diff --git a/src/intel/vulkan/genX_acceleration_structure.c b/src/intel/vulkan/genX_acceleration_structure.c
index 31859c3e9e4..db5c34cdcdb 100644
--- a/src/intel/vulkan/genX_acceleration_structure.c
+++ b/src/intel/vulkan/genX_acceleration_structure.c
@@ -835,8 +835,7 @@ cmd_build_acceleration_structures(
                    &data, sizeof(data));
    }
 
-   if (anv_cmd_buffer_is_render_queue(cmd_buffer))
-      genX(flush_pipeline_select_gpgpu)(cmd_buffer);
+   genX(flush_pipeline_select_gpgpu)(cmd_buffer);
 
    /* Due to the nature of GRL and its heavy use of jumps/predication, we
     * cannot tell exactly in what order the CFE_STATE we insert are going to
diff --git a/src/intel/vulkan/genX_cmd_compute.c b/src/intel/vulkan/genX_cmd_compute.c
index 9a3a09e78ce..ce1e7d9f2fb 100644
--- a/src/intel/vulkan/genX_cmd_compute.c
+++ b/src/intel/vulkan/genX_cmd_compute.c
@@ -587,8 +587,7 @@ genX(cmd_buffer_dispatch_kernel)(struct anv_cmd_buffer *cmd_buffer,
 
    genX(cmd_buffer_config_l3)(cmd_buffer, kernel->l3_config);
 
-   if (anv_cmd_buffer_is_render_queue(cmd_buffer))
-      genX(flush_pipeline_select_gpgpu)(cmd_buffer);
+   genX(flush_pipeline_select_gpgpu)(cmd_buffer);
 
    /* Apply any pending pipeline flushes we may have.  We want to apply them
     * now because, if any of those flushes are for things like push constants,