diff --git a/python/sglang/srt/arg_groups/overrides.py b/python/sglang/srt/arg_groups/overrides.py
index 597c19f098a1..04402f21aba3 100644
--- a/python/sglang/srt/arg_groups/overrides.py
+++ b/python/sglang/srt/arg_groups/overrides.py
@@ -1446,11 +1446,11 @@ def _moe_runner_backend_quant_constraints(view: Any) -> dict:
     if (
         moe_runner_backend == "auto"
         and view.quantization == "modelopt_fp4"
-        and get_platform().is_sm120
+        and (get_platform().is_sm120 or get_platform().is_sm110)
     ):
         moe_runner_backend = "flashinfer_cutlass"
         logger.info(
-            "Use flashinfer_cutlass as MoE runner backend on SM120 for "
+            "Use flashinfer_cutlass as MoE runner backend on SM110/SM120 for "
             "modelopt_fp4 (trtllm-gen MoE kernels are SM100-only)"
         )
     if moe_runner_backend != view.moe_runner_backend:
diff --git a/python/sglang/srt/layers/quantization/fp8_utils.py b/python/sglang/srt/layers/quantization/fp8_utils.py
index 2c4a70be9fff..883af39a4c31 100755
--- a/python/sglang/srt/layers/quantization/fp8_utils.py
+++ b/python/sglang/srt/layers/quantization/fp8_utils.py
@@ -1009,6 +1009,9 @@ def initialize_fp8_gemm_config() -> None:
     backend = get_exec().kernel.fp8_gemm_runner_backend
     if backend == "auto" and get_platform().is_sm120:
         backend = "cutlass"
+    elif backend == "auto" and get_platform().is_sm110:
+        # DeepGEMM's block-FP8 kernel does not support SM110.
+        backend = "triton"
 
     backend = Fp8GemmRunnerBackend(backend)
 
diff --git a/python/sglang/srt/runtime_context.py b/python/sglang/srt/runtime_context.py
index a361146e76fa..a1920a115624 100644
--- a/python/sglang/srt/runtime_context.py
+++ b/python/sglang/srt/runtime_context.py
@@ -1776,6 +1776,7 @@ def pre_capture_activation_reserve_mb(gpu_mem: float | None) -> float:
     "is_sm90": "is_sm90_supported",
     "is_sm100": "is_sm100_supported",
     "is_sm100_or_sm110": "is_sm100_or_sm110_supported",
+    "is_sm110": "is_sm110_supported",
     "is_sm120": "is_sm120_supported",
     "is_blackwell": "is_blackwell_supported",
     "is_hopper_with_cuda_12_3": "is_hopper_with_cuda_12_3",
diff --git a/python/sglang/srt/utils/common.py b/python/sglang/srt/utils/common.py
index a05df0388836..b6ba5788d6cd 100644
--- a/python/sglang/srt/utils/common.py
+++ b/python/sglang/srt/utils/common.py
@@ -311,6 +311,11 @@ def _check_cuda_device_version(
         cuda_version=(12, 8),
     )
 )
+is_sm110_supported = lru_cache(maxsize=1)(
+    partial(
+        _check_cuda_device_version, device_capability_majors=[11], cuda_version=(12, 8)
+    )
+)
 is_sm80_supported = lru_cache(maxsize=1)(
     partial(
         _check_cuda_device_version, device_capability_majors=[8], cuda_version=(11, 0)