[compilation error] nvcc fatal : Unsupported gpu architecture 'compute_89'

[fmo-mt]

What's the issue, what's expected?:
I was following the documentation as this, I compile msccl with
make -j src.build NVCC_GENCODE="-gencode=arch=compute_80,code=sm_80"
and when acting
python3 -m pip install .
the python-pip threw an error, I looked up the error log and found that this failure is caused by:

          FAILED: CMakeFiles/cmTC_419f6.dir/main.cu.o
          /usr/local/cuda/bin/nvcc -forward-unknown-to-host-compiler   -std=c++17 "--generate-code=arch=compute_70,code=[compute_70,sm_70]" "--generate-code=arch=com
pute_80,code=[compute_80,sm_80]" "--generate-code=arch=compute_89,code=[compute_89,sm_89]" "--generate-code=arch=compute_90,code=[compute_90,sm_90]" -MD -MT CMakeFil
es/cmTC_419f6.dir/main.cu.o -MF CMakeFiles/cmTC_419f6.dir/main.cu.o.d -x cu -c /tmp/tmplaptwlbv/CMakeFiles/CMakeScratch/TryCompile-yzB6GO/main.cu -o CMakeFiles/cmTC_
419f6.dir/main.cu.o
          nvcc fatal   : Unsupported gpu architecture 'compute_89'
          ninja: build stopped: subcommand failed.

My environment:

• GPUs: A100 80G
• cuda driver: 11.7
• driver version: 515.48.07
• nvcc: 11.7