveShadingRateSupportedWithViewportIndexing : 0
VariableShadingRateTier : D3D12_VARIABLE_SHADING_RATE_TIER_NOT_SUPPORTED (0)
ShadingRateImageTileSize : 0
BackgroundProcessingSupported : 1
MeshShaderTier : D3D12_MESH_SHADER_TIER_NOT_SUPPORTED (0)
SamplerFeedbackTier : D3D12_SAMPLER_FEEDBACK_TIER_NOT_SUPPORTED (0)
UnalignedBlockTexturesSupported : 1
MeshShaderPipelineStatsSupported : 0
MeshShaderSupportsFullRangeRenderTargetArrayIndex : 0
AtomicInt64OnTypedResourceSupported : 1
AtomicInt64OnGroupSharedSupported : 1
DerivativesInMeshAndAmplificationShadersSupported : 0
WaveMMATier : D3D12_WAVE_MMA_TIER_NOT_SUPPORTED (0)
VariableRateShadingSumCombinerSupported : 0
MeshShaderPerPrimitiveShadingRateSupported : 0
AtomicInt64OnDescriptorHeapResourceSupported : 1
EnhancedBarriersSupported : 1
MSPrimitivesPipelineStatisticIncludesCulledPrimitives : 1
RelaxedFormatCastingSupported : 1
UnrestrictedBufferTextureCopyPitchSupported : 1
UnrestrictedVertexElementAlignmentSupported : 1
InvertedViewportHeightFlipsYSupported : 1
InvertedViewportDepthFlipsZSupported : 1
TextureCopyBetweenDimensionsSupported : 1
AlphaBlendFactorSupported : 1
AdvancedTextureOpsSupported : 1
WriteableMSAATexturesSupported : 1
IndependentFrontAndBackStencilRefMaskSupported : 1
TriangleFanSupported : 1
DynamicIndexBufferStripCutSupported : 0
DisplayableTexture : 0
DisplayableTexture.SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_0 (0)
DirectML maximum feature level : DML_FEATURE_LEVEL_5_0 (0x5000)
Metacommands enumerated : 0
 
Vega with Agility driver:
"Radeon RX Vega"
VEN_1002, DEV_687F, SUBSYS_6B761002, REV_C1
Dedicated video memory : 8119.2 MB (8513564672 bytes)
Total video memory : 24481.6 MB (25670813696 bytes)
Video driver version : 30.0.21023.5005
WDDM version : KMT_DRIVERVERSION_WDDM_3_0 (3000)
Virtual memory model : GPUMMU
Hardware-accelerated scheduler : Disabled, DXGK_FEATURE_SUPPORT_ALWAYS_OFF (0)
GraphicsPreemptionGranularity : DXGI_GRAPHICS_PREEMPTION_PRIMITIVE_BOUNDARY (1)
ComputePreemptionGranularity : DXGI_COMPUTE_PREEMPTION_DMA_BUFFER_BOUNDARY (0)
Maximum feature level : D3D_FEATURE_LEVEL_12_1 (0xc100)
DoublePrecisionFloatShaderOps : 1
OutputMergerLogicOp : 1
MinPrecisionSupport : D3D12_SHADER_MIN_PRECISION_SUPPORT_16_BIT (2) (0b0000'0010)
TiledResourcesTier : D3D12_TILED_RESOURCES_TIER_4 (4)
ResourceBindingTier : D3D12_RESOURCE_BINDING_TIER_3 (3)
PSSpecifiedStencilRefSupported : 1
TypedUAVLoadAdditionalFormats : 1
ROVsSupported : 1
ConservativeRasterizationTier : D3D12_CONSERVATIVE_RASTERIZATION_TIER_3 (3)
StandardSwizzle64KBSupported : 0
CrossNodeSharingTier : D3D12_CROSS_NODE_SHARING_TIER_NOT_SUPPORTED (0)
CrossAdapterRowMajorTextureSupported : 0
VPAndRTArrayIndexFromAnyShaderFeedingRasterizerSupportedWithoutGSEmulation : 1
ResourceHeapTier : D3D12_RESOURCE_HEAP_TIER_2 (2)
MaxGPUVirtualAddressBitsPerResource : 44
MaxGPUVirtualAddressBitsPerProcess : 44
Adapter Node 0: TileBasedRenderer: 0, UMA: 0, CacheCoherentUMA: 0, IsolatedMMU: 1, HeapSerializationTier: 0, ProtectedResourceSession.Support: 0
