Well, thank you all for your contributions, I learned from this a lot. Now when I think about it again, I realize that I was doing too many permutations prior to partitioning which resulted in too many repeated sets, most of them are not prime. Here I reversed that order and permuted only needed subsets. It is significantly faster now with few allocations.
using Primes
using Combinatorics
mprime(k) = isprime(evalpoly(10,k))
function brute1180(n)
total = 0
for i in partitions(1:n)
mul = 1
for j in i
mul *= sum(mprime, permutations(j))
mul == 0 && break
end
total += mul
end
total
end
@btime brute1180(7)
4.647 ms (77712 allocations: 6.37 MiB)
1205