Bottlenecks analysis_shaderALU_bound name Shader ALU Bound description is shader ALU bound analysis_TextureSamplingInst_bound name Texture Sampling Instructions Bound description is shader texture sampling instruction bound analysis_pixelRate_bound name Pixel Rate Bound description is pixel rate bound analysis_primitiveRate_bound name Primitive Rate Bound description is primitive rate bound analysis_memory_bound name Memory Bandwidth Bound description is memory bandwidth bound Version 1.0 Instruments DefaultSamplingInterval 500 Profiles Limiters1 DerivedCounters Name VSOccupancy Name PSOccupancy Name CSOccupancy Name LLCRead Name LLCWrite Name ALUutilization DerivedCounters VSOccupancy counters DerivedVSOccupancy DerivedCPF name Vertex Shader Occupancy description Vertex Shader occupancy type Percentage PSOccupancy counters DerivedPSOccupancy DerivedCPF name Fragment Shader Occupancy description Fragment Shader occupancy type Percentage CSOccupancy counters DerivedCSOccupancy DerivedCSNOccupancy DerivedCPF name Compute Shader Occupancy description Compute Shader occupancy type Percentage LLCRead counters DerivedLLCRead GpuTime name Last Level Cache Read description Last Level Cache Read # of Bytes type Rate LLCWrite counters DerivedLLCWrite GpuTime name Last Level Cache Written description Last Level Cache Write # of Bytes type Rate ALUutilization counters DerivedWavesInFlight DerivedALUActiveInst DerivedALUThreadCycles GpuTime name ALUutilization description Utilization of instructions in SIMD type Rate GPUTime counters MTLStat_nSec name GPU Time description GPU Time in nSec type Count GPUBusy counters AMDStat_GPU_Engine_Ticks AMDStat_GPU_Engine_Busy_Ticks name GPU Busy % description Percentage of time GPU is doing work type Percentage GPUStall counters GPUBusy name GPU Stall % description Percentage of time GPU is being stalled type Percentage ShaderCoreUtilization counters MTLStatShaderCost MTLStatComputeCost name Shader Core Utilization % description Percentage of time Shader core is doing work type Percentage ShaderCoreALUActive counters MTLStatComputeCost AMDStat_Shader_SALU_Cycles AMDStat_Shader_VALU_Cycles AMDStat_Compute_SALU_Insts AMDStat_Compute_VALU_Inst_Cycles AMDStat_GPU_Engine_Busy_Ticks name Shader Core ALU Active % description Percentage of time Shader core is doing actual work (ALU instructions) in total GPU time type Percentage ShaderCoreVertexALUActive counters AMDStat_Vertex_SALU_Insts AMDStat_Vertex_VALU_Inst_Cycles AMDStat_GPU_Engine_Busy_Ticks name ALU active % in Vertex Shader description Percentage of time shader core is running ALU instructions in vertex shader in total GPU time type Percentage ShaderCoreFragmentALUActive counters AMDStat_Fragment_SALU_Insts AMDStat_Fragment_VALU_Inst_Cycles AMDStat_GPU_Engine_Busy_Ticks name ALU active % in Fragment Shader description Percentage of time shader core is running ALU instructions in fragment shader in total GPU time type Percentage MemoryStallPerWavefrontCompute counters AMDStat_Compute_Wait_Inst_VMEM AMDStat_Compute_WAIT_CNT_VMVS AMDStat_Compute_Waves_Executed AverageWavefrontLatencyCompute name Memory Stall % in Compute Shader Wavefronts description Percentage of time a compute shader wavefront is stalled by memory operation type Percentage MemoryStallPerWavefrontVertex counters AMDStat_Vertex_Wait_Inst_VMEM AMDStat_Vertex_WAIT_CNT_VMVS AMDStat_Vertex_Waves_Executed AverageWavefrontLatencyVertex name Memory Stall % in Vertex Shader Wavefronts description Percentage of time a vertex shader wavefront is stalled by memory operation type Percentage MemoryStallPerWavefrontFragment counters AMDStat_Fragment_Wait_Inst_VMEM AMDStat_Fragment_WAIT_CNT_VMVS AMDStat_Fragment_Waves_Executed AverageWavefrontLatencyFragment name Memory Stall % in Fragment Shader Wavefronts description Percentage of time a fragment shader wavefront is stalled by memory operation type Percentage AverageWavefrontLatencyCompute counters AMDStat_Compute_Accum_Prev_Cycles AMDStat_Compute_Waves_Executed name Average latency of a Compute Shader Wavefront description Average cycles each Compute shader wavefront takes type Count AverageWavefrontLatencyVertex counters AMDStat_Vertex_Accum_Prev_Cycles AMDStat_Vertex_Waves_Executed name Average latency of a Vertex Shader Wavefront description Average cycles each vertex shader wavefront takes type Count AverageWavefrontLatencyFragment counters AMDStat_Fragment_Accum_Prev_Cycles AMDStat_Fragment_Waves_Executed name Average latency of a Fragment Shader Wavefront description Average cycles each fragment shader wavefront takes type Count AverageWavesInflightCompute counters AMDStat_Compute_Accum_Prev_Cycles AMDStat_Compute_Busy_Cycles name Compute Shader Wavefront SIMD Occupancy description Average Number of Compute Shader Wavefront In Flight per SIMD type Count AverageWavesInflightVertex