HighestShaderModel : D3D12_SHADER_MODEL_6_7 (0x0067)
WaveOps : 1
WaveLaneCountMin : 64
WaveLaneCountMax : 64
TotalLaneCount : 3584
ExpandedComputeResourceStates : 1
Int64ShaderOps : 1
RootSignature.HighestVersion : D3D_ROOT_SIGNATURE_VERSION_1_1 (2)
DepthBoundsTestSupported : 1
ProgrammableSamplePositionsTier : D3D12_PROGRAMMABLE_SAMPLE_POSITIONS_TIER_2 (2)
ShaderCache.SupportFlags : D3D12_SHADER_CACHE_SUPPORT_SINGLE_PSO | LIBRARY | AUTOMATIC_INPROC_CACHE | AUTOMATIC_DISK_CACHE | DRIVER_MANAGED_CACHE | SHADER_CONTROL_CLEAR | SHADER_SESSION_DELETE (127) (0b0111'1111)
CopyQueueTimestampQueriesSupported : 1
CastingFullyTypedFormatSupported : 1
WriteBufferImmediateSupportFlags : D3D12_COMMAND_LIST_SUPPORT_FLAG_DIRECT | BUNDLE | COMPUTE | COPY (15) (0b0000'1111)
ViewInstancingTier : D3D12_VIEW_INSTANCING_TIER_1 (1)
BarycentricsSupported : 0
ExistingHeaps.Supported : 1
MSAA64KBAlignedTextureSupported : 1
SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_2 (2)
Native16BitShaderOpsSupported : 1
AtomicShaderInstructions : 0
SRVOnlyTiledResourceTier3 : 1
RenderPassesTier : D3D12_RENDER_PASS_TIER_0 (0)
RaytracingTier : D3D12_RAYTRACING_TIER_NOT_SUPPORTED (0)
AdditionalShadingRatesSupported : 0
PerPrimitiveShadingRateSupportedWithViewportIndexing : 0
VariableShadingRateTier : D3D12_VARIABLE_SHADING_RATE_TIER_NOT_SUPPORTED (0)
ShadingRateImageTileSize : 0
BackgroundProcessingSupported : 0
MeshShaderTier : D3D12_MESH_SHADER_TIER_NOT_SUPPORTED (0)
SamplerFeedbackTier : D3D12_SAMPLER_FEEDBACK_TIER_NOT_SUPPORTED (0)
UnalignedBlockTexturesSupported : 1
MeshShaderPipelineStatsSupported : 0
MeshShaderSupportsFullRangeRenderTargetArrayIndex : 0
AtomicInt64OnTypedResourceSupported : 1
AtomicInt64OnGroupSharedSupported : 1
DerivativesInMeshAndAmplificationShadersSupported : 0
WaveMMATier : D3D12_WAVE_MMA_TIER_NOT_SUPPORTED (0)
VariableRateShadingSumCombinerSupported : 0
MeshShaderPerPrimitiveShadingRateSupported : 0
AtomicInt64OnDescriptorHeapResourceSupported : 1
EnhancedBarriersSupported : 1
MSPrimitivesPipelineStatisticIncludesCulledPrimitives : 0
RelaxedFormatCastingSupported : 1
UnrestrictedBufferTextureCopyPitchSupported : 1
UnrestrictedVertexElementAlignmentSupported : 1
InvertedViewportHeightFlipsYSupported : 1
InvertedViewportDepthFlipsZSupported : 1
TextureCopyBetweenDimensionsSupported : 1
AlphaBlendFactorSupported : 1
AdvancedTextureOpsSupported : 1
WriteableMSAATexturesSupported : 1
IndependentFrontAndBackStencilRefMaskSupported : 1
TriangleFanSupported : 1
DynamicIndexBufferStripCutSupported : 0
DisplayableTexture : 0
DisplayableTexture.SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_0 (0)
DirectML maximum feature level : DML_FEATURE_LEVEL_5_0 (0x5000)
Metacommands enumerated : 4
Metacommands [parameters per stage]: Conv (Convolution) [84][1][6], Conv (Convolution) [108][5][6], GEMM (General matrix multiply) [67][1][6], GEMM (General matrix multiply) [91][5][6]
 
