When issue_width is 3 and there are two threads are running, how many instructions can we fetch in one cycle? Is it 3 or 3*2?

It should be the same as issue_width. so 3.

Can we set MAX_THREAD as some numbers?

Yes, you can set MAX_THREAD as 5 (for example.) 5 should be good enough.

Do we have to worry about memory disambiguation problem in this assignment?

No, We assume that there is a perfect memory disambiguation predictor.

What should I use to find out memory instructions? is checking mem_type sufficient or do I have to check opcode?

if opcode is OP_ST mem_type is MEM_ST and if opcode is OP_LD mem_type is MEM_LD. so just checking mem_type itself is sufficient.

What will be the cache miss penalty? Is this KNOB_DCACHE_HIT_LATENCY+KNOB_MEM_LATENCY or just KNOB_MEM_LATENCY?

It is KNOB_DCACHE_HIT_LATENCY+ KNOB_MEM_LATENCY

Do we need to implement load-store forwarding?

You could. but you don't have to do.

Do we need to differentiate load and store?

no, for this assignment, you can treat them equally,