counters AMDStat_Vertex_Accum_Prev_Cycles AMDStat_Vertex_Busy_Cycles name Vertex Shader Wavefront SIMD Occupancy description Average Number of Vertex Shader Wavefront In Flight per SIMD type Count AverageWavesInflightFragment counters AMDStat_Fragment_Accum_Prev_Cycles AMDStat_Fragment_Busy_Cycles name Fragment Shader Wavefront SIMD Occupancy description Average Number of Fragment Shader Wavefront In Flight per SIMD type Count ShaderCoreStall counters ShaderCoreALUActive AMDStat_GPU_Engine_Busy_Ticks AMDStat_Shader_VMEM_Read_Cycles AMDStat_Shader_VMEM_Write_Cycles AMDStat_Shader_EXP_Cycles AMDStat_Shader_GDS_Cycles MTLStatShaderCost name % of Stall Time in Shader Core in total GPU time description Percentage of time shader core is being stalled in total GPU time type Percentage SamplerBusy counters AMDStat_Sampler_Busy name % of Busy Time in Sampler description Percentage of time Sampler is busy type Percentage L2CacheThroughput counters AMDStat_L2_Cache_Request AMDStat_L2_Busy AMDStat_GPU_Engine_Busy_Ticks name L2 Cache Throughput description Number of requests (cache lines) being sent to L2 Cache per GPU clock type Rate L2CacheDramBandwidth counters AMDStat_L2_Write_Size AMDStat_L2_Read_Size AMDStat_L2_Busy AMDStat_GPU_Engine_Busy_Ticks name L2 Cache Dram Bandwidth description Number of GB being accessed in memory by L2 cache per second type Rate VSInvocation counters AMDStat_DS_Invocations AMDStat_VS_Invocations name Vertex Shader Invocations description Number of times vertex shader is invoked type Count CSInvocation counters AMDStat_CS_Invocations AMDStat_VS_Invocations AMDStat_PS_Invocations name Compute Shader Invocations description Number of times compute shader is invoked type Count PSInvocation counters AMDStat_PS_Invocations AMDStat_VS_Invocations name Fragment Shader Invocations description Number of times fragment shader is invoked type Count VertexCost counters AMDStat_Vertex_Accum_Prev_Cycles AMDStat_Fragment_Accum_Prev_Cycles name Vertex Shader Cost description % of total compute unit cycles spend running vertex shader type Percentage FragmentCost counters AMDStat_Vertex_Accum_Prev_Cycles AMDStat_Fragment_Accum_Prev_Cycles name Fragment Shader Cost description % of total compute unit cycles spend running fragment shader type Percentage VertexDuration counters MTLStatFragmentCost MTLStatVertexCost name Vertex Shader Duration description % of GPU busy time the vertex shader is in flight type Percentage FragmentDuration counters MTLStatFragmentCost MTLStatVertexCost name Fragment Shader Duration description % of GPU busy time the fragment shader is in flight type Percentage PixelToVertexRatio counters AMDStat_GE_Vertices PixelsRasterized name Pixel To Vertex Ratio description Number of pixels per vertex type Rate PixelPerTriangle counters PrimitivesSubmitted PixelsRasterized name Pixel Per Triangle description Number of pixels per triangle type Rate VerticesSubmitted counters AMDStat_HS_Invocations AMDStat_GE_Vertices AMDStat_VS_Vertices_In name Vertices Submitted description Number of vertices being submitted to input assembly type Count NumberOfPatchesSubmitted counters AMDStat_HS_Invocations AMDStat_GE_Vertices name Number of Patches Submitted description Number of patches being submitted to input assembly type Count VerticesReused counters AMDStat_Vertices_Reuse AMDStat_GE_Vertices AMDStat_VS_Invocations name Number of Vertices Reused description Number of number of vertices being reused in the total number of submitted vertices type Count VerticesReusedPercentage counters VerticesReused AMDStat_GE_Vertices name Percentage of Vertices Reused description Percentage of number of vertices being reused in the total number of submitted vertices type Percentage AverageTessFactor counters AMDStat_HS_Invocations AMDStat_VS_Vertices_In AMDStat_GE_Vertices name Average Tessellation Factor description Average of actually calculated tessellation factor type Count VerticesRendered counters AMDStat_VS_Vertices_In name Number of Vertices Rendered description Number of number of vertices being passed to shader core in total number of vertices submitted type Count VerticesRenderedPercentage counters AMDStat_VS_Vertices_In VerticesSubmitted name Percentage of Vertices Rendered description Percentage of number of vertices being passed to shader core in total number of vertices submitted type Percentage PrimitivesSubmitted counters AMDStat_HS_Invocations AMDStat_C_Invocations AMDStat_GE_Primitives name Primitives Submitted description Number of primitives gathered by input assembly type Count PrimitivesRendered counters AMDStat_C_Primitives name Number of Primitives Rendered description Number of Number of primitives actually rendered in total number of primitives submitted type Count PrimitivesRenderedPercentage counters AMDStat_C_Primitives PrimitivesSubmitted