Skip to content

Conversation

@RuslanKutdusov
Copy link

Hi,

Recently, I decided to measure the performance of DX12 root descriptors using this project. The first step was to port it to DX12. After doing so, I noticed a performance difference between the DX11 and DX12 versions on Nvidia GPUs. Specifically, there were significant discrepancies in:

Buffer<T>.Load uniform
cbuffer loads
Texture2D<T>.Load uniform

For example, here are the timings from DX11:

Buffer<R8>.Load uniform: 0.804ms 51.368x
Buffer<R8>.Load linear: 39.710ms 1.040x
Buffer<R8>.Load random: 39.815ms 1.037x

And here from DX12:

Buffer<R8>.Load uniform: 42.198ms 0.966x
Buffer<R8>.Load linear: 42.034ms 0.970x
Buffer<R8>.Load random: 41.828ms 0.974x

As you can see, uniform access in DX12 performs similarly to linear and random access, which is quite different from DX11 behavior.

It turns out this difference occurs when shaders are compiled with DXC and Shader Model 6.x. DX12 also supports shaders compiled with FXC and Shader Model 5.1—in that case, the timings are similar to DX11:

Buffer<R8>.Load uniform: 0.768ms 52.836x
Buffer<R8>.Load linear: 39.499ms 1.027x
Buffer<R8>.Load random: 39.450ms 1.028x

I inspected the generated SASS code, and it seems that in the case of DX11+FXC and DX12+FXC, the driver’s compiler aggressively unrolls loops for uniform typed buffer loads, cbuffer loads and uniform texture loads. FXC and DXC preserve loops in resulting DXBC/DXIL. In case of DX12+DXC loops are preserved in the final ISAs.
The resulting ISAs from DX11+FXC and DX12+FXC aren’t exactly the same—perhaps this explains some differences in the timing of other access patterns, although they’re not as dramatic.

I tested it on 2060(driver 572.83 March 2025) and on 3060(driver 576.02 April 2025).

