multiple GPUs are not working properly in distribution tutorial

I ran this tutorial on google cloud compute engine.
My instance has 2 GPUs(A100),and tensorflow recognizes my GPUs.

from tensorflow.python.client import device_lib
device_lib.list_local_devices()

[name: "/device:CPU:0"
 device_type: "CPU"
 memory_limit: 268435456
 locality {
 }
 incarnation: 7266000849052307782,
 name: "/device:GPU:0"
 device_type: "GPU"
 memory_limit: 40332623872
 locality {
   bus_id: 1
   links {
     link {
       device_id: 1
       type: "StreamExecutor"
       strength: 1
     }
   }
 }
 incarnation: 12716067537431769392
 physical_device_desc: "device: 0, name: A100-SXM4-40GB, pci bus id: 0000:00:04.0, compute capability: 8.0",
 name: "/device:GPU:1"
 device_type: "GPU"
 memory_limit: 40332623872
 locality {
   bus_id: 1
   links {
     link {
       type: "StreamExecutor"
       strength: 1
     }
   }
 }
 incarnation: 6366335206738999217
 physical_device_desc: "device: 1, name: A100-SXM4-40GB, pci bus id: 0000:00:05.0, compute capability: 8.0"]

But, during training, my 2nd GPU’s utilization is around 1~2%.(1st GPU’s utilization is around 40%).

How can I get the second GPU to work properly?

Start by running the distribution tutorial linked in the issue and review its training setup alongside the device_lib.list_local_devices() output. Investigate why the second GPU remains at 1–2% utilization, then verify that training uses both detected A100 GPUs as expected.