name Percentage of Primitives Rendered description Percentage of Number of primitives actually rendered in total number of primitives submitted type Percentage NumberOfCulledPrimitives counters AMDStat_Culled_Prims name Number of Primitives Culled description Number of Number of primitives being culled in total number of primitives submitted type Count NumberOfCulledPrimitivesPercentage counters AMDStat_Culled_Prims PrimitivesSubmitted name Percentage of Primitives Culled description Percentage of Number of primitives being culled in total number of primitives submitted type Percentage NumberOfClippedPrimitives counters AMDStat_Clipped_Prims name Number of Primitives Clipped description Number of Number of primitives being clipped in total number of primitives submitted type Count NumberOfClippedPrimitivesPercentage counters AMDStat_Clipped_Prims PrimitivesSubmitted name Percentage of Primitives Clipped description Percentage of Number of primitives being clipped in total number of primitives submitted type Percentage HierarchicalZTotalTilesCount counters AMDStat_HiZTiles_Total_Count name Hierarchical ZTotal Tiles Count description Number of tiles sent to hierarchical Z test type Count HierarchicalZFailPercentage counters AMDStat_HiZTiles_Culled_Count AMDStat_HiZTiles_Total_Count name Percentage of Hierarchical Z Fail Count description Percentage of number of tiles that fails hierarchical Z test in the total number of tiles sent to HiZ test type Percentage PreZPassSampleCount counters AMDStat_PreZ_Samples_PassingZ name PreZ Z test Pass Sample Count description Number of samples that pass preZ Z test type Count PreZFailSampleCount counters AMDStat_PreZ_Samples_FailingZ name PreZ Z test Fail Count description Number of samples that fail preZ Z test type Count PreZStencilFailSampleCount counters AMDStat_PreZ_Samples_FailingS name PreZ Stencil Test Fail SampleCount description Number of samples that fail preZ stencil test type Count PostZPassSampleCount counters AMDStat_PostZ_Samples_PassingZ name PostZ Z test Pass Sample Count description Number of samples that pass postZ Z test type Count PostZFailSampleCount counters AMDStat_PostZ_Samples_FailingZ name PostZ Z test Fail Sample Count description Number of samples that fail postZ Z test type Count PostZStencilFailSampleCount counters AMDStat_PostZ_Samples_FailingS name PostZ Stencil test Fail Sample Count description Number of samples that fail postZ stencil test type Count PreZFailPercentage counters AMDStat_PreZ_Quads_Count AMDStat_FS_Quads_Count AMDStat_PreHiZ_Total_Quads_Count name Fail % in PreZ Z and Stencil test description Percentage of pixels that failed preZ Z and stencil test in total number of rasterized pixels type Percentage PostZFailPercentage counters AMDStat_PostZ_Samples_FailingZ AMDStat_PostZ_Samples_FailingS AMDStat_PostZ_Samples_PassingZ AMDStat_PreZ_Quads_Count AMDStat_PostFS_Pixels AMDStat_PreHiZ_Total_Quads_Count name Fail % in PostZ Z and Stencil test description Percentage of pixels that failed postZ Z and stencil test in total number of rasterized pixels type Percentage PixelsRasterized counters AMDStat_PreHiZ_Total_Quads_Count name Pixels Rasterized description Total number of pixels rasterized type Count PixelsDrawn counters AMDStat_Pixels_Drawn NumMRT name Pixels Drawn description Number of pixels written to color buffer type Count PixelsDrawnPercentage counters NumMRT AMDStat_Quads_Drawn AMDStat_PreHiZ_Total_Quads_Count name % of Pixels Drawn description Percentage of number of pixels written to color buffer in total number of rasterized pixels type Percentage FragmentsDrawn counters PixelsDrawn AMDStat_Quads_Drawn AMDStat_Quad_Fragments_Drawn NumMRT name Fragments Drawn description Number of fragments written to color buffer type Count PixelsDiscardedPercentage counters AMDStat_PS_Invocations AMDStat_PostFS_Pixels AMDStat_PreHiZ_Total_Quads_Count name % of Pixels Discarded in Shader description Percentage of number of pixels being discarded in shader in total number of pixels rasterized type Percentage PixelRate counters AMDStat_Pixels_Drawn AMDStat_CBDrawn_Busy AMDStat_GPU_Engine_Busy_Ticks name Pixel Rate description Number of pixels being rendered per GPU clock type Rate ALUInstructionPerInvocationCompute counters AMDStat_Compute_Waves_Executed AMDStat_Compute_SALU_Insts AMDStat_Compute_VALU_Insts name ALU Instructions Executed Per Compute Shader Invocation description Total number of ALU instructions executed in each compute shader invocation type Count ALUInstructionPerInvocationVertex counters AMDStat_Vertex_SALU_Insts AMDStat_Vertex_Waves_Executed AMDStat_Vertex_VALU_Insts name ALU Instructions Executed Per Vertex Shader Invocation description Total number of ALU instructions executed in each vertex shader invocation type Count ALUInstructionPerInvocationFragment