  DX11 5_0     DX12 5_1     DX12 6_0    
  Total Average Std dev Total Average Std dev Total Average Std dev
Buffer<R8>.Load uniform 0.804ms 0.027ms 0.001ms 0.768ms 0.026ms 0.001ms 42.198ms 1.407ms 0.028ms
Buffer<R8>.Load linear 39.710ms 1.324ms 0.012ms 39.499ms 1.317ms 0.017ms 42.034ms 1.401ms 0.027ms
Buffer<R8>.Load random 39.815ms 1.327ms 0.014ms 39.450ms 1.315ms 0.016ms 41.828ms 1.394ms 0.022ms
Buffer<RG8>.Load uniform 1.008ms 0.034ms 0.014ms 1.054ms 0.035ms 0.007ms 41.766ms 1.392ms 0.024ms
Buffer<RG8>.Load linear 39.948ms 1.332ms 0.012ms 39.373ms 1.312ms 0.016ms 41.571ms 1.386ms 0.024ms
Buffer<RG8>.Load random 39.951ms 1.332ms 0.016ms 39.331ms 1.311ms 0.014ms 41.365ms 1.379ms 0.027ms
Buffer<RGBA8>.Load uniform 1.226ms 0.041ms 0.009ms 1.286ms 0.043ms 0.007ms 41.244ms 1.375ms 0.026ms
Buffer<RGBA8>.Load linear 41.410ms 1.380ms 0.017ms 40.771ms 1.359ms 0.013ms 40.938ms 1.365ms 0.024ms
Buffer<RGBA8>.Load random 41.298ms 1.377ms 0.019ms 40.556ms 1.352ms 0.018ms 40.761ms 1.359ms 0.023ms
Buffer<R16f>.Load uniform 0.927ms 0.031ms 0.007ms 0.895ms 0.030ms 0.007ms 40.762ms 1.359ms 0.023ms
Buffer<R16f>.Load linear 39.662ms 1.322ms 0.017ms 39.160ms 1.305ms 0.019ms 40.655ms 1.355ms 0.019ms
Buffer<R16f>.Load random 39.543ms 1.318ms 0.015ms 39.161ms 1.305ms 0.016ms 40.552ms 1.352ms 0.018ms
Buffer<RG16f>.Load uniform 1.008ms 0.034ms 0.007ms 1.055ms 0.035ms 0.007ms 40.505ms 1.350ms 0.018ms
Buffer<RG16f>.Load linear 39.467ms 1.316ms 0.018ms 39.147ms 1.305ms 0.015ms 40.350ms 1.345ms 0.016ms
Buffer<RG16f>.Load random 39.521ms 1.317ms 0.015ms 39.186ms 1.306ms 0.015ms 40.409ms 1.347ms 0.017ms
Buffer<RGBA16f>.Load uniform 1.237ms 0.041ms 0.007ms 1.311ms 0.044ms 0.007ms 40.564ms 1.352ms 0.017ms
Buffer<RGBA16f>.Load linear 40.931ms 1.364ms 0.016ms 40.615ms 1.354ms 0.015ms 40.442ms 1.348ms 0.015ms
Buffer<RGBA16f>.Load random 40.870ms 1.362ms 0.015ms 40.567ms 1.352ms 0.015ms 40.433ms 1.348ms 0.015ms
Buffer<R32f>.Load uniform 0.954ms 0.032ms 0.007ms 0.959ms 0.032ms 0.007ms 40.424ms 1.347ms 0.018ms
Buffer<R32f>.Load linear 39.427ms 1.314ms 0.014ms 39.165ms 1.305ms 0.016ms 40.494ms 1.350ms 0.020ms
Buffer<R32f>.Load random 39.462ms 1.315ms 0.011ms 39.037ms 1.301ms 0.018ms 40.618ms 1.354ms 0.021ms
Buffer<RG32f>.Load uniform 1.060ms 0.035ms 0.007ms 1.065ms 0.036ms 0.007ms 40.850ms 1.362ms 0.022ms
Buffer<RG32f>.Load linear 39.348ms 1.312ms 0.013ms 38.992ms 1.300ms 0.015ms 40.916ms 1.364ms 0.023ms
Buffer<RG32f>.Load random 39.402ms 1.313ms 0.012ms 38.831ms 1.294ms 0.015ms 41.087ms 1.370ms 0.025ms
