To Refresh or Not to Refresh: x86 Servers

Author(s): Vince Londini

Enterprises strike a balance between application delivery and cost containment when establishing a server hardware refresh rate. While the survey average sits at four years for commodity x86 servers, barely one-fifth of enterprises actually refresh on a four year cycle.

Most Pick Three or Five

Figure 1 illustrates the two current schools of thought regarding server hardware refresh cycles. Smaller enterprises are more evenly split; there being almost no statistical difference between the first three groups in the “<501 Employees” bar. But, on the whole, enterprises tend to choose either a  three year or a five year cycle.