Scale Up Server
What is a Scale Up Server?
A scale up server increases its workload capacity by adding memory and processing power, i.e., building “up” rather than “out,” which would mean adding server units. There are two fundamental approaches to addressing increasing workloads: "Scale up" meets the challenge through a bigger, more powerful server with added memory and compute capacity. This vertical building "up" of the server contrasts with the “scale out” approach, which adds server units to the workload horizontally. A scale up server is one that is designed to enable the simple and cost-effective addition of system resources to itself, enabling the "scale up" approach to work.
Why scale up servers?
Scale up servers are valuable in a number of scenarios, including:
- When the workload itself is best performed on fewer machines, each with a high level of compute power
- When physical space and/or electrical power is limited and a small number of high-powered servers are more suitable than a large number of low-powered ones
- When software licensing factors favor the use of fewer servers
- When the complexity of setting up a “scale out” collection of servers is not justified by the workload
HPE scale up servers
The HPE ProLiant family of servers offers some of the industry’s best scale up features. These include the HPE ProLiant DL560 and BL660c Gen9 Servers, which provide dense compute, right-sized storage and a large memory footprint. These two servers took the #1 4P world records for both max-jOPS and critical-jOPS in the MultiJVM category on the SPECjbb20152 benchmark. The DL560 Gen9 is the world’s most energy-efficient 4P server, winning three world records on the SPECpower_ssj20082 benchmark.