Mill Computing, Inc. Forums The Mill Architecture Belt saturation in short belts Reply To: Belt saturation in short belts

Ivan Godard
Post count: 629

This is a test case to confirm that the specializer can handle belt overflow. It is deliberately simplified to the minimum necessary to expose the event of interest. Belt overflow also occurs in real codes, but real codes have so much else going on that it’s hard even for us to understand the behavior.

When can you play with it? Waiting is!