using DLA but not accelerate

I set DLA with these code,and many warning come out.I think some layer is worked on DLA,but it seems not obviously accelerated.How can I confirm the DLA is working?

_builder->setHalf2Mode(HalfFloat[chooseHalfFloat].half);
	_builder->allowGPUFallback(true);
	DeviceType type = (DeviceType)1;
	_builder->setDefaultDeviceType(type);

here is log :

WARNING: Default DLA is enabled but layer layer1-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer2-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer3-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer4-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer6-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer7-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer8-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer10-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer11-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer13-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer14-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer15-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer17-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer18-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer20-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer21-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer23-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer24-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer26-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer27-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer29-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer30-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer32-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer33-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer35-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer36-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer38-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer39-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer40-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer42-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer43-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer45-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer46-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer48-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer49-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer51-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer52-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer54-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer55-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer57-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer58-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer60-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer61-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer63-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer64-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer65-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer67-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer68-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer70-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer71-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer73-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer74-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer76-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer77-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer78-act is not running on DLA, falling back to GPU.
WARNING: layer80-route: Concatenation on DLA requires at least two inputs.
WARNING: Default DLA is enabled but layer layer80-route is not running on DLA, falling back to GPU.
WARNING: layer81-maxpool: DLA only supports windows in the range of [1-8].
WARNING: Default DLA is enabled but layer layer81-maxpool is not running on DLA, falling back to GPU.
WARNING: layer82-route: Concatenation on DLA requires at least two inputs.
WARNING: Default DLA is enabled but layer layer82-route is not running on DLA, falling back to GPU.
WARNING: layer83-maxpool: DLA only supports windows in the range of [1-8].
WARNING: Default DLA is enabled but layer layer83-maxpool is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer85-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer86-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer87-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer88-act is not running on DLA, falling back to GPU.
WARNING: layer91-route: Concatenation on DLA requires at least two inputs.
WARNING: Default DLA is enabled but layer layer91-route is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer92-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer93-upsample is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer95-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer96-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer97-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer98-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer99-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer100-act is not running on DLA, falling back to GPU.
WARNING: layer103-route: Concatenation on DLA requires at least two inputs.
WARNING: Default DLA is enabled but layer layer103-route is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer104-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer105-upsample is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer107-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer108-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer109-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer110-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer111-act is not running on DLA, falling back to GPU.
WARNING: Default DLA is enabled but layer layer112-act is not running on DLA, falling back to GPU.
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer11-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer13-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer14-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer15-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer17-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer18-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer20-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer21-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer23-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer24-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer26-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer27-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer29-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer30-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer32-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer33-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer35-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer36-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer38-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer39-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer40-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer42-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer43-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer45-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer46-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer48-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer49-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer51-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer52-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer54-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer55-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer57-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer58-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer60-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer61-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer63-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer64-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer65-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer67-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer68-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer70-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer71-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer73-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer74-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer76-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer77-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer78-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer85-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer86-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer87-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer88-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer92-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer95-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer96-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer97-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer98-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer99-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer100-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer104-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer107-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer108-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer109-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer110-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer111-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer112-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer79-maxpool
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer89-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer101-conv
WARNING: DLA supports only 8 subgraphs per DLA core. Switching to GPU for layer layer113-conv

Hi,

Based on your log, lots of layers inside your model are fall-backed to the GPU.

WARNING: Default DLA is enabled but layer layer1-act is not running on DLA, falling back to GPU.

You can find the support matrix of our DLA here:
https://docs.nvidia.com/deeplearning/sdk/tensorrt-developer-guide/index.html#dla_layers

Please noticed that there are some overhead in data transfer if an intermediate layer is fallbacked to the GPU.

Thanks.