From 9536952f0cb20eeeeed2c4302e5828ab922430b3 Mon Sep 17 00:00:00 2001
From: Tristan Russell <tristanr@trussnexus.com>
Date: Tue, 22 Jul 2025 09:47:20 -0400
Subject: [PATCH] feat: Initial Commit

---
 .clangd |  6 ++++++
 main.cu | 41 +++++++++++++++++++++++++++++++++++++++++
 2 files changed, 47 insertions(+)
 create mode 100644 .clangd
 create mode 100644 main.cu

diff --git a/.clangd b/.clangd
new file mode 100644
index 0000000..31209a5
--- /dev/null
+++ b/.clangd
@@ -0,0 +1,6 @@
+CompileFlags:
+  Add:
+    - --cuda-path=/opt/cuda
+    - --cuda-gpu-arch=sm_89   # Replace XX with your actual GPU architecture, e.g., 86
+    - -I/opt/cuda/include
+    - -L/opt/cuda/lib64
diff --git a/main.cu b/main.cu
new file mode 100644
index 0000000..1cbe9a0
--- /dev/null
+++ b/main.cu
@@ -0,0 +1,41 @@
+#include <stdio.h>
+
+__global__ void cuda_hello(int* number, int* out) {
+    *out = *number;
+    for(int i = 0; i < 255; i++) {
+        *out += 1;
+    }
+}
+
+int main() {
+    int *number, *out;
+    int *d_number, *d_out;
+    number = (int*)malloc(sizeof(int));
+    out = (int*)malloc(sizeof(int));
+
+    *number = 22;
+    *out = 0;
+
+    cudaMalloc((void**)&d_number, sizeof(int));
+    cudaMemcpy(d_number, number, sizeof(int), cudaMemcpyHostToDevice);
+    cudaMalloc((void**)&d_out, sizeof(int));
+    cudaMemcpy(d_out, out, sizeof(int), cudaMemcpyHostToDevice);
+
+
+    cuda_hello<<<1,1>>>(d_number, d_out);
+
+    cudaDeviceSynchronize();
+
+    cudaMemcpy(out, d_out, sizeof(int), cudaMemcpyDeviceToHost);
+
+    cudaFree(d_number);
+    cudaFree(d_out);
+
+
+    printf("Number: %i\n", *out);
+
+    free(number);
+    free(out);
+
+    return 0;
+}