Buffer<RGBA32f>.Load uniform 1.216ms 0.041ms 0.007ms 1.209ms 0.040ms 0.007ms 81.479ms 2.716ms 0.046ms
Buffer<RGBA32f>.Load linear 78.140ms 2.605ms 0.018ms 76.989ms 2.566ms 0.025ms 82.003ms 2.733ms 0.037ms
Buffer<RGBA32f>.Load random 78.200ms 2.607ms 0.019ms 77.060ms 2.569ms 0.024ms 82.658ms 2.755ms 0.032ms
ByteAddressBuffer.Load uniform 19.650ms 0.655ms 0.008ms 18.832ms 0.628ms 0.006ms 24.822ms 0.827ms 0.011ms
ByteAddressBuffer.Load linear 16.377ms 0.546ms 0.006ms 15.461ms 0.515ms 0.007ms 27.609ms 0.920ms 0.009ms
ByteAddressBuffer.Load random 16.575ms 0.553ms 0.007ms 15.456ms 0.515ms 0.009ms 25.325ms 0.844ms 0.011ms
ByteAddressBuffer.Load2 uniform 28.287ms 0.943ms 0.008ms 24.841ms 0.828ms 0.008ms 28.982ms 0.966ms 0.008ms
ByteAddressBuffer.Load2 linear 43.529ms 1.451ms 0.014ms 42.593ms 1.420ms 0.014ms 45.600ms 1.520ms 0.013ms
ByteAddressBuffer.Load2 random 24.964ms 0.832ms 0.008ms 22.694ms 0.756ms 0.008ms 29.578ms 0.986ms 0.007ms
ByteAddressBuffer.Load3 uniform 34.826ms 1.161ms 0.010ms 31.245ms 1.041ms 0.009ms 35.902ms 1.197ms 0.007ms
ByteAddressBuffer.Load3 linear 45.923ms 1.531ms 0.013ms 44.830ms 1.494ms 0.012ms 47.757ms 1.592ms 0.012ms
ByteAddressBuffer.Load3 random 37.580ms 1.253ms 0.010ms 36.329ms 1.211ms 0.007ms 40.342ms 1.345ms 0.008ms
ByteAddressBuffer.Load4 uniform 44.199ms 1.473ms 0.012ms 40.829ms 1.361ms 0.006ms 42.621ms 1.421ms 0.011ms
ByteAddressBuffer.Load4 linear 108.608ms 3.620ms 0.042ms 105.488ms 3.516ms 0.043ms 112.637ms 3.755ms 0.041ms
ByteAddressBuffer.Load4 random 85.310ms 2.844ms 0.057ms 84.554ms 2.818ms 0.060ms 87.858ms 2.929ms 0.059ms
ByteAddressBuffer.Load2 unaligned uniform 28.740ms 0.958ms 0.013ms 25.386ms 0.846ms 0.014ms 29.640ms 0.988ms 0.024ms
ByteAddressBuffer.Load2 unaligned linear 45.321ms 1.511ms 0.038ms 45.373ms 1.512ms 0.022ms 46.832ms 1.561ms 0.041ms
ByteAddressBuffer.Load2 unaligned random 25.361ms 0.845ms 0.012ms 23.241ms 0.775ms 0.010ms 30.367ms 1.012ms 0.024ms
ByteAddressBuffer.Load4 unaligned uniform 45.088ms 1.503ms 0.021ms 42.077ms 1.403ms 0.019ms 43.720ms 1.457ms 0.034ms
ByteAddressBuffer.Load4 unaligned linear 113.447ms 3.782ms 0.080ms 112.015ms 3.734ms 0.039ms 116.133ms 3.871ms 0.091ms
ByteAddressBuffer.Load4 unaligned random 89.205ms 2.973ms 0.070ms 86.827ms 2.894ms 0.033ms 91.349ms 3.045ms 0.052ms
StructuredBuffer<float>.Load uniform 18.226ms 0.608ms 0.019ms 14.976ms 0.499ms 0.009ms 21.078ms 0.703ms 0.012ms
StructuredBuffer<float>.Load linear 15.250ms 0.508ms 0.010ms 15.167ms 0.506ms 0.009ms 24.837ms 0.828ms 0.016ms
