Parallel hardware about karuta HOT 2 OPEN

Thank you for your encouraging comment!

I'am afraid that loop unrolling is not implemented yet, but it should be interesting. I am going to implement it probably from next month or so.

Karuta's priority until now was to describe parallelism between threads (FSMs) and between objects (modules). Possible alternative solutions are to clone objects or to create a set of replica threads like:

  @(num=4)
  process f(threadIndex int) { // 4 FSMs are generated
    for var i = 0; ... { ... }
  }

However, this might cause some overhead by the arbitration between shared resources. So Please stay tuned for a while.

from karuta.

Ah okay haha yes I think I've tried to do that but the resulting verilog did not seem too efficient.

Alright then, I'll patiently await the arrival of the loop unrolling. Thank you so much for entertaining this request and please take care!

from karuta.