diff --git a/python/sglang/srt/layers/attention/flashinfer_backend.py b/python/sglang/srt/layers/attention/flashinfer_backend.py
index 6b66e12d6..a5b207c77 100644
--- a/python/sglang/srt/layers/attention/flashinfer_backend.py
+++ b/python/sglang/srt/layers/attention/flashinfer_backend.py
@@ -1187,7 +1187,7 @@ class FlashInferMultiStepDraftBackend:
 
     def init_cuda_graph_state(self, max_bs: int, max_num_tokens: int):
         self.cuda_graph_kv_indices = torch.zeros(
-            (self.speculative_num_steps, max_bs * self.topk * self.max_context_len),
+            (self.speculative_num_steps, max_bs * self.max_context_len),
             dtype=torch.int32,
             device="cuda",
         )
@@ -1349,10 +1349,6 @@ def fast_decode_plan(
                 self.device, non_blocking=non_blocking
             )
 
-    # TODO:
-    # We want to cache `empty_q_data`, `empty_kv_cache`, `last_page_len_host` (if it is ones) in the wrapper
-    # so that we do not need to create them every time.
-
     # Create empty tensors for dtype info if needed
     empty_q_data = torch.empty(
         0,
diff --git a/sgl-kernel/CMakeLists.txt b/sgl-kernel/CMakeLists.txt
index 5867f95f5..80f29921f 100644
--- a/sgl-kernel/CMakeLists.txt
+++ b/sgl-kernel/CMakeLists.txt
@@ -81,7 +81,7 @@ FetchContent_Populate(repo-triton)
 FetchContent_Declare(
     repo-flashinfer
     GIT_REPOSITORY https://github.com/flashinfer-ai/flashinfer.git
-    GIT_TAG        1a85c439a064c1609568675aa580a409a53fb183
+    GIT_TAG        018b551825c8e5579206e6eb9d3229fa679202b3
     GIT_SHALLOW    OFF
 )
 FetchContent_Populate(repo-flashinfer)