StructuredBuffer<float>.Load random 15.134ms 0.504ms 0.013ms 15.179ms 0.506ms 0.010ms 23.217ms 0.774ms 0.013ms
StructuredBuffer<float2>.Load uniform 22.109ms 0.737ms 0.017ms 21.494ms 0.716ms 0.010ms 23.222ms 0.774ms 0.013ms
StructuredBuffer<float2>.Load linear 47.728ms 1.591ms 0.031ms 24.512ms 0.817ms 0.011ms 26.543ms 0.885ms 0.015ms
StructuredBuffer<float2>.Load random 23.936ms 0.798ms 0.011ms 22.762ms 0.759ms 0.010ms 24.927ms 0.831ms 0.016ms
StructuredBuffer<float4>.Load uniform 43.675ms 1.456ms 0.016ms 42.435ms 1.414ms 0.018ms 45.322ms 1.511ms 0.022ms
StructuredBuffer<float4>.Load linear 58.199ms 1.940ms 0.016ms 55.740ms 1.858ms 0.020ms 59.210ms 1.974ms 0.026ms
StructuredBuffer<float4>.Load random 51.907ms 1.730ms 0.016ms 50.062ms 1.669ms 0.015ms 52.831ms 1.761ms 0.023ms
cbuffer{float4} load uniform 1.702ms 0.057ms 0.006ms 24.469ms 0.816ms 0.012ms 25.815ms 0.860ms 0.015ms
cbuffer{float4} load linear 938.054ms 31.268ms 0.247ms 1003.056ms 33.435ms 0.214ms 1040.154ms 34.672ms 0.316ms
cbuffer{float4} load random 314.245ms 10.475ms 0.050ms 571.129ms 19.038ms 0.119ms 595.134ms 19.838ms 0.086ms
Texture2D<R8>.Load uniform 1.135ms 0.038ms 0.015ms 0.984ms 0.033ms 0.007ms 40.854ms 1.362ms 0.010ms
Texture2D<R8>.Load linear 38.942ms 1.298ms 0.023ms 38.183ms 1.273ms 0.009ms 40.828ms 1.361ms 0.010ms
Texture2D<R8>.Load random 38.812ms 1.294ms 0.021ms 38.189ms 1.273ms 0.010ms 40.752ms 1.358ms 0.010ms
Texture2D<RG8>.Load uniform 1.050ms 0.035ms 0.007ms 1.162ms 0.039ms 0.007ms 40.867ms 1.362ms 0.009ms
Texture2D<RG8>.Load linear 38.469ms 1.282ms 0.019ms 37.984ms 1.266ms 0.010ms 40.899ms 1.363ms 0.006ms
Texture2D<RG8>.Load random 38.386ms 1.280ms 0.022ms 37.789ms 1.260ms 0.010ms 40.721ms 1.357ms 0.009ms
Texture2D<RGBA8>.Load uniform 1.391ms 0.046ms 0.008ms 1.435ms 0.048ms 0.008ms 40.895ms 1.363ms 0.008ms
Texture2D<RGBA8>.Load linear 39.351ms 1.312ms 0.017ms 38.672ms 1.289ms 0.009ms 41.253ms 1.375ms 0.008ms
Texture2D<RGBA8>.Load random 39.403ms 1.313ms 0.018ms 38.450ms 1.282ms 0.008ms 41.032ms 1.368ms 0.007ms
Texture2D<R16F>.Load uniform 0.853ms 0.028ms 0.007ms 0.920ms 0.031ms 0.007ms 40.765ms 1.359ms 0.008ms
Texture2D<R16F>.Load linear 38.354ms 1.278ms 0.016ms 38.018ms 1.267ms 0.008ms 40.765ms 1.359ms 0.008ms
Texture2D<R16F>.Load random 38.384ms 1.279ms 0.016ms 38.038ms 1.268ms 0.008ms 40.739ms 1.358ms 0.008ms
Texture2D<RG16F>.Load uniform 1.103ms 0.037ms 0.007ms 1.140ms 0.038ms 0.007ms 40.728ms 1.358ms 0.008ms
Texture2D<RG16F>.Load linear 38.310ms 1.277ms 0.014ms 37.922ms 1.264ms 0.008ms 40.938ms 1.365ms 0.010ms
