Network Bandwidth & Throughput Calculator
Calculate theoretical bandwidth, actual throughput, file transfer time, and network efficiency based on connection parameters and protocol overhead.
TCP/IP overhead ~20%, Ethernet ~7%, combined ~25%
Practical max utilization before congestion
Ping / RTT in milliseconds
Formulas Used
Effective Throughput:
Throughput = Bandwidth × (Utilization / 100) × (1 − Overhead / 100)
Transfer Time:
Transfer Time = File Size (bits) / Throughput (bps)
Bandwidth-Delay Product (BDP):
BDP = Bandwidth (bps) × RTT (s)
Represents the amount of data "in flight" on the network at any moment — the pipe volume.
TCP Maximum Throughput (window-limited):
TCP Throughput_max = Window Size (bits) / RTT (s)
Derived from the TCP sliding window mechanism. If Window Size < BDP, TCP cannot fully saturate the link.
Network Efficiency:
Efficiency (%) = (1 − Overhead / 100) × (Utilization / 100) × 100
Assumptions & References
- Bandwidth values use SI decimal prefixes (1 Mbps = 1,000,000 bps); file sizes use SI as well (1 MB = 1,000,000 bytes = 8,000,000 bits).
- Protocol overhead of ~20% is typical for TCP/IP over Ethernet (IP header 20 B, TCP header 20 B, Ethernet frame 26 B on a 1500 B MTU ≈ 4.4% framing + TCP/IP ≈ 2.7% → combined with ACKs, retransmissions, and control traffic ≈ 15–25%).
- Practical link utilization is capped below 100% to avoid congestion collapse; 70–80% is a common engineering rule of thumb (Cisco best practices).
- The TCP window-size formula is a simplification of the Mathis equation: Throughput ≤ MSS / (RTT × √p) where packet loss p → 0 reduces to Window/RTT.
- BDP concept from Jacobson & Karels (1988); TCP window scaling defined in RFC 7323.
- Latency (RTT) includes propagation delay, transmission delay, queuing delay, and processing delay.
- This calculator does not model multi-path, QUIC, or UDP-based protocols.