I've made an update to the Direct3D12 feature checker tool to report D3D12_FEATURE_D3D12_OPTIONS12, D3D12_FEATURE_D3D12_OPTIONS13 in Windows 11 Insider Preview Nickel (build 22621), as well as D3D12_FEATURE_D3D12_OPTIONS14, D3D12_FEATURE_D3D12_OPTIONS15, D3D12_FEATURE_D3D12_OPTIONS16, and shader model 6_8 in Insider Preview Copper (build 25193).

For Windows 10 users, the same features can be reported (except options 16) with the Agile version of the tool - this requires the redistributable runtime from the Direct3D 12 Agility SDK 1706.4 Preview with Developer mode enabled.
Download the latest NuGet package and extract build\native\bin\x64\D3D12Core.dll to the D3D12\ subfolder using either NuGetPackageExplorer, 7-Zip, WinRAR, or File Explorer.

This is how new features are reported by Adrenalin driver 22.6.1 for Agility SDK 1706 (build 30.0.21023.5005) on my Radeon 5700 XT:

Code: 
MSPrimitivesPipelineStatisticIncludesCulledPrimitives : 0
EnhancedBarriersSupported : 1
RelaxedFormatCastingSupported : 1
UnrestrictedBufferTextureCopyPitchSupported : 1
UnrestrictedVertexElementAlignmentSupported : 1
InvertedViewportHeightFlipsYSupported : 1
InvertedViewportDepthFlipsZSupported : 1
TextureCopyBetweenDimensionsSupported : 1
AlphaBlendFactorSupported : 1
AdvancedTextureOpsSupported : 1
WriteableMSAATexturesSupported : 1
IndependentFrontAndBackStencilRefMaskSupported : 1
TriangleFanSupported : 1
DynamicIndexBufferStripCutSupported : 0
DynamicDepthBiasSupported: 0
GPUUploadHeapSupported: 0

Also shader model 6.7 and experimental shader model 6.8 are supported.

Note that if you have an older video driver that doesn't support MSPrimitivesPipelineStatisticIncludesCulledPrimitives, it will report Unknown (-1) instead of True (1) or False (0).


PS. I will focus on updating the Agile executable for new Agility SDK releases going forward; regular versions should continue to work in Windows 10 and new Insider Preview builds, but I do not plan to installl Windows 11 10X in the foreseeable future.