Texture2D<RG16F>.Load random 38.220ms 1.274ms 0.011ms 37.929ms 1.264ms 0.008ms 40.910ms 1.364ms 0.015ms
Texture2D<RGBA16F>.Load uniform 1.446ms 0.048ms 0.007ms 1.511ms 0.050ms 0.007ms 41.294ms 1.376ms 0.021ms
Texture2D<RGBA16F>.Load linear 39.002ms 1.300ms 0.010ms 38.745ms 1.291ms 0.008ms 41.856ms 1.395ms 0.020ms
Texture2D<RGBA16F>.Load random 75.330ms 2.511ms 0.011ms 74.951ms 2.498ms 0.008ms 82.200ms 2.740ms 0.080ms
Texture2D<R32F>.Load uniform 1.037ms 0.035ms 0.011ms 0.975ms 0.032ms 0.007ms 42.103ms 1.403ms 0.050ms
Texture2D<R32F>.Load linear 38.070ms 1.269ms 0.008ms 38.039ms 1.268ms 0.008ms 42.379ms 1.413ms 0.052ms
Texture2D<R32F>.Load random 38.043ms 1.268ms 0.008ms 38.063ms 1.269ms 0.008ms 42.490ms 1.416ms 0.056ms
Texture2D<RG32F>.Load uniform 1.065ms 0.036ms 0.007ms 1.171ms 0.039ms 0.007ms 42.903ms 1.430ms 0.056ms
Texture2D<RG32F>.Load linear 38.073ms 1.269ms 0.008ms 37.893ms 1.263ms 0.009ms 43.381ms 1.446ms 0.056ms
Texture2D<RG32F>.Load random 75.534ms 2.518ms 0.011ms 74.989ms 2.500ms 0.009ms 86.579ms 2.886ms 0.081ms
Texture2D<RGBA32F>.Load uniform 1.454ms 0.048ms 0.007ms 1.492ms 0.050ms 0.007ms 87.532ms 2.918ms 0.056ms
Texture2D<RGBA32F>.Load linear 75.944ms 2.531ms 0.022ms 75.648ms 2.522ms 0.032ms 88.144ms 2.938ms 0.028ms
Texture2D<RGBA32F>.Load random 76.341ms 2.545ms 0.036ms 77.046ms 2.568ms 0.041ms 88.533ms 2.951ms 0.016ms
Texture2D<R8>.Sample(nearest) uniform 38.702ms 1.290ms 0.022ms 39.429ms 1.314ms 0.021ms 45.075ms 1.502ms 0.013ms
Texture2D<R8>.Sample(nearest) linear 38.693ms 1.290ms 0.020ms 39.605ms 1.320ms 0.019ms 44.926ms 1.498ms 0.011ms
Texture2D<R8>.Sample(nearest) random 38.749ms 1.292ms 0.022ms 39.721ms 1.324ms 0.017ms 44.963ms 1.499ms 0.011ms
Texture2D<RG8>.Sample(nearest) uniform 38.733ms 1.291ms 0.020ms 39.561ms 1.319ms 0.013ms 45.087ms 1.503ms 0.011ms
Texture2D<RG8>.Sample(nearest) linear 38.901ms 1.297ms 0.015ms 39.734ms 1.324ms 0.010ms 45.192ms 1.506ms 0.012ms
Texture2D<RG8>.Sample(nearest) random 38.957ms 1.299ms 0.015ms 39.612ms 1.320ms 0.011ms 45.205ms 1.507ms 0.009ms
Texture2D<RGBA8>.Sample(nearest) uniform 39.873ms 1.329ms 0.012ms 40.709ms 1.357ms 0.012ms 47.117ms 1.571ms 0.009ms
Texture2D<RGBA8>.Sample(nearest) linear 39.775ms 1.326ms 0.013ms 40.351ms 1.345ms 0.012ms 47.748ms 1.592ms 0.010ms
Texture2D<RGBA8>.Sample(nearest) random 40.031ms 1.334ms 0.011ms 40.415ms 1.347ms 0.012ms 47.570ms 1.586ms 0.012ms
Texture2D<R16F>.Sample(nearest) uniform 39.289ms 1.310ms 0.012ms 39.565ms 1.319ms 0.013ms 44.963ms 1.499ms 0.012ms
Texture2D<R16F>.Sample(nearest) linear 39.332ms 1.311ms 0.011ms 39.735ms 1.325ms 0.012ms 44.913ms 1.497ms 0.009ms
