linux perf sum clang vs gcc

#gcc
$ gcc --version
gcc (Ubuntu 5.4.0-6ubuntu1~16.04.4) 5.4.0 20160609
$ gcc -O3 -o sum sum.c
$ time ./sum 5000
./sum 5000  1.17s user 0.00s system 99% cpu 1.171 total
$ sudo perf stat ./sum 5000

 Performance counter stats for './sum 5000':

       1168.298144      task-clock (msec)         #    1.000 CPUs utilized
                29      context-switches          #    0.025 K/sec
                 0      cpu-migrations            #    0.000 K/sec
                55      page-faults               #    0.047 K/sec
     5,023,172,901      cycles                    #    4.300 GHz
    18,773,198,728      instructions              #    3.74  insn per cycle
     1,256,435,755      branches                  # 1075.441 M/sec
         1,011,161      branch-misses             #    0.08% of all branches

       1.168475610 seconds time elapsed

#clang
$ clang --version
clang version 3.8.0-2ubuntu4 (tags/RELEASE_380/final)
$ clang -O3 -o sum sum.c
$ time ./sum 5000
./sum 5000  4.65s user 0.00s system 99% cpu 4.656 total
$ sudo perf stat ./sum 5000

 Performance counter stats for './sum 5000':

       4656.012983      task-clock (msec)         #    1.000 CPUs utilized
                16      context-switches          #    0.003 K/sec
                 0      cpu-migrations            #    0.000 K/sec
                52      page-faults               #    0.011 K/sec
    20,019,074,780      cycles                    #    4.300 GHz
    53,775,258,783      instructions              #    2.69  insn per cycle
     1,255,158,509      branches                  #  269.578 M/sec
         1,027,384      branch-misses             #    0.08% of all branches

       4.655998791 seconds time elapsed