PPS. Microsoft broke OS versioning again, so major/minor version, ProductName and ReleaseID were frozen for some reason. Working around it requires a disproportionate amount of effort, therefore OS version will be reported as is and ReleaseID will be superseded by DisplayVersion if available. If they have to insist OS version remains at "10.0", so be it. Build number, UBR (update build release) number and release branch name are still being updated.
 
Last edited:
Windows 10 version 2009 (build 19045.1889 vb_release) x64

ADAPTER 0
"NVIDIA GeForce RTX 3060 Ti"
VEN_10DE, DEV_2489, SUBSYS_88271043, REV_A1
Dedicated video memory : 8050.0 MB (8441036800 bytes)
Total video memory : 16205.4 MB (16992622592 bytes)
Video driver version : 31.0.15.1694
WDDM version : KMT_DRIVERVERSION_WDDM_2_7 (2700)
Virtual memory model : GPUMMU
Hardware-accelerated scheduler : Disabled
GraphicsPreemptionGranularity : DXGI_GRAPHICS_PREEMPTION_PIXEL_BOUNDARY (3)
ComputePreemptionGranularity : DXGI_COMPUTE_PREEMPTION_DISPATCH_BOUNDARY (1)
Failed to create Direct3D 12 device
E r r o r 0 X 8 8 7 E 0 0 0 3 : D i e D 3 D 1 2 S D K - V e r s i o n s k o n f i g u r a t i o n v o n H o s t . e x e i s t u n g ü l t i g .
Click to expand...

Hmm, not sure what I did wrong?
 
Ampere on Win11 2022:
Code: 
Direct3D 12 feature checker (August 2022) by DmitryKo (x64)
https://forum.beyond3d.com/posts/1840641/

Windows 10 version 2009 (build 22621.521 ni_release) x64

ADAPTER 0
"NVIDIA GeForce RTX 3080"
VEN_10DE, DEV_2206, SUBSYS_87AC1043, REV_A1
Dedicated video memory : 10078.0 MB (10567548928 bytes)
Total video memory : 42802.6 MB (44881825792 bytes)
Video driver version : 31.0.15.1740
WDDM version : KMT_DRIVERVERSION_WDDM_3_1 (3100)
Virtual memory model : GPUMMU
Hardware-accelerated scheduler : Enabled, DXGK_FEATURE_SUPPORT_STABLE (2)
GraphicsPreemptionGranularity : DXGI_GRAPHICS_PREEMPTION_PIXEL_BOUNDARY (3)
ComputePreemptionGranularity : DXGI_COMPUTE_PREEMPTION_DISPATCH_BOUNDARY (1)
Maximum feature level : D3D_FEATURE_LEVEL_12_2 (0xc200)
DoublePrecisionFloatShaderOps : 1
OutputMergerLogicOp : 1
MinPrecisionSupport : D3D12_SHADER_MIN_PRECISION_SUPPORT_16_BIT (2) (0b0000'0010)
TiledResourcesTier : D3D12_TILED_RESOURCES_TIER_3 (3)
ResourceBindingTier : D3D12_RESOURCE_BINDING_TIER_3 (3)
PSSpecifiedStencilRefSupported : 0
TypedUAVLoadAdditionalFormats : 1
ROVsSupported : 1
ConservativeRasterizationTier : D3D12_CONSERVATIVE_RASTERIZATION_TIER_3 (3)
StandardSwizzle64KBSupported : 0
CrossNodeSharingTier : D3D12_CROSS_NODE_SHARING_TIER_NOT_SUPPORTED (0)
CrossAdapterRowMajorTextureSupported : 0
VPAndRTArrayIndexFromAnyShaderFeedingRasterizerSupportedWithoutGSEmulation : 1
ResourceHeapTier : D3D12_RESOURCE_HEAP_TIER_2 (2)
MaxGPUVirtualAddressBitsPerResource : 40
MaxGPUVirtualAddressBitsPerProcess : 40
Adapter Node 0:     TileBasedRenderer: 0, UMA: 0, CacheCoherentUMA: 0, IsolatedMMU: 1, HeapSerializationTier: 0, ProtectedResourceSession.Support: 1, ProtectedResourceSessionTypeCount: 1 D3D12_PROTECTED_RESOURCES_SESSION_HARDWARE_PROTECTED
HighestShaderModel : D3D12_SHADER_MODEL_6_6 (0x0066)
WaveOps : 1
WaveLaneCountMin : 32
WaveLaneCountMax : 32
TotalLaneCount : 8704
ExpandedComputeResourceStates : 1
Int64ShaderOps : 1
RootSignature.HighestVersion : D3D_ROOT_SIGNATURE_VERSION_1_1 (2)
DepthBoundsTestSupported : 1
ProgrammableSamplePositionsTier : D3D12_PROGRAMMABLE_SAMPLE_POSITIONS_TIER_2 (2)
ShaderCache.SupportFlags : D3D12_SHADER_CACHE_SUPPORT_SINGLE_PSO | LIBRARY | DRIVER_MANAGED_CACHE | SHADER_CONTROL_CLEAR | SHADER_SESSION_DELETE (115) (0b0111'0011)
CopyQueueTimestampQueriesSupported : 1
CastingFullyTypedFormatSupported : 1
WriteBufferImmediateSupportFlags : D3D12_COMMAND_LIST_SUPPORT_FLAG_DIRECT | BUNDLE | COMPUTE | COPY | VIDEO_DECODE | VIDEO_PROCESS | VIDEO_ENCODE (127) (0b0111'1111)
ViewInstancingTier : D3D12_VIEW_INSTANCING_TIER_3 (3)
BarycentricsSupported : 1
ExistingHeaps.Supported : 1
MSAA64KBAlignedTextureSupported : 1
SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_2 (2)
Native16BitShaderOpsSupported : 1
AtomicShaderInstructions : 0
SRVOnlyTiledResourceTier3 : 1
RenderPassesTier : D3D12_RENDER_PASS_TIER_0 (0)
RaytracingTier : D3D12_RAYTRACING_TIER_1_1 (11)
AdditionalShadingRatesSupported : 1
PerPrimitiveShadingRateSupportedWithViewportIndexing : 1
VariableShadingRateTier : D3D12_VARIABLE_SHADING_RATE_TIER_2 (2)
ShadingRateImageTileSize : 16
BackgroundProcessingSupported : 1
MeshShaderTier : D3D12_MESH_SHADER_TIER_1 (10)
SamplerFeedbackTier : D3D12_SAMPLER_FEEDBACK_TIER_0_9 (90)
UnalignedBlockTexturesSupported : 1
MeshShaderPipelineStatsSupported : 1
MeshShaderSupportsFullRangeRenderTargetArrayIndex : 1
AtomicInt64OnTypedResourceSupported : 1
AtomicInt64OnGroupSharedSupported : 1
DerivativesInMeshAndAmplificationShadersSupported : 0
WaveMMATier : D3D12_WAVE_MMA_TIER_NOT_SUPPORTED (0)
VariableRateShadingSumCombinerSupported : 1
MeshShaderPerPrimitiveShadingRateSupported : 1
AtomicInt64OnDescriptorHeapResourceSupported : 1
DisplayableTexture : 0
DisplayableTexture.SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_0 (0)
MSPrimitivesPipelineStatisticIncludesCulledPrimitives : 1
EnhancedBarriersSupported : 0
RelaxedFormatCastingSupported : 0
UnrestrictedBufferTextureCopyPitchSupported : 1
UnrestrictedVertexElementAlignmentSupported : 1
InvertedViewportHeightFlipsYSupported : 1
InvertedViewportDepthFlipsZSupported : 1
TextureCopyBetweenDimensionsSupported : 1
AlphaBlendFactorSupported : 1
DirectML maximum feature level : DML_FEATURE_LEVEL_5_0 (0x5000)
Metacommands enumerated : 8
Metacommands [parameters per stage]: Conv (Convolution) [84][1][6], CopyTensor [3][1][31], MVN (Mean Variance Normalization) [67][1][6], GEMM (General matrix multiply) [67][1][6], Conv (Convolution) [108][5][6], GEMM (General matrix multiply) [91][5][6], MVN (Mean Variance Normalization) [91][5][6], Pooling [56][3][4]