counters AMDStat_Fragment_Waves_Executed AMDStat_Fragment_SALU_Insts AMDStat_Fragment_VALU_Insts name ALU Instructions Executed Per Fragment Shader Invocation description Total number of ALU instructions executed in each fragment shader invocation type Count MemInstructionPerInvocationCompute counters AMDStat_Compute_Waves_Executed AMDStat_Compute_VMEM_Write_Cycles AMDStat_Compute_VMEM_Read_Cycles AMDStat_Compute_SMEM_Insts name Memory Instructions Executed Per Compute Shader Invocation description Total number of memory instructions executed in each compute shader invocation type Count MemInstructionPerInvocationVertex counters AMDStat_Vertex_VMEM_Write_Cycles AMDStat_Vertex_Waves_Executed AMDStat_Vertex_VMEM_Read_Cycles AMDStat_Vertex_SMEM_Insts name Memory Instructions Executed Per Vertex Shader Invocation description Total number of memory instructions executed in each vertex shader invocation type Count MemInstructionPerInvocationFragment counters AMDStat_Fragment_Waves_Executed AMDStat_Fragment_VMEM_Write_Cycles AMDStat_Fragment_VMEM_Read_Cycles AMDStat_Fragment_SMEM_Insts name Memory Instructions Executed Per Fragment Shader Invocation description Total number of memory Iinstructions executed in each fragment shader invocation type Count ControlInstructionPerInvocationCompute counters AMDStat_Compute_Waves_Executed AMDStat_Compute_Branch_Insts name Control Instructions Executed Per Compute Shader Invocation description Total number of control instructions executed in each compute shader invocation type Count ControlInstructionPerInvocationVertex counters AMDStat_Vertex_Waves_Executed AMDStat_Vertex_Branch_Insts name Control Instructions Executed Per Vertex Shader Invocation description Total number of control instructions executed in each vertex shader invocation type Count ControlInstructionPerInvocationFragment counters AMDStat_Fragment_Waves_Executed AMDStat_Fragment_Branch_Insts name Control Instructions Executed Per Fragment Shader Invocation description Total number of control instructions executed in each fragment shader invocation type Count ALUToMemRatioCompute counters ALUInstructionPerInvocationCompute MemInstructionPerInvocationCompute name ALU to Memory Ratio in Compute Shader description Number of ALU instructions per memory instruction in compute shader type Rate ALUToMemRatioVertex counters ALUInstructionPerInvocationVertex MemInstructionPerInvocationVertex name ALU to Memory Ratio in Vertex Shader description Number of ALU instructions per memory instruction in vertex shader type Rate ALUToMemRatioFragment counters ALUInstructionPerInvocationFragment MemInstructionPerInvocationFragment name ALU to Memory Ratio in Fragment Shader description Number of ALU instructions per memory instruction in fragment shader type Rate LdsStallPerWavefrontCompute counters AMDStat_Compute_WAIT_CNT_LGKM AMDStat_Compute_LDS_Wait AMDStat_Compute_Waves_Executed AverageWavefrontLatencyCompute name % of LDS Stall in a Compute Wavefront description Percentage of a compute shader wavefront being stalled by LDS access type Percentage LdsStallPerWavefrontVertex counters AMDStat_Vertex_LDS_Wait AMDStat_Vertex_WAIT_CNT_LGKM AMDStat_Vertex_Waves_Executed AverageWavefrontLatencyVertex name % of LDS Stall in a Vertex Wavefront description Percentage of a vertex shader wavefront being stalled by LDS access type Percentage LdsStallPerWavefrontFragment counters AMDStat_Fragment_LDS_Wait AMDStat_Fragment_WAIT_CNT_LGKM AMDStat_Fragment_Waves_Executed AverageWavefrontLatencyFragment name % of LDS Stall in a Fragment Wavefront description Percentage of a fragment wavefront being stalled by LDS access type Percentage ROPStall counters AMDStat_ROP_Stalled_Ticks AMDStat_GPU_Engine_Busy_Ticks name ROP Stall description Percentage of GPU busy time being stalled by ROP type Percentage ExportStallPerWavefrontVertex counters AMDStat_Vertex_Wait_Export_Alloc AMDStat_Vertex_WAIT_CNT_EXP AMDStat_Vertex_Waves_Executed AverageWavefrontLatencyVertex name Export Stall % in Vertex Shader Wavefronts description Percentage of time a vertex shader wavefront is stalled by export instructions type Percentage ExportStallPerWavefrontFragment counters AMDStat_Fragment_Wait_Export_Alloc AMDStat_Fragment_WAIT_CNT_EXP AMDStat_Fragment_Waves_Executed AverageWavefrontLatencyFragment name Export Stall % in Fragment Shader Wavefronts description Percentage of time a fragment shader wavefront is stalled by export instructions type Percentage ZeroAreaCullPrims counters AMDStat_ZeroArea_Culled_Prims name Number of Zero Area Culled Primitives description Number of number of primitives being culled due to having zero area coverage in total number of primitives submitted type Count ZeroAreaCullPrimsPercentage counters AMDStat_ZeroArea_Culled_Prims PrimitivesSubmitted name Percentage of Zero Area Culled Primitives