Adjustable divide data in row

[pepi37]

Ok, as always I have some strange problems: if I can say at start dont judge my definitions or decisions and tell me what is better or not:thanks for that.
So problem is: I have 5 computers running linux: two are Ryzen 3700X one is Ryzen 3600 , one is Ryzen 3900 and one is Intel I5-9600K
So I need to divide data for processing ( CRUS sequences) equal and in row .
This is real data
3700X 4 workers - 2 cores per worker (*2 since I got same CPU)
3900X 4 workers - 3 cores per worker
3600 2 worker - 3 cores per worker
Intel 1 worker with 6 cores.

So I can sum all number of cores ( in this case 40 cores) divide it with number or candidates lets say 1000 candidates : so every core got
25 workunits)
For that I use script Batalov wrote and it works perfectly

echo 'SPLIT [number of workers]'
range=$(seq 1 $1);
for i in `seq 1 $1`; do
echo '[Worker #'$i']'; awk 'a++%'$1'=='$i'-1' in.txt
done > final.txt

So that is first part. and it is easy.

Second part is puzzling me
For 3700 I need to take first 8 sections from "final.txt- (Batalov script) since it has 8 cores: sort it again to be in row and split it in 4 equal parts.
So I take 8 sections, sort it and using Batalov script got perfect answer.

But can be done faster then this? - that is my question
I would like that answer can be adjustable since for other situations number of workers can be changes.

Prefer linux script as answer.

Thanks for any reply