And Agile with Developer mode On:

Code: 
Direct3D 12 feature checker (August 2022) by DmitryKo (x64) Agile v706
https://forum.beyond3d.com/posts/1840641/

Windows 10 version 2009 (build 22621.521 ni_release) x64
Checking for experimental features SM6 TR4 META

ADAPTER 0
"NVIDIA GeForce RTX 3080"
VEN_10DE, DEV_2206, SUBSYS_87AC1043, REV_A1
Dedicated video memory : 10078.0 MB (10567548928 bytes)
Total video memory : 42802.6 MB (44881825792 bytes)
Video driver version : 31.0.15.1740
WDDM version : KMT_DRIVERVERSION_WDDM_3_1 (3100)
Virtual memory model : GPUMMU
Hardware-accelerated scheduler : Enabled, DXGK_FEATURE_SUPPORT_STABLE (2)
GraphicsPreemptionGranularity : DXGI_GRAPHICS_PREEMPTION_PIXEL_BOUNDARY (3)
ComputePreemptionGranularity : DXGI_COMPUTE_PREEMPTION_DISPATCH_BOUNDARY (1)
Maximum feature level : D3D_FEATURE_LEVEL_12_2 (0xc200)
DoublePrecisionFloatShaderOps : 1
OutputMergerLogicOp : 1
MinPrecisionSupport : D3D12_SHADER_MIN_PRECISION_SUPPORT_16_BIT (2) (0b0000'0010)
TiledResourcesTier : D3D12_TILED_RESOURCES_TIER_4 (4)
ResourceBindingTier : D3D12_RESOURCE_BINDING_TIER_3 (3)
PSSpecifiedStencilRefSupported : 0
TypedUAVLoadAdditionalFormats : 1
ROVsSupported : 1
ConservativeRasterizationTier : D3D12_CONSERVATIVE_RASTERIZATION_TIER_3 (3)
StandardSwizzle64KBSupported : 0
CrossNodeSharingTier : D3D12_CROSS_NODE_SHARING_TIER_NOT_SUPPORTED (0)
CrossAdapterRowMajorTextureSupported : 0
VPAndRTArrayIndexFromAnyShaderFeedingRasterizerSupportedWithoutGSEmulation : 1
ResourceHeapTier : D3D12_RESOURCE_HEAP_TIER_2 (2)
MaxGPUVirtualAddressBitsPerResource : 40
MaxGPUVirtualAddressBitsPerProcess : 40
Adapter Node 0:     TileBasedRenderer: 0, UMA: 0, CacheCoherentUMA: 0, IsolatedMMU: 1, HeapSerializationTier: 0, ProtectedResourceSession.Support: 1, ProtectedResourceSessionTypeCount: 1 D3D12_PROTECTED_RESOURCES_SESSION_HARDWARE_PROTECTED
HighestShaderModel : D3D12_SHADER_MODEL_6_7 (0x0067)
WaveOps : 1
WaveLaneCountMin : 32
WaveLaneCountMax : 32
TotalLaneCount : 8704
ExpandedComputeResourceStates : 1
Int64ShaderOps : 1
RootSignature.HighestVersion : D3D_ROOT_SIGNATURE_VERSION_1_1 (2)
DepthBoundsTestSupported : 1
ProgrammableSamplePositionsTier : D3D12_PROGRAMMABLE_SAMPLE_POSITIONS_TIER_2 (2)
ShaderCache.SupportFlags : D3D12_SHADER_CACHE_SUPPORT_SINGLE_PSO | LIBRARY | DRIVER_MANAGED_CACHE | SHADER_CONTROL_CLEAR | SHADER_SESSION_DELETE (115) (0b0111'0011)
CopyQueueTimestampQueriesSupported : 1
CastingFullyTypedFormatSupported : 1
WriteBufferImmediateSupportFlags : D3D12_COMMAND_LIST_SUPPORT_FLAG_DIRECT | BUNDLE | COMPUTE | COPY | VIDEO_DECODE | VIDEO_PROCESS | VIDEO_ENCODE (127) (0b0111'1111)
ViewInstancingTier : D3D12_VIEW_INSTANCING_TIER_3 (3)
BarycentricsSupported : 1
ExistingHeaps.Supported : 1
MSAA64KBAlignedTextureSupported : 1
SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_2 (2)
Native16BitShaderOpsSupported : 1
AtomicShaderInstructions : 0
SRVOnlyTiledResourceTier3 : 1
RenderPassesTier : D3D12_RENDER_PASS_TIER_0 (0)
RaytracingTier : D3D12_RAYTRACING_TIER_1_1 (11)
AdditionalShadingRatesSupported : 1
PerPrimitiveShadingRateSupportedWithViewportIndexing : 1
VariableShadingRateTier : D3D12_VARIABLE_SHADING_RATE_TIER_2 (2)
ShadingRateImageTileSize : 16
BackgroundProcessingSupported : 1
MeshShaderTier : D3D12_MESH_SHADER_TIER_1 (10)
SamplerFeedbackTier : D3D12_SAMPLER_FEEDBACK_TIER_0_9 (90)
UnalignedBlockTexturesSupported : 1
MeshShaderPipelineStatsSupported : 1
MeshShaderSupportsFullRangeRenderTargetArrayIndex : 1
AtomicInt64OnTypedResourceSupported : 1
AtomicInt64OnGroupSharedSupported : 1
DerivativesInMeshAndAmplificationShadersSupported : 0
WaveMMATier : D3D12_WAVE_MMA_TIER_NOT_SUPPORTED (0)
VariableRateShadingSumCombinerSupported : 1
MeshShaderPerPrimitiveShadingRateSupported : 1
AtomicInt64OnDescriptorHeapResourceSupported : 1
DisplayableTexture : 0
DisplayableTexture.SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_0 (0)
MSPrimitivesPipelineStatisticIncludesCulledPrimitives : 1
EnhancedBarriersSupported : 1
RelaxedFormatCastingSupported : 0
UnrestrictedBufferTextureCopyPitchSupported : 1
UnrestrictedVertexElementAlignmentSupported : 1
InvertedViewportHeightFlipsYSupported : 1
InvertedViewportDepthFlipsZSupported : 1
TextureCopyBetweenDimensionsSupported : 1
AlphaBlendFactorSupported : 1
AdvancedTextureOpsSupported : 0
WriteableMSAATexturesSupported : 0
IndependentFrontAndBackStencilRefMaskSupported : 0
TriangleFanSupported : 0
DynamicIndexBufferStripCutSupported : 0
DirectML maximum feature level : DML_FEATURE_LEVEL_5_0 (0x5000)
Metacommands enumerated : 9
Metacommands [parameters per stage]: Conv (Convolution) [84][1][6], CopyTensor [3][1][31], 2x2 Nearest neighbour Upsample [15][1][2], MVN (Mean Variance Normalization) [67][1][6], GEMM (General matrix multiply) [67][1][6], Conv (Convolution) [108][5][6], GEMM (General matrix multiply) [91][5][6], MVN (Mean Variance Normalization) [91][5][6], Pooling [56][3][4]
 
Last edited:
Ampere using Win10 with Agility SDK:

Code: 
Direct3D 12 feature checker (August 2022) by DmitryKo (x64) (Agile)
[URL unfurl="true"]https://forum.beyond3d.com/posts/1840641/[/URL]