description Percentage of number of primitives being culled due to having zero area coverage in total number of primitives submitted type Percentage ClipperCullPrims counters AMDStat_Clipping_Culled_Prims name Number of Clipper Culled Primitives description Number of number of primitives being culled by clipper in total number of primitives submitted type Count ClipperCullPrimsPercentage counters AMDStat_Clipping_Culled_Prims PrimitivesSubmitted name Percentage of Clipper Culled Primitives description Percentage of number of primitives being culled by clipper in total number of primitives submitted type Percentage MemUnitBusy counters AMDStat_TexUnit_Busy AMDStat_GPU_Engine_Busy_Ticks name Memory Unit Stall description Percentage of GPUTime the memory unit is active type Percentage MemUnitBusyCycles counters AMDStat_TexUnit_Busy_Ticks name Memory Unit Busy Cycles description Number of GPU cycles the memory unit is active type Count MemUnitStalled counters AMDStat_TexUnit_Stall AMDStat_GPU_Engine_Busy_Ticks name Memory Unit Stall Cycles description The percentage of GPUTime the memory unit is stalled type Percentage MemUnitStalledCycles counters AMDStat_TexUnit_Stalled_Ticks name Memory Unit Stall Cycles description Number of GPU cycles the memory unit is stalled type Count WriteUnitStalled counters AMDStat_L2_Write_Unit_Stalled AMDStat_GPU_Engine_Busy_Ticks name Write Unit Stall Cycles description The percentage of GPUTime the write unit is stalled type Percentage WriteUnitStalledCycles counters AMDStat_L2_Write_Unit_Stalled name Write Unit Stall Cycles description Number of GPU cycles the write unit is stalled type Count LocalVidMemBytes counters AMDStat_LocalVidMemBytes name Write Unit Stall Cycles description Number of GPU cycles the write unit is stalled type Count XGMIBytes counters AMDStat_XGMIBytes name Write Unit Stall Cycles description Number of GPU cycles the write unit is stalled type Count PcieBytes counters AMDStat_PcieBytes name PCIEbytes description Number of GPU cycles the write unit is stalled type Count TextureUnitStall counters AMDStat_TexUnit_Stall name Texture Unit Stall description Percentage of GPU busy time the texture unit is stalled type Percentage TextureUnitBusy counters AMDStat_TexUnit_Busy name Texture Unit Busy description Percentage of GPU busy time the texture unit is busy type Percentage TextureCacheMissRate counters AMDStat_TextureCache_Hit AMDStat_TextureCache_Miss name Texture Cache Miss Rate description Percentage of time Texture Cache access is a Miss type Percentage L0CacheMissRate counters AMDStat_L0_Cache_Request AMDStat_L0_Cache_Miss name L0 Cache Miss Rate description Percentage of time L0 Cache access is a Miss type Percentage L0CacheHitRate counters AMDStat_L0_Cache_Request AMDStat_L0_Cache_Miss name L0 Cache Hit Rate description Percentage of time L0 Cache access is a Hit type Percentage L0CacheHitCount counters AMDStat_L0_Cache_Request AMDStat_L0_Cache_Miss name L0 Cache Hit Count description Count of L0 Cache Hits type Count L0CacheMissCount counters AMDStat_L0_Cache_Miss name L0 Cache Miss Count description Count of L0 Cache Misses type Count L1CacheMissRate counters AMDStat_L1_Cache_Miss AMDStat_L1_Cache_Request name L1 Cache Miss Rate description Percentage of time L1 Cache access is a Miss type Percentage L1CacheHitRate counters AMDStat_L1_Cache_Miss AMDStat_L1_Cache_Request name L1 Cache Hit Rate description Percentage of time L1 Cache access is a Hit type Percentage L1CacheHitCount counters AMDStat_L1_Cache_Miss AMDStat_L1_Cache_Request name L1 Cache Hit Count description L1 Cache Hit Count type Count L1CacheMissCount counters AMDStat_L1_Cache_Miss name L1 Cache Miss Count description L1 Cache Miss Count type Count L2CacheHitRate counters AMDStat_L2_Cache_Miss AMDStat_L2_Cache_Request name L2 Cache Hit Rate description Percentage of time L2 Cache access is a Hit type Percentage L2CacheMissRate counters AMDStat_L2_Cache_Miss AMDStat_L2_Cache_Request name L2 Cache Miss Rate description Percentage of time L2 Cache access is a Miss type Percentage L2CacheHitCount counters AMDStat_L2_Cache_Miss AMDStat_L2_Cache_Request name L2 Cache Hit Count description Count of L2 Cache access is a Hit type Count L2CacheMissCount counters AMDStat_L2_Cache_Miss name L2 Cache Miss Count description Count of L2 Cache access is a Miss type Count FetchSize counters AMDStat_Tex_Fetch_Size_32B AMDStat_Tex_Fetch_Size_64B AMDStat_Tex_Fetch_Size_96B AMDStat_Tex_Fetch_Size_128B name FetchSize description Number of total bytes fetched from the video memory type Count WriteSize counters AMDStat_L2_Write_Size AMDStat_Tex_Write_Size name WriteSize description Number of total bytes written to the video memory type Count VertexMemoryFetchLatency counters AMDStat_GE_MemRead_Bin0 AMDStat_GE_MemRead_Bin1 AMDStat_GE_MemRead_Bin2 AMDStat_GE_MemRead_Bin3 