diff --git a/test/registered/jit/benchmark/bench_spec_topk1.py b/test/registered/jit/benchmark/bench_spec_topk1.py
index 1635a1046..fada9272c 100644
--- a/test/registered/jit/benchmark/bench_spec_topk1.py
+++ b/test/registered/jit/benchmark/bench_spec_topk1.py
@@ -12,11 +12,12 @@ from sglang.jit_kernel.benchmark.utils import (
     run_benchmark,
 )
 from sglang.kernels.ops.speculative.topk1 import draft_topk1_postprocess
-from sglang.test.ci.ci_register import register_cuda_ci
+from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
 
 register_cuda_ci(
     est_time=30, stage="base-b-kernel-benchmark", runner_config="1-gpu-large"
 )
+register_amd_ci(est_time=30, stage="jit-kernel-benchmark", runner_config="amd")
 
 
 BATCH_SIZE_RANGE = get_benchmark_range(
diff --git a/test/registered/jit/benchmark/bench_vocab_parallel_embedding.py b/test/registered/jit/benchmark/bench_vocab_parallel_embedding.py
index 4c5b87239..2082a1925 100644
--- a/test/registered/jit/benchmark/bench_vocab_parallel_embedding.py
+++ b/test/registered/jit/benchmark/bench_vocab_parallel_embedding.py
@@ -8,11 +8,12 @@ from sglang.kernels.ops.embeddings.vocab_parallel_embedding import (
     vocab_parallel_embedding,
 )
 from sglang.srt.layers.vocab_parallel_embedding import get_masked_input_and_mask
-from sglang.test.ci.ci_register import register_cuda_ci
+from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
 
 register_cuda_ci(
     est_time=10, stage="base-b-kernel-benchmark", runner_config="1-gpu-large"
 )
+register_amd_ci(est_time=10, stage="jit-kernel-benchmark", runner_config="amd")
 
 # Key order must match the perf_report x_names.
 DEFAULTS = dict(