Project 0: Aditya Hota

README.md

@@ -3,12 +3,44 @@ Project 0 Getting Started
 
 **University of Pennsylvania, CIS 565: GPU Programming and Architecture, Project 0**
 
-* (TODO) YOUR NAME HERE
-  * (TODO) [LinkedIn](), [personal website](), [twitter](), etc.
-* Tested on: (TODO) Windows 22, i7-2222 @ 2.22GHz 22GB, GTX 222 222MB (Moore 2222 Lab)
+* Aditya Hota
+  * [LinkedIn](https://www.linkedin.com/in/aditya-hota)
+* Tested on: Windows 11, i7-8750H @ 2.20 GHz 20 GB, GTX 1050 Ti with Max-Q Design 6 GB (personal laptop)
 
-### (TODO: Your README)
+## Project Deliverables (Part 3)
+## Part 3.1: CUDA
+Compute Capability: `6.1`
 
-Include screenshots, analysis, etc. (Remember, this is public, so don't put
-anything here that you don't want to share with the world.)
+### Part 3.1.1: Modify CUDA Program
+Screenshot of CUDA project executable (with updated colors as of 9/1/2021):
 
+<img src="images/cuda_3_1_1_name2.png" width="500">
+
+### Part 3.1.2: Analyze Program
+Application trace timeline:
+
+<img src="images/cuda_3_1_2_trace.png" width="750">
+
+### Part 3.1.3: Nsight Debugging
+The value for index 2300 is computed on thread `(12, 2, 0)` of block `(43, 0, 0)`. This is shown with the yellow arrow being on the 13th thread of the block.
+
+<img src="images/cuda_3_1_3_warp_info.png" width="750">
+
+## Part 3.2: WebGL
+My Chrome browser supports both WebGL 1 and WebGL 2.
+
+<img src="images/cuda_3_2_webgl_1.png" width="500">
+<br>
+<img src="images/cuda_3_2_webgl_2.png" width="500">
+
+## Part 3.3: DXR
+My computer supports the software fallback layer only (FL). I made my triangle purple.
+
+<img src="images/cuda_3_3_dxr_test.png" width="500">
+
+## Other contributions
+On Optimus machines, you may get the error `Cuda error: Kernel failed!: all CUDA-capable devices are busy or unavailable.` It looks like Windows is starting to manage automatic GPU switching rather than the NVIDIA driver, so if this issue comes up, you might be able to fix it by going to the Windows Settings app > System > Display > Graphics and manually adding the path to the compiled executable that was built in VS code. Then, you can set the GPU to NVIDIA.
+
+<img src="images/cuda_getting_started_set_GPU.png" width="250">
+<br>
+<img src="images/cuda_getting_started_set_GPU_confirm.png" width="250">

cuda-getting-started/src/main.cpp

@@ -10,8 +10,7 @@
  * C main function.
  */
 int main(int argc, char* argv[]) {
-    // TODO: Change this line to use your name!
-    m_yourName = "TODO: YOUR NAME HERE";
+    m_yourName = "Aditya Hota";
 
     if (init(argc, argv)) {
         mainLoop();

dxr-support/src/D3D12RaytracingHelloWorld/Raytracing.hlsl

@@ -69,7 +69,7 @@ void MyRaygenShader()
 [shader("closesthit")]
 void MyClosestHitShader(inout RayPayload payload, in MyAttributes attr)
 {
-    float3 barycentrics = float3(1 - attr.barycentrics.x - attr.barycentrics.y, attr.barycentrics.x, attr.barycentrics.y);
+    float3 barycentrics = float3(0.5f, 0.f, 1.f);
     payload.color = float4(barycentrics, 1);
 }