Calculate the composite EIGRP metric from bandwidth, delay, reliability and load.
EIGRP computes a composite metric from the slowest (minimum) bandwidth along the entire path and the cumulative delay of every link in the path, combined using tunable K-value weights. With Cisco's default K-values (K1=1, K3=1, and K2=K4=K5=0), the formula reduces to 256 times the sum of a bandwidth term (10,000,000 divided by the minimum bandwidth in Kbps) and a delay term (the total delay in tens of microseconds). Because it uses the single slowest link rather than summing bandwidth, and adds cumulative delay, EIGRP's metric reflects real path characteristics more richly than RIP's simple hop count, while remaining computationally lighter than full link-state flooding.
EIGRP composite metric (default K-values)
metric = 256 × ((10^7 / min_bandwidth_Kbps) × K1 + (total_delay / 10) × K3)
A multi-hop path can never move data faster than its slowest link — that link is the bottleneck for the entire path — so EIGRP correctly uses only the minimum bandwidth encountered along the route rather than summing or averaging bandwidths, which would misrepresent the path's true throughput ceiling.
K1 through K5 weight the contribution of bandwidth, load, delay, reliability, and MTU respectively to the composite metric formula. Cisco's default enables only K1 (bandwidth) and K3 (delay), effectively ignoring load, reliability, and MTU unless an administrator explicitly changes the K-value profile — all routers in an EIGRP domain must use matching K-values or they cannot form neighbor relationships.
The scaling factor of 256 originates from EIGRP's predecessor, IGRP, which used an 8-bit (0-255) metric internally; EIGRP retained backward compatibility with IGRP by scaling its wider composite metric by 256, preserving the ability to redistribute routes between the two protocols with a consistent conversion factor.