AMDStat_GE_MemRead_Bin4 AMDStat_GE_MemRead_Bin5 AMDStat_GE_MemRead_Bin6 AMDStat_GE_MemRead_Bin7 name Vertex Memory Fetch Latency description Average latency of vertex memory fetch in number of GPU clocks type Count NumMRT counters AMDStat_PostFS_Pixels AMDStat_PostFS_Exports name Number of MRT description Number of MRT being used type Count PrimitiveRate counters AMDStat_FrontEnd_Busy_SE0PA0 AMDStat_FrontEnd_Busy_SE0PA1 AMDStat_FrontEnd_Busy_SE1PA0 AMDStat_FrontEnd_Busy_SE1PA1 AMDStat_FrontEnd_Busy_SE2PA0 AMDStat_FrontEnd_Busy_SE2PA1 AMDStat_FrontEnd_Busy_SE3PA0 AMDStat_FrontEnd_Busy_SE3PA1 AMDStat_FrontEnd_Stall_SE0PA0 AMDStat_FrontEnd_Stall_SE0PA1 AMDStat_FrontEnd_Stall_SE1PA0 AMDStat_FrontEnd_Stall_SE1PA1 AMDStat_FrontEnd_Stall_SE2PA0 AMDStat_FrontEnd_Stall_SE2PA1 AMDStat_FrontEnd_Stall_SE3PA0 AMDStat_FrontEnd_Stall_SE3PA1 AMDStat_FrontEnd_StarvedBusy_SE0PA0 AMDStat_FrontEnd_StarvedBusy_SE0PA1 AMDStat_FrontEnd_StarvedBusy_SE1PA0 AMDStat_FrontEnd_StarvedBusy_SE1PA1 AMDStat_FrontEnd_StarvedBusy_SE2PA0 AMDStat_FrontEnd_StarvedBusy_SE2PA1 AMDStat_FrontEnd_StarvedBusy_SE3PA0 AMDStat_FrontEnd_StarvedBusy_SE3PA1 name Primitive Rate description Number of primitives processed per GPU clock type Rate VertexRate counters AMDStat_VS_Vertices_In AMDStat_VS_Stalled_Ticks AMDStat_VS_StarvedBusy_Ticks name Vertex Rate description Number of vertices processed per GPU clock during vertex pipe busy time type Rate PeakPrimitiveRate name Peak Primitive Rate per Clock description Peak primitive rate per GPU clock supported by this GPU. To reach peak primitive rate you have to have perfect vertex reuse. type Rate PeakVertexRate name Peak Vertex Rate per Clock description Peak vertex rate per GPU clock supported by this GPU type Rate PeakPixelRate name Peak Pixel Rate per Clock description Peak pixel rate per GPU clock supported by this GPU type Rate VertexRatePercentage counters PeakVertexRate VertexRate name Vertex Rate Percentage description Percentage of actual vertex rate in peak vertex rate type Percentage PrimitiveRatePercentage counters PeakPrimitiveRate PrimitiveRate name Primitive Rate Percentage description Percentage of actual primitiveR rate in peak primitive rate type Percentage PixelRatePercentage counters PeakPixelRate PixelRate name Pixel Rate Percentage description Percentage of actual pixel rate in peak pixel rate type Percentage ParameterCacheStall counters AMDStat_ParameterCache_Stall name % of GPU time stalled by Full Parameter Cache description Percentage of GPU time VS workload is stalled due to parameter cache space. This can also be caused by backend not releasing parameter cache space fast enough. type Percentage ShaderCoreVertexStall counters AMDStat_Vertex_SALU_Insts AMDStat_Vertex_VALU_Inst_Cycles AverageThreadsPerVertexWavefront AMDStat_Vertex_VMEM_Read_Cycles AMDStat_Vertex_VMEM_Write_Cycles AMDStat_Vertex_VMEM_Inst_Cycles AMDStat_Vertex_EXP_Insts AMDStat_GPU_Engine_Busy_Ticks MTLStatVertexCost name Shader Core Stall % in Vertex Shader description Percentage of time shader core is being stalled in vertex shader in total GPU time type Percentage ShaderCoreFragmentStall counters AMDStat_Fragment_SALU_Insts AMDStat_Fragment_VALU_Inst_Cycles AverageThreadsPerFragmentWavefront AMDStat_Fragment_VMEM_Inst_Cycles AMDStat_Fragment_VMEM_Read_Cycles AMDStat_Fragment_VMEM_Write_Cycles AMDStat_Fragment_SMEM_Insts AMDStat_Fragment_EXP_Insts AMDStat_GPU_Engine_Busy_Ticks MTLStatFragmentCost name Shader Core Stall % in Fragment Shader description Percentage of time shader core is being stalled in fragment shader in total GPU time type Percentage TessellatorBusy counters AMDStat_Tessellator_Busy name Tessellator Busy description Percentage of GPU busy time the tessellator is busy type Percentage CBDrawnBusyRate counters AMDStat_CBDrawn_Busy AMDStat_CB_Busy name Color Buffer Draw to Busy Rate description Percentage of time color buffer is busy drawing to total color buffer busy time type Percentage ColorBlockL2Throughput counters AMDStat_CBDrawn_Busy AMDStat_CB_Mem_Read AMDStat_CB_Mem_Written name Traffic send out by Color Buffer To L2 Cache description Bytes of cache access made by color buffer each clock when color buffer is busy type Rate CBMemRead counters AMDStat_CB_Mem_Read name Color Buffer Memory Read description Number of bytes read from the color buffer. type Count CBColorAndMaskRead counters AMDStat_CB_Mem_Read AMDStat_CB_FC_MC_DCC_Mem_Read AMDStat_CB_CM_Mem_Read AMDStat_CB_FC_Mem_Read name Color Buffer and Mask buffer Memory Read description Total number of bytes read from the color and mask buffers. type Count CBMemWritten counters AMDStat_CB_Mem_Written name Color Buffer Memory Written description Number of bytes written to the color buffer. type Count CBColorAndMaskWritten counters AMDStat_CB_Mem_Written AMDStat_CB_FC_MC_DCC_Mem_Written