From fc1b93f15c1a5b431f1b06bc7cbdecb2b4a7d030 Mon Sep 17 00:00:00 2001
From: Junda Chen <32371474+GindaChen@users.noreply.github.com>
Date: Thu, 15 Oct 2026 00:50:58 -0700
Subject: [PATCH] [docs] Clarify CUDA startup timing boundaries

---
 docs/contributing/startup_profiling.md | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/docs/contributing/startup_profiling.md b/docs/contributing/startup_profiling.md
index e3e0b77f0..a26387d17 100644
--- a/docs/contributing/startup_profiling.md
+++ b/docs/contributing/startup_profiling.md
@@ -21,6 +21,10 @@ The script measures torch and FastVideo/helper imports separately, device and
 process-group setup, VAE loading, and the first decode of a fixed seeded latent
 input. It preserves launcher-provided rendezvous settings; otherwise it chooses a
 local rendezvous port. It uses one GPU, FP32, and the default cached Wan VAE path.
+Imports and configuration can themselves initialize CUDA. The
+`cuda_initialized_before_device_setup` field records whether the explicit
+device-setup interval begins with an existing context; import time includes
+any initialization side effects.
 It verifies output shape and finiteness; those checks alone do not establish
 parity. Compare the saved tensors against an independent unchanged-source run.