Texture2D<R16F>.Sample(nearest) random 39.503ms 1.317ms 0.011ms 39.855ms 1.329ms 0.011ms 44.872ms 1.496ms 0.009ms
Texture2D<RG16F>.Sample(nearest) uniform 39.413ms 1.314ms 0.014ms 39.790ms 1.326ms 0.011ms 45.028ms 1.501ms 0.009ms
Texture2D<RG16F>.Sample(nearest) linear 39.513ms 1.317ms 0.017ms 39.912ms 1.330ms 0.011ms 45.145ms 1.505ms 0.010ms
Texture2D<RG16F>.Sample(nearest) random 39.564ms 1.319ms 0.019ms 39.926ms 1.331ms 0.012ms 45.123ms 1.504ms 0.011ms
Texture2D<RGBA16F>.Sample(nearest) uniform 40.298ms 1.343ms 0.015ms 41.083ms 1.369ms 0.013ms 47.096ms 1.570ms 0.012ms
Texture2D<RGBA16F>.Sample(nearest) linear 40.323ms 1.344ms 0.020ms 40.790ms 1.360ms 0.013ms 47.823ms 1.594ms 0.013ms
Texture2D<RGBA16F>.Sample(nearest) random 78.754ms 2.625ms 0.032ms 79.363ms 2.645ms 0.016ms 89.415ms 2.981ms 0.031ms
Texture2D<R32F>.Sample(nearest) uniform 40.228ms 1.341ms 0.022ms 40.141ms 1.338ms 0.012ms 45.494ms 1.516ms 0.017ms
Texture2D<R32F>.Sample(nearest) linear 40.366ms 1.346ms 0.017ms 40.291ms 1.343ms 0.012ms 45.707ms 1.524ms 0.023ms
Texture2D<R32F>.Sample(nearest) random 40.531ms 1.351ms 0.020ms 40.297ms 1.343ms 0.012ms 45.804ms 1.527ms 0.022ms
Texture2D<RG32F>.Sample(nearest) uniform 40.474ms 1.349ms 0.023ms 40.198ms 1.340ms 0.012ms 46.028ms 1.534ms 0.024ms
Texture2D<RG32F>.Sample(nearest) linear 40.487ms 1.350ms 0.021ms 40.168ms 1.339ms 0.010ms 46.161ms 1.539ms 0.025ms
Texture2D<RG32F>.Sample(nearest) random 80.605ms 2.687ms 0.034ms 79.714ms 2.657ms 0.012ms 90.298ms 3.010ms 0.034ms
Texture2D<RGBA32F>.Sample(bilinear) uniform 80.803ms 2.693ms 0.024ms 79.754ms 2.658ms 0.012ms 90.635ms 3.021ms 0.030ms
Texture2D<RGBA32F>.Sample(nearest) linear 81.105ms 2.703ms 0.019ms 79.810ms 2.660ms 0.014ms 90.781ms 3.026ms 0.027ms
Texture2D<RGBA32F>.Sample(nearest) random 81.129ms 2.704ms 0.015ms 79.843ms 2.661ms 0.014ms 90.535ms 3.018ms 0.027ms
Texture2D<R8>.Sample(bilinear) uniform 40.967ms 1.366ms 0.011ms 40.328ms 1.344ms 0.011ms 45.736ms 1.525ms 0.013ms
Texture2D<R8>.Sample(bilinear) linear 40.973ms 1.366ms 0.011ms 40.380ms 1.346ms 0.011ms 45.621ms 1.521ms 0.014ms
Texture2D<R8>.Sample(bilinear) random 40.907ms 1.364ms 0.013ms 40.346ms 1.345ms 0.011ms 45.517ms 1.517ms 0.011ms
Texture2D<RG8>.Sample(bilinear) uniform 40.644ms 1.355ms 0.013ms 40.316ms 1.344ms 0.011ms 45.738ms 1.525ms 0.012ms
Texture2D<RG8>.Sample(bilinear) linear 40.756ms 1.359ms 0.009ms 40.263ms 1.342ms 0.011ms 45.844ms 1.528ms 0.016ms
Texture2D<RG8>.Sample(bilinear) random 40.682ms 1.356ms 0.011ms 40.354ms 1.345ms 0.008ms 45.741ms 1.525ms 0.016ms
