On a 100 Mbps PROFINET network in RT mode, you can typically support up to 100-200 IO Devices with Update Times of 10 ms without saturation, provided the topology stays reasonable (no more than about 4 hops).
True
Classical sizing of a 100 Mbps PROFINET RT network gives a cyclic bandwidth of roughly 10-100 kbps per device depending on the I/O count, so 100 devices at about 50 kbps each consume only 5 Mbps, i.e. 5 % of the 100 Mbps pipe, which leaves a comfortable margin. The real limit is not raw bandwidth but the number of hops and the cumulative store-and-forward latency they introduce. Beyond 200 devices, moving to 1 Gbps becomes cost-effective, and IRT requires a larger margin because time slots are reserved. Sizing tools such as Siemens SINETPLAN (free) or Hirschmann Network Sizing automate this calculation.
PROFINET bank in preparation
The full PROFINET bank isn't available yet. Drop your email to get notified at launch and grab an early-bird discount.
Join the waitlist →See the 9 other PROFINET practice questions
Related questions
- Which PROFINET real-time class delivers the strictest determinism (cycle <= 31.25 us with jitter <= 1 us), making it suitable for motion-control applications?2. Real-time classes · RT vs. IRT vs. NRT
- A PROFINET network running in IRT mode requires industrial Ethernet switches that are specifically IRT-certified and implement precise time-based scheduling (typically prioritised cut-through forwarding).2. Real-time classes · IRT-Switches
- The MRP (Media Redundancy Protocol) handles redundancy on a PROFINET ring using one "MRP Manager" node and several "MRP Clients", guaranteeing a failover time below 200 ms in the event of a cable break.3. Topology & cabling · MRP (Media Redundancy)
- PROFINET "S2 redundancy" denotes an architecture in which two IO Controllers (one primary, one standby) are connected to the same IO Device, with automatic failover in case of primary failure and without loss of communication.3. Redundancy · S2-Redundanz
- A PROFINET IO Device can spontaneously raise a "Diagnostic Alarm" to its Controller in the event of an anomaly (sensor cable break, overheating, short-circuit), without waiting for a read request.4. Diagnostics · Diagnose-Alarme