AMDStat_CB_CM_Mem_Written AMDStat_CB_FC_Mem_Written name Color Buffer and Mask buffer Memory Written description Total number of bytes written to the color and mask buffers. type Count CBSlowPixelPct counters AMDStat_Quad_Fragments_32ABGR_Drawn AMDStat_Quad_Fragments_Drawn name Color Buffer slow draws description Percentage of pixels written to the color buffer using a half-rate or quarter-rate format. type Percentage CBSlowPixelCount counters AMDStat_Quad_Fragments_32ABGR_Drawn name Color Buffer slow draws pixel count description Number of pixels written to the color buffer using a half-rate or quarter-rate format type Count DepthBlockL2Throughput counters AMDStat_ZUnit_Busy AMDStat_DB_Tile_Read AMDStat_DB_Quad_Read AMDStat_DB_Tile_Write AMDStat_DB_Quad_Write name Traffic send out by Depth Buffer To L2 Cache description Bytes of cache access made by depth buffer each clock when Depth buffer is busy type Rate ZUnitStalled counters AMDStat_ZUnit_Stalled AMDStat_GPU_Engine_Busy_Ticks name ZUnit Stall rate description The percentage of GPUTime the depth buffer spends waiting for the color buffer to be ready to accept data. High figures here indicate a bottleneck in color buffer operations. type Percentage ZUnitStalledCycles counters AMDStat_ZUnit_Stalled_Ticks name ZUnit Stall cycles description Number of GPU cycles the depth buffer spends waiting for the color buffer to be ready to accept data. Larger numbers indicate a bottleneck in color buffer operations. type Count DBMemRead counters AMDStat_DB_Tile_Read AMDStat_DB_Quad_Read name Depth Buffer Memory Read description Number of bytes read from the Depth buffer. type Count DBMemWritten counters AMDStat_DB_Tile_Write AMDStat_DB_Quad_Write name Depth Buffer Memory Written description Number of bytes written to the Depth buffer. type Count TexTriFilteringPct counters AMDStat_TexUnit_TriFilteringCount AMDStat_TexUnit_NoTriFilteringCount name Tri Filtering Tex description Percentage of pixels that received trilinear filtering. Note that not all pixels for which trilinear filtering is enabled will receive it (e.g. if the texture is magnified) type Percentage TexTriFilteringCount counters AMDStat_TexUnit_TriFilteringCount name Trilinear Filtering Count of pixels description Count of pixels that received trilinear filtering. Note that not all pixels for which trilinear filtering is enabled will receive it (e.g. if the texture is magnified). type Count NoTexTriFilteringCount counters AMDStat_TexUnit_NoTriFilteringCount name Non Trilinear Filtering Count of pixels description Count of pixels that did not receive trilinear filtering. type Count TexVolFilteringPct counters AMDStat_TexUnit_TexVolFilteringCount AMDStat_TexUnit_NoTexVolFilteringCount name Volume Filtered Pixel Percentage description Percentage of pixels that received volume filtering. type Percentage TexVolFilteringCount counters AMDStat_TexUnit_TexVolFilteringCount name Volume Filtering Count of pixels description Count of pixels that received volume filtering. type Count NoTexVolFilteringCount counters AMDStat_TexUnit_NoTexVolFilteringCount name Non Volume Filtering Count of pixels description Count of pixels that did not receive volume filtering. type Count TexAveAnisotropy counters AMDStat_TexUnit_Average_Anisotropy_1 AMDStat_TexUnit_Average_Anisotropy_2 AMDStat_TexUnit_Average_Anisotropy_4 AMDStat_TexUnit_Average_Anisotropy_6 AMDStat_TexUnit_Average_Anisotropy_8 AMDStat_TexUnit_Average_Anisotropy_10 AMDStat_TexUnit_Average_Anisotropy_12 AMDStat_TexUnit_Average_Anisotropy_14 AMDStat_TexUnit_Average_Anisotropy_16 name Average Anisotropy description The average degree of anisotropy applied. A number between 1 and 16. The anisotropic filtering algorithm only applies samples where they are required (e.g. there will be no extra anisotropic samples if the view vector is perpendicular to the surface) so this can be much lower than the requested anisotropy. type Percentage CSVFetchInsts counters AMDStat_Compute_Tex_Load_Inst AMDStat_Compute_Waves_Executed name Average vector fetch instructions description The average number of vector fetch instructions from the video memory executed per work-item (affected by flow control). type Percentage CSSFetchInsts counters AMDStat_Compute_SMEM_Insts AMDStat_Compute_Waves_Executed name Average scalar fetch instructions description The average number of scalar fetch instructions from the video memory executed per work-item (affected by flow control). type Percentage CSVWriteInsts counters AMDStat_Compute_Tex_Store_Inst AMDStat_Compute_Waves_Executed name Average vector fetch instructions description The average number of vector fetch instructions from the video memory executed per work-item (affected by flow control). type Percentage CSVALUBusy counters AMDStat_Compute_VALU_Inst_Cycles AMDStat_GPU_Engine_Busy_Ticks name