robertanderson
Member
I’m planning to pick up some QSFP-40G-SR-BD modules to use 40G over existing duplex multimode fiber, and while most folks talk about compatibility or link stability, I keep running into some slightly different questions that I haven’t seen good answers for yet. Before I order, I figured it would be great to get real feedback from people who have actually deployed these.
Here’s what I’ve been wondering and haven’t seen much discussion about:
1. Does SR-BD affect link latency compared to MPO SR4?
I know SR-BD and SR4 are both 40G, but does anyone actually measure a difference in latency or performance in real traffic patterns? Does SR-BD introduce any processing overhead on switches compared to a straight SR4 run?
2. Do patch panels/duplex jumpers ever cause unexpected polarity quirks?
With SR4 and MPO you already have to be careful about polarity. I am curious whether SR-BD over duplex fiber triggers weird transceiver alarms (like mis-termination warnings) on some switch OSes if the fiber isn’t perfectly balanced or swapped.
3. Does using SR-BD limit how future upgrades are handled?
For example, if we ever want to break out 40G to 4x10G on the same fiber plant later using breakout modules, does having SR-BD in place now make it harder or more expensive to transition?
4. Any behavior differences under thermals or high-density ports?
Since these are often used in dense switches, does SR-BD run hotter or draw more power than traditional SR4 optics? And if so, do certain chassis or vendor platforms handle that better than others?
If you’ve used QSFP-40G-SR-BD in production and seen any of these questions play out, I’d really appreciate your input. Does SR-BD behave as smoothly as SR4 once it’s connected, or are there subtle things that only show up in real world deployments? Thanks!
Here’s what I’ve been wondering and haven’t seen much discussion about:
1. Does SR-BD affect link latency compared to MPO SR4?
I know SR-BD and SR4 are both 40G, but does anyone actually measure a difference in latency or performance in real traffic patterns? Does SR-BD introduce any processing overhead on switches compared to a straight SR4 run?
2. Do patch panels/duplex jumpers ever cause unexpected polarity quirks?
With SR4 and MPO you already have to be careful about polarity. I am curious whether SR-BD over duplex fiber triggers weird transceiver alarms (like mis-termination warnings) on some switch OSes if the fiber isn’t perfectly balanced or swapped.
3. Does using SR-BD limit how future upgrades are handled?
For example, if we ever want to break out 40G to 4x10G on the same fiber plant later using breakout modules, does having SR-BD in place now make it harder or more expensive to transition?
4. Any behavior differences under thermals or high-density ports?
Since these are often used in dense switches, does SR-BD run hotter or draw more power than traditional SR4 optics? And if so, do certain chassis or vendor platforms handle that better than others?
If you’ve used QSFP-40G-SR-BD in production and seen any of these questions play out, I’d really appreciate your input. Does SR-BD behave as smoothly as SR4 once it’s connected, or are there subtle things that only show up in real world deployments? Thanks!