Windows 10 version 2009 (build 19045.1889 vb_release) x64

ADAPTER 0
"NVIDIA GeForce RTX 3060 Ti"
VEN_10DE, DEV_2489, SUBSYS_88271043, REV_A1
Dedicated video memory : 8050.0 MB (8441036800 bytes)
Total video memory : 16205.4 MB (16992622592 bytes)
Video driver version : 31.0.15.1694
WDDM version : KMT_DRIVERVERSION_WDDM_2_7 (2700)
Virtual memory model : GPUMMU
Hardware-accelerated scheduler : Disabled
GraphicsPreemptionGranularity : DXGI_GRAPHICS_PREEMPTION_PIXEL_BOUNDARY (3)
ComputePreemptionGranularity : DXGI_COMPUTE_PREEMPTION_DISPATCH_BOUNDARY (1)
Maximum feature level : D3D_FEATURE_LEVEL_12_2 (0xc200)
DoublePrecisionFloatShaderOps : 1
OutputMergerLogicOp : 1
MinPrecisionSupport : D3D12_SHADER_MIN_PRECISION_SUPPORT_16_BIT (2) (0b0000'0010)
TiledResourcesTier : D3D12_TILED_RESOURCES_TIER_3 (3)
ResourceBindingTier : D3D12_RESOURCE_BINDING_TIER_3 (3)
PSSpecifiedStencilRefSupported : 0
TypedUAVLoadAdditionalFormats : 1
ROVsSupported : 1
ConservativeRasterizationTier : D3D12_CONSERVATIVE_RASTERIZATION_TIER_3 (3)
StandardSwizzle64KBSupported : 0
CrossNodeSharingTier : D3D12_CROSS_NODE_SHARING_TIER_NOT_SUPPORTED (0)
CrossAdapterRowMajorTextureSupported : 0
VPAndRTArrayIndexFromAnyShaderFeedingRasterizerSupportedWithoutGSEmulation : 1
ResourceHeapTier : D3D12_RESOURCE_HEAP_TIER_2 (2)
MaxGPUVirtualAddressBitsPerResource : 40
MaxGPUVirtualAddressBitsPerProcess : 40
Adapter Node 0:     TileBasedRenderer: 0, UMA: 0, CacheCoherentUMA: 0, IsolatedMMU: 1, HeapSerializationTier: 0, ProtectedResourceSession.Support: 1, ProtectedResourceSessionTypeCount: 1 D3D12_PROTECTED_RESOURCES_SESSION_HARDWARE_PROTECTED
HighestShaderModel : D3D12_SHADER_MODEL_6_6 (0x0066)
WaveOps : 1
WaveLaneCountMin : 32
WaveLaneCountMax : 32
TotalLaneCount : 4864
ExpandedComputeResourceStates : 1
Int64ShaderOps : 1
RootSignature.HighestVersion : D3D_ROOT_SIGNATURE_VERSION_1_1 (2)
DepthBoundsTestSupported : 1
ProgrammableSamplePositionsTier : D3D12_PROGRAMMABLE_SAMPLE_POSITIONS_TIER_2 (2)
ShaderCache.SupportFlags : D3D12_SHADER_CACHE_SUPPORT_SINGLE_PSO | LIBRARY (3) (0b0000'0011)
CopyQueueTimestampQueriesSupported : 1
CastingFullyTypedFormatSupported : 1
WriteBufferImmediateSupportFlags : D3D12_COMMAND_LIST_SUPPORT_FLAG_DIRECT | BUNDLE | COMPUTE | COPY | VIDEO_DECODE | VIDEO_PROCESS | VIDEO_ENCODE (127) (0b0111'1111)
ViewInstancingTier : D3D12_VIEW_INSTANCING_TIER_3 (3)
BarycentricsSupported : 1
ExistingHeaps.Supported : 1
MSAA64KBAlignedTextureSupported : 1
SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_2 (2)
Native16BitShaderOpsSupported : 1
AtomicShaderInstructions : 0
SRVOnlyTiledResourceTier3 : 1
RenderPassesTier : D3D12_RENDER_PASS_TIER_0 (0)
RaytracingTier : D3D12_RAYTRACING_TIER_1_1 (11)
AdditionalShadingRatesSupported : 1
PerPrimitiveShadingRateSupportedWithViewportIndexing : 1
VariableShadingRateTier : D3D12_VARIABLE_SHADING_RATE_TIER_2 (2)
ShadingRateImageTileSize : 16
BackgroundProcessingSupported : 1
MeshShaderTier : D3D12_MESH_SHADER_TIER_1 (10)
SamplerFeedbackTier : D3D12_SAMPLER_FEEDBACK_TIER_0_9 (90)
UnalignedBlockTexturesSupported : 1
MeshShaderPipelineStatsSupported : 1
MeshShaderSupportsFullRangeRenderTargetArrayIndex : 1
AtomicInt64OnTypedResourceSupported : 1
AtomicInt64OnGroupSharedSupported : 1
DerivativesInMeshAndAmplificationShadersSupported : 0
WaveMMATier : D3D12_WAVE_MMA_TIER_NOT_SUPPORTED (0)
VariableRateShadingSumCombinerSupported : 1
MeshShaderPerPrimitiveShadingRateSupported : 1
AtomicInt64OnDescriptorHeapResourceSupported : 1
DisplayableTexture : 0
DisplayableTexture.SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_0 (0)
MSPrimitivesPipelineStatisticIncludesCulledPrimitives : 1
EnhancedBarriersSupported : 1
RelaxedFormatCastingSupported : 0
UnrestrictedBufferTextureCopyPitchSupported : 1
UnrestrictedVertexElementAlignmentSupported : 1
InvertedViewportHeightFlipsYSupported : 1
InvertedViewportDepthFlipsZSupported : 1
TextureCopyBetweenDimensionsSupported : 1
AlphaBlendFactorSupported : 1
AdvancedTextureOpsSupported : 0
WriteableMSAATexturesSupported : 0
IndependentFrontAndBackStencilRefMaskSupported : 0
TriangleFanSupported : 0
DynamicIndexBufferStripCutSupported : 0
DirectML maximum feature level : DML_FEATURE_LEVEL_2_0 (0x2000)
Metacommands enumerated : 9
Metacommands [parameters per stage]: Conv (Convolution) [84][1][6], CopyTensor [3][1][31], 2x2 Nearest neighbour Upsample [15][1][2], MVN (Mean Variance Normalization) [67][1][6], GEMM (General matrix multiply) [67][1][6], Conv (Convolution) [108][5][6], GEMM (General matrix multiply) [91][5][6], MVN (Mean Variance Normalization) [91][5][6], Pooling [56][3][4]
 
DirectML feature level is not related to driver capabilities; it's a "software" feature to specify a set of supported operators in each major release. DirectML runtime in Widows 10 only supports level 2_0, you need to download the latest DirectML distributable package and extract bin\x64-win\DirectML.dll to the tool folder, so it would load instead of a system-wide DLL.