Percentage of CS vector ALU instructions description The percentage of GPUTime vector ALU instructions are processed. Value range: 0% (bad) to 100% (optimal). type Percentage CSVALUBusyCycles counters AMDStat_Compute_VALU_Inst_Cycles name Number of CS vector ALU instruction cycles description Number of GPU cycles where vector ALU instructions are processed. type Count CSSALUBusy counters AMDStat_Compute_SALU_Insts AMDStat_GPU_Engine_Busy_Ticks name Percentage of CS scalar ALU instructions description The percentage of GPUTime scalar ALU instructions are processed. Value range: 0% (bad) to 100% (optimal). type Percentage CSSALUBusyCycles counters AMDStat_Compute_SALU_Insts name Number of CS scalar ALU instruction cycles description Number of GPU cycles where scalar ALU instructions are processed type Count CSMemUnitBusy counters AMDStat_TexUnit_Busy AMDStat_CSG_number_of_threadGrps AMDStat_CSN_number_of_threadGrps AMDStat_GPU_Engine_Busy_Ticks name CS Memory Unit Busy description The percentage of GPUTime the memory unit is active. The result includes the stall time (MemUnitStalled). This is measured with all extra fetches and writes and any cache or memory effects taken into account. Value range: 0% to 100% (fetch-bound). type Percentage CSMemUnitBusyCycles counters AMDStat_TexUnit_Busy_Ticks AMDStat_CSG_number_of_threadGrps AMDStat_CSN_number_of_threadGrps name CS Memory Unit Busy Cycles description Number of GPU cycles the memory unit is active type Count CSMemUnitStalled counters AMDStat_L0_Stall AMDStat_CSG_number_of_threadGrps AMDStat_CSN_number_of_threadGrps AMDStat_GPU_Engine_Busy_Ticks name CS Memory Unit Stall Cycles from CS instructions description The percentage of GPUTime the memory unit is stalled. Try reducing the number or size of fetches and writes if possible. Value range: 0% (optimal) to 100% (bad). type Percentage CSMemUnitStalledCycles counters AMDStat_L0_Stall AMDStat_CSG_number_of_threadGrps AMDStat_CSN_number_of_threadGrps name CS Memory Unit Stall Cycles description Number of GPU cycles the memory unit is stalled type Count CSWriteUnitStalled counters AMDStat_L2_Write_Unit_Stalled AMDStat_CSG_number_of_threadGrps AMDStat_CSN_number_of_threadGrps AMDStat_GPU_Engine_Busy_Ticks name CS Write Unit Stall Cycles from CS instructions description The percentage of GPUTime the write unit is stalled. type Percentage CSWriteUnitStalledCycles counters AMDStat_L2_Write_Unit_Stalled AMDStat_CSG_number_of_threadGrps AMDStat_CSN_number_of_threadGrps name CS Write Unit Stall Cycles description Number of GPU cycles the write unit is stalled. type Count CSGDSInsts counters AMDStat_Compute_GDS_Insts AMDStat_Compute_Waves_Executed name Average GDS instructions description The average number of GDS read or GDS write instructions executed per work item (affected by flow control). type Percentage CSLDSInsts counters AMDStat_Compute_LDS_Insts AMDStat_Compute_Waves_Executed name Average LDS instructions description The average number of LDS read or LDS write instructions executed per work item (affected by flow control). type Percentage CSALUStalledByLDS counters AMDStat_Compute_LDS_Wait AMDStat_GPU_Engine_Busy_Ticks AMDStat_Compute_Waves_Executed name Average Stall time due to LDS instructions description The percentage of GPUTime ALU units are stalled by the LDS input queue being full or the output queue being not ready. If there are LDS bank conflicts, reduce them. Otherwise, try reducing the number of LDS accesses if possible. Value range: 0% (optimal) to 100% (bad). type Percentage CSLDSBankConflict counters AMDStat_CSG_number_of_threadGrps AMDStat_CSN_number_of_threadGrps AMDStat_GPU_Engine_Busy_Ticks AMDStat_Compute_LDS_Bank_Conflict name Average LDS Stall time due to bank conflicts description The percentage of GPUTime LDS is stalled by bank conflicts. Value range: 0% (optimal) to 100% (bad). type Percentage CSLDSBankConflictCycles counters AMDStat_CSG_number_of_threadGrps AMDStat_CSN_number_of_threadGrps AMDStat_Compute_LDS_Bank_Conflict name Number LDS Stall cycles due to bank conflicts description Number of GPU cycles the LDS is stalled by bank conflicts. Value range: 0 (optimal) to GPUBusyCycles (bad). type Count AverageThreadsPerVertexWavefront counters AMDStat_Vertex_Waves_Executed AMDStat_VS_Invocations name Average Threads Per Vertex Wavefront description Average number of active threads per vertex wavefront type Count AverageThreadsPerFragmentWavefront counters AMDStat_Fragment_Waves_Executed AMDStat_PS_Invocations name Average Threads Per Fragment Wavefront description Average number of active threads per fragment wavefront type Count AverageThreadsPerComputeWavefront counters AMDStat_Compute_Waves_Executed AMDStat_CS_Invocations name Average Threads Per Compute Wavefront description Average number of active threads per compute wavefront type Count