Need for parallel sequence

Hi,

An interface is driven by a driver. This driver can drive a single transaction at a given point in time. If so, what is the need for parallel sequences as explained in OVM/Sequences/Generation | Verification Academy? Driver can’t drive both of them right? Can anyone used this knowing the significance, please explain?

Murali

You could use parallel sequences to interleave other accesses to the bus with the ones in the main stimulus stream in order to try to find problems with the DUT. This is sometimes referred to as irritant traffic.

With some agents, able to handle more than one transaction at a time, it’s a way of increasing the stress on the DUT.

In reply to mperyer:

I found the usage of this now. My OVC had to send Reqs and Resps to DUT in the same interface. But Resps are conditional; had to be sent only when OVC receives a Req from DUT. Having a resp_seq and running it forever in the higher level virtual sequence is good option. And then the req_seq can be randomly run. Thus the same sequencer schedules both Req/Resp items correctly.