But it doesn't really have any effect on actual "hardware" features, i.e. which Direct3D 12 metacommands are implemented on each specific GPU. So maybe I'd remove it from the report in a future update.

PS. I've removed the reporting of DirectML feature level and also fixed a few small issues; the ZIP archive attached to the original post has been updated.
 
Last edited:
UHD730 integrated graphics in Intel Core i5-12400 (Alder Lake):

Code: 
Direct3D 12 feature checker (August 2022) by DmitryKo (x64) (Agility SDK v706)
https://forum.beyond3d.com/posts/1840641/

Windows 10 version 22H2 (build 19045.2006 vb_release) x64
Checking for experimental features SM6 TR4 META

ADAPTER 0
"Intel(R) UHD Graphics 730"
VEN_8086, DEV_4692, SUBSYS_7D251462, REV_0C
Dedicated video memory : 128.0 MB (134217728 bytes)
Total video memory : 16405.9 MB (17202868224 bytes)
BIOS string : Intel Video BIOS
Video driver version : 31.0.101.3430
WDDM version : KMT_DRIVERVERSION_WDDM_2_7 (2700)
Virtual memory model : GPUMMU
Hardware-accelerated scheduler : Not supported
GraphicsPreemptionGranularity : DXGI_GRAPHICS_PREEMPTION_TRIANGLE_BOUNDARY (2)
ComputePreemptionGranularity : DXGI_COMPUTE_PREEMPTION_THREAD_GROUP_BOUNDARY (2)
Maximum feature level : D3D_FEATURE_LEVEL_12_1 (0xc100)
DoublePrecisionFloatShaderOps : 0
OutputMergerLogicOp : 1
MinPrecisionSupport : D3D12_SHADER_MIN_PRECISION_SUPPORT_16_BIT (2) (0b0000'0010)
TiledResourcesTier : D3D12_TILED_RESOURCES_TIER_3 (3)
ResourceBindingTier : D3D12_RESOURCE_BINDING_TIER_3 (3)
PSSpecifiedStencilRefSupported : 1
TypedUAVLoadAdditionalFormats : 1
ROVsSupported : 1
ConservativeRasterizationTier : D3D12_CONSERVATIVE_RASTERIZATION_TIER_3 (3)
StandardSwizzle64KBSupported : 0
CrossNodeSharingTier : D3D12_CROSS_NODE_SHARING_TIER_NOT_SUPPORTED (0)
CrossAdapterRowMajorTextureSupported : 1
VPAndRTArrayIndexFromAnyShaderFeedingRasterizerSupportedWithoutGSEmulation : 1
ResourceHeapTier : D3D12_RESOURCE_HEAP_TIER_2 (2)
MaxGPUVirtualAddressBitsPerResource : 44
MaxGPUVirtualAddressBitsPerProcess : 48
Adapter Node 0:     TileBasedRenderer: 0, UMA: 1, CacheCoherentUMA: 1, IsolatedMMU: 1, HeapSerializationTier: 0, ProtectedResourceSession.Support: 1, ProtectedResourceSessionTypeCount: 1 D3D12_PROTECTED_RESOURCES_SESSION_HARDWARE_PROTECTED
HighestShaderModel : D3D12_SHADER_MODEL_6_8 (0x0068)
WaveOps : 1
WaveLaneCountMin : 16
WaveLaneCountMax : 16
TotalLaneCount : 384
ExpandedComputeResourceStates : 1
Int64ShaderOps : 1
RootSignature.HighestVersion : D3D_ROOT_SIGNATURE_VERSION_1_1 (2)
DepthBoundsTestSupported : 1
ProgrammableSamplePositionsTier : D3D12_PROGRAMMABLE_SAMPLE_POSITIONS_TIER_1 (1)
ShaderCache.SupportFlags : D3D12_SHADER_CACHE_SUPPORT_SINGLE_PSO | LIBRARY | AUTOMATIC_INPROC_CACHE | AUTOMATIC_DISK_CACHE (15) (0b0000'1111)
CopyQueueTimestampQueriesSupported : 1
CastingFullyTypedFormatSupported : 1
WriteBufferImmediateSupportFlags : D3D12_COMMAND_LIST_SUPPORT_FLAG_DIRECT | BUNDLE | COMPUTE | COPY (15) (0b0000'1111)
ViewInstancingTier : D3D12_VIEW_INSTANCING_TIER_2 (2)
BarycentricsSupported : 0
ExistingHeaps.Supported : 1
MSAA64KBAlignedTextureSupported : 1
SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_2 (2)
Native16BitShaderOpsSupported : 1
AtomicShaderInstructions : 0
SRVOnlyTiledResourceTier3 : 1
RenderPassesTier : D3D12_RENDER_PASS_TIER_0 (0)
RaytracingTier : D3D12_RAYTRACING_TIER_NOT_SUPPORTED (0)
AdditionalShadingRatesSupported : 1
PerPrimitiveShadingRateSupportedWithViewportIndexing : 0
VariableShadingRateTier : D3D12_VARIABLE_SHADING_RATE_TIER_1 (1)
ShadingRateImageTileSize : 0
BackgroundProcessingSupported : 1
MeshShaderTier : D3D12_MESH_SHADER_TIER_NOT_SUPPORTED (0)
SamplerFeedbackTier : D3D12_SAMPLER_FEEDBACK_TIER_0_9 (90)
UnalignedBlockTexturesSupported : 1
MeshShaderPipelineStatsSupported : 0
MeshShaderSupportsFullRangeRenderTargetArrayIndex : 0
AtomicInt64OnTypedResourceSupported : 0
AtomicInt64OnGroupSharedSupported : 0
DerivativesInMeshAndAmplificationShadersSupported : 0
WaveMMATier : D3D12_WAVE_MMA_TIER_NOT_SUPPORTED (0)
VariableRateShadingSumCombinerSupported : 0
MeshShaderPerPrimitiveShadingRateSupported : 0
AtomicInt64OnDescriptorHeapResourceSupported : 0
DisplayableTexture : 0
DisplayableTexture.SharedResourceCompatibilityTier : D3D12_SHARED_RESOURCE_COMPATIBILITY_TIER_0 (0)
MSPrimitivesPipelineStatisticIncludesCulledPrimitives : -1
EnhancedBarriersSupported : 0
RelaxedFormatCastingSupported : 0
UnrestrictedBufferTextureCopyPitchSupported : 0
UnrestrictedVertexElementAlignmentSupported : 0
InvertedViewportHeightFlipsYSupported : 0
InvertedViewportDepthFlipsZSupported : 0
TextureCopyBetweenDimensionsSupported : 0
AlphaBlendFactorSupported : 0
AdvancedTextureOpsSupported : 0
WriteableMSAATexturesSupported : 0
IndependentFrontAndBackStencilRefMaskSupported : 0
TriangleFanSupported : 0
DynamicIndexBufferStripCutSupported : 0
Metacommands enumerated : 9
Metacommands [parameters per stage]: Conv (Convolution) [84][1][6], GEMM (General matrix multiply) [67][1][6], Pooling [44][1][4], Conv (Convolution) [108][5][6], GEMM (General matrix multiply) [91][5][6], MVN (Mean Variance Normalization) [91][5][6], Pooling [56][3][4], LSTM (Long Short-Term Memory) [252][10][13], DStorageCustom Metacommand [1][0][9]

