Just consider the codes given below:

hashcat32 -a 3 -m 1000 hash.txt ?l?l?l?l?l?l


mp32 ?l?l?l?l?l?l | hashcat32 -a 0 -m 1000 hash.txt

As you guys can see I've done the same thing in two different methods,but the latter method is significantly slower than the former one,why is this so?can anyone please explain?:-)
read this thread https://hashcat.net/forum/thread-6839.html
Thanks man…!
The answer to my question,which I found out after reading the above mentioned thread was:Mask processor is CPU-accelerated while the mask candidate generator in HashCat is GPU-accelerated,correct me if I'm wrong.

If it's so then why is it designed like that? Mask Processor could've guaranteed more speed if it was GPU-accelerated,don't you guys think so?
Not quite. The difference is with mp (or any other external candidate generator), candidates are generated on the CPU, piped into hashcat via stdin, and then streamed to the compute devices, whereas the builtin mask attack generates its own candidates in the kernel loop on the device.
Yeah,I understood that,but since it's slow without adding any rules,what would people normally use as an alternative for -r or -q parameters in MP?
Come up with different solutions if they're attacking fast hash modes. PACK for example can do the same thing by generating hcmask files.
Tnx,PACK is an amazing tool,but is there a version or an alternative available for windows?
PACK is just a collection of python scripts. You can run python on almost any OS, including Windows.
Alright,so I downloaded PACK and I got scripts named:rulegen.py,statsgen.py,maskgen.py&policygen.py-which should I use?