Texture2D<RGBA8>.Sample(bilinear) uniform 41.475ms 1.382ms 0.011ms 41.411ms 1.380ms 0.008ms 47.403ms 1.580ms 0.017ms
Texture2D<RGBA8>.Sample(bilinear) linear 41.356ms 1.379ms 0.008ms 41.067ms 1.369ms 0.008ms 48.027ms 1.601ms 0.020ms
Texture2D<RGBA8>.Sample(bilinear) random 41.544ms 1.385ms 0.012ms 41.222ms 1.374ms 0.008ms 47.776ms 1.593ms 0.015ms
Texture2D<R16F>.Sample(bilinear) uniform 40.684ms 1.356ms 0.011ms 40.228ms 1.341ms 0.010ms 45.263ms 1.509ms 0.015ms
Texture2D<R16F>.Sample(bilinear) linear 40.568ms 1.352ms 0.014ms 40.175ms 1.339ms 0.011ms 45.252ms 1.508ms 0.017ms
Texture2D<R16F>.Sample(bilinear) random 40.622ms 1.354ms 0.011ms 40.039ms 1.335ms 0.013ms 45.178ms 1.506ms 0.016ms
Texture2D<RG16F>.Sample(bilinear) uniform 40.528ms 1.351ms 0.012ms 39.964ms 1.332ms 0.013ms 45.288ms 1.510ms 0.014ms
Texture2D<RG16F>.Sample(bilinear) linear 40.518ms 1.351ms 0.010ms 40.021ms 1.334ms 0.010ms 45.517ms 1.517ms 0.011ms
Texture2D<RG16F>.Sample(bilinear) random 40.472ms 1.349ms 0.010ms 39.986ms 1.333ms 0.009ms 45.491ms 1.516ms 0.013ms
Texture2D<RGBA16F>.Sample(bilinear) uniform 41.330ms 1.378ms 0.010ms 41.165ms 1.372ms 0.009ms 47.369ms 1.579ms 0.014ms
Texture2D<RGBA16F>.Sample(bilinear) linear 41.279ms 1.376ms 0.009ms 40.793ms 1.360ms 0.010ms 48.151ms 1.605ms 0.019ms
Texture2D<RGBA16F>.Sample(bilinear) random 80.355ms 2.679ms 0.014ms 79.374ms 2.646ms 0.013ms 89.522ms 2.984ms 0.028ms
Texture2D<R32F>.Sample(bilinear) uniform 40.659ms 1.355ms 0.011ms 40.164ms 1.339ms 0.008ms 45.560ms 1.519ms 0.015ms
Texture2D<R32F>.Sample(bilinear) linear 40.642ms 1.355ms 0.010ms 40.279ms 1.343ms 0.010ms 45.644ms 1.521ms 0.016ms
Texture2D<R32F>.Sample(bilinear) random 40.703ms 1.357ms 0.011ms 40.201ms 1.340ms 0.010ms 45.749ms 1.525ms 0.018ms
Texture2D<RG32F>.Sample(bilinear) uniform 40.602ms 1.353ms 0.011ms 40.045ms 1.335ms 0.009ms 46.040ms 1.535ms 0.018ms
Texture2D<RG32F>.Sample(bilinear) linear 40.598ms 1.353ms 0.009ms 40.016ms 1.334ms 0.009ms 46.230ms 1.541ms 0.017ms
Texture2D<RG32F>.Sample(bilinear) random 80.478ms 2.683ms 0.015ms 79.367ms 2.646ms 0.016ms 90.422ms 3.014ms 0.027ms
Texture2D<RGBA32F>.Sample(bilinear) uniform 160.777ms 5.359ms 0.024ms 158.589ms 5.286ms 0.022ms 180.288ms 6.010ms 0.046ms
Texture2D<RGBA32F>.Sample(bilinear) linear 160.286ms 5.343ms 0.025ms 158.861ms 5.295ms 0.026ms 180.274ms 6.009ms 0.039ms
Texture2D<RGBA32F>.Sample(bilinear) random 239.558ms 7.985ms 0.044ms 237.757ms 7.925ms 0.038ms 267.076ms 8.903ms 0.066ms

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant