JTCat
Member
I am working on a CLX that has to message tag values to and from a DH+ network that is somewhat congested. It is in the works to remove nodes from the DH+ network and put them on Ethernet with 5/05's.
Am I correct in assuming that 'write' instructions use less bandwidth than 'read' instructions?
It seems to me that a read instruction has to communicate both ways to retrieve data. First being to request the data from the other processor, then the data has to be sent back to the requesting processor.
Whereas a write instruction just blasts the information out there without regard to whether there is a processor listening for it or not.
I understand the whole troubleshooting read vs writes, but that's not an issue in this regard. I just want to know if using write operations is less bandwidth intensive than read operations.
Thanks,
JTC
Am I correct in assuming that 'write' instructions use less bandwidth than 'read' instructions?
It seems to me that a read instruction has to communicate both ways to retrieve data. First being to request the data from the other processor, then the data has to be sent back to the requesting processor.
Whereas a write instruction just blasts the information out there without regard to whether there is a processor listening for it or not.
I understand the whole troubleshooting read vs writes, but that's not an issue in this regard. I just want to know if using write operations is less bandwidth intensive than read operations.
Thanks,
JTC