Solved the probelm batch size in TensorRT 7?

Hi all,
In the TensorRT 5.x and 6.x, batch size was a problem. when I feed the image to pure tensorrt model then give me results in the one by one mode time, i.e if one batch is take 30ms then 4 batch image are taken about 120ms.
I tested this problem with ssd-mobile-v1 and v2 model in the jetapck 4.2 and 4.3, I want to know, Is the problem solved in the tensorrt 7?
in this post, I don’t know, what’s means?

If execution of single batch is utilizing the complete compute resource then execution time will increase proportional to the batch size as mentioned in the post.

Can you check the CPU/GPU compute usage while executing single batch size?

Also,in TRT 7 you can try creating optimization profile for multiple batch sizes and check if that is resulting in better performance in your case.
https://docs.nvidia.com/deeplearning/sdk/tensorrt-archived/tensorrt-710-ea/tensorrt-developer-guide/index.html#opt_profiles

Thanks