From 50693920194ef8e860b1c86a565319ada0ceb602 Mon Sep 17 00:00:00 2001
From: Varsius <42544122+Varsius@users.noreply.github.com>
Date: Sun, 6 Sep 2026 22:15:51 +0200
Subject: [PATCH] fix(v2026-spot-gpu): replace dummy paper link with acm url

---
 cluster-trace-v2026-spot-gpu/README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/cluster-trace-v2026-spot-gpu/README.md b/cluster-trace-v2026-spot-gpu/README.md
index 9ae0c90..9978421 100644
--- a/cluster-trace-v2026-spot-gpu/README.md
+++ b/cluster-trace-v2026-spot-gpu/README.md
@@ -5,7 +5,7 @@ The node dataset captures the operational characteristics of **4278 GPU nodes**
 
 We categorize AI jobs as as _High-Priority_ (HP) workloads with **strict SLOs** and _Spot_ workloads that used  **spot instances as opportunistic resources**. The job details are presented in the workload dataset.
 
-Based on such traces, we have also developed a scheduling framework that maximizes the SLO guarantee for spot workloads with spot GPU provisioning predictions. For detailed description, please refer to [our ASPLOS'26 paper](https://www.xxx).
+Based on such traces, we have also developed a scheduling framework that maximizes the SLO guarantee for spot workloads with spot GPU provisioning predictions. For detailed description, please refer to [our ASPLOS'26 paper](https://dl.acm.org/doi/abs/10.1145/3760250.3762231).
 
 ## 🗄️ Dataset Details
 The node dataset is provided in the file `node_info_df.csv`.