Is it expected running `sudo pcm 1 -nc -csv -f` on a 224 CPU host results in pcm using 1,125 threads?

I'm running Intel pcm on a host with 224 CPUs and was surprised to see that Intel pcm is using 1,125 threads. Is that expected? Seems like a lot, or?

$ ps -eLo psr,pid,tid,comm | perl -lane 'if(m~( pcm$|PSR)~){ print; }' | head
PSR     PID     TID COMMAND
  0 3173348 3173348 pcm
  0 3173348 3173349 pcm
223 3173348 3173350 pcm
222 3173348 3173351 pcm
223 3173348 3173352 pcm
222 3173348 3173353 pcm
 38 3173348 3173354 pcm
  0 3173348 3173355 pcm
  4 3173348 3173356 pcm

$ ps -eLo psr,pid,tid,comm | perl -lane 'if(m~( pcm$|PSR)~){ print; }' | wc --lines
1126

$ ps auxwww | egrep "([U]SER|[p]cm )"
USER         PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME COMMAND
root     3173347  0.0  0.0  10248  3584 ?        S    16:30   0:00 sudo /usr/sbin/pcm 1 -nc -csv -f
root     3173348  3.7  0.0 24458416 7248 ?       Sl   16:30   0:36 /usr/sbin/pcm 1 -nc -csv -f

$ nproc
224

$ cat /etc/os-release | egrep -v URL
PRETTY_NAME="Debian GNU/Linux 12 (bookworm)"
NAME="Debian GNU/Linux"
VERSION_ID="12"
VERSION="12 (bookworm)"
VERSION_CODENAME=bookworm
ID=debian

Start by reproducing sudo pcm 1 -nc -csv -f on a 224-CPU Debian 12 host and compare the process and thread counts shown by the commands in the issue. Trace where PCM creates worker threads, then determine whether 1,125 threads is expected; done means providing a documented explanation or a narrowly scoped correction if the count is unintended.