Also known as: Server blade

A blade server is a stripped-down server module that slides into a shared enclosure; the enclosure supplies power, cooling, and networking to many blades at once.¹

Overview

Where a rack server is a complete, self-contained machine, a blade keeps only the essentials — CPU, memory, and often local storage — and offloads power supplies, fans, and switching to the chassis. A single enclosure can hold a dozen or more blades, cutting cabling and power overhead and packing more compute into each data center cabinet. The trade-off is vendor lock-in to that enclosure and a larger up-front cost, so blades pay off mainly at scale.

Trade-offs

Blades shine when you need many similar servers in a small footprint — for example as the compute pool behind heavy virtualization or a high-availability cluster. For a handful of machines, plain rack servers are simpler and cheaper. A GopherTrunk deployment almost never needs blade density; its bottleneck is RF capture at the antenna, not raw server compute.

Sources

1. Blade server — Wikipedia, on blade form factors and enclosures. ↩

See also

• Rack server
• Server
• Data center
• Dedicated server
• Virtualization
• High availability