PS. DStorageCustom Metacommand with 9 execution parameters??! I wonder what meta commands have to do with DirectStorage, when they are intended to replace standard DirectML operators with optimized hardware-specific implementations in the driver...

Microsoft Learn documentation on ID3D12MetaCommand interface: "A meta command is a Direct3D 12 object representing an algorithm that is accelerated by independent hardware vendors (IHVs). It's an opaque reference to a command generator that is implemented by the driver... You can think of the signature of a meta command as being similar to a C-style function, with multiple in/out parameters, and no return value."

Nvidia Developer blog post Accelerating WinML and NVIDIA Tensor Cores: "Metacommands — Mechanism by which independent hardware providers (such as NVIDIA) can implement overridden versions of operations making the best use of the hardware".
 
Last edited:
FYI here are the parameters of the various stages of the "DStorageCustom Metacommand" in the latest drivers for Intel UHD630 / UHD730:

Code: 
CREATION       [1]:
_In_ UINT64 MaxStreams

INITIALIZATION [0]

EXECUTION      [9]:
_In_ GPU_VIRTUAL_ADDRESS InputBuffer,
_In_ UINT64 InputBufferSize,
_Out_ GPU_VIRTUAL_ADDRESS OutputBuffer,
_In_ UINT64 OutputBufferSize,
_Inout_ GPU_VIRTUAL_ADDRESS ControlBuffer,
_In_ UINT64 ControlBufferSize,
_Inout_ GPU_VIRTUAL_ADDRESS ScratchBuffer,
_In_ UINT64 ScratchBufferSize,
_In_ UINT64 StreamCount

Note there are pointers to the GPU virtual address space, however in the IOMMU virtual memory model used by Intel integrated graphics drivers, virtual address space is shared with (and is actually managed by) the host CPU, because only system memory is physically present.
 
Last edited:
EXECUTION looks like a descriptor for a decompression task, doesn't it? Curious way to turn the decompression algorithms into a guaranteed proprietary blackbox you can't even dump or debug into.

The actual storage access part of DirectStorage shouldn't show end up in the meta commands for any vendor in the foreseeable future though, that ain't GPU controlled.
 
Isn't it a bit early to presume that the Execute Metacommand is the only way that decompression is executed?

With so much on GPUs being proprietary black boxes, why is decompression being yet another one a problem?
 
Jawed said:
With so much on GPUs being proprietary black boxes, why is decompression being yet another one a problem?
Click to expand...
Too many algorithms to choose from, and a good chance that at least one IHV will screw up their black blox algorithm for the compression scheme you would like to target. And yet as the author of the assets, you do have to pick one specific format. It's just yet another chance to catch the backfire in place of the IHVs. And if you think "in that case I will roll my own decompression algorithm" - sure, but will you get the budget granted to do that? Especially if you think more complex/efficient patterns, like e.g. WebP to BC5 transcoding?

Not to call any names, but will you take the fall when special IHV N will tell you "we just added support for format X from feature level Y on our new lineup, but screw performance on the old one"? Or IHV I tells you "oh, that's just a functional stub, not intended for production use. But if you pay extra we can implement it efficiently for you". Or will you stick with the safest, least efficient choice? Or even accept fallback to CPU decoding?

You are right, it's not the first blackbox. But it's yet another one which is going to be costly to work around when needed. And you are assuming fast(er) asset streaming, you are going to cry in pain when you find you don't get it on one platform even though the algorithm would had been portable with ease.
 
Last edited:
You must log in or register to reply here.
Back
Top