What is Scale-Up and Scale-Out

Home - What is Scale-Up and Scale-Out

SCALE-UP

Scale-Up or Vertical Scaling (Stacking up servers vertically on a single compute node)

Scale-Up is done by adding more resources to an existing system or to a compute node to increase the performance of the node. Scaling up provides a large memory foot-print across multiple CPU sockets (8S, 16S, 32S, 64s, etc.). A higher number of CPU sockets will then constitute a larger system with correspondingly larger memory to handle applications with large memory and compute requirements. Among the applications that benefit from Scale-Up systems are SAP/Hana, transactional databases, scientific simulations, Advanced Analytics, Real-time Streaming, AI, etc. A common requirement for all of those applications is to have access to large amounts of memory across the CPUs (sockets) in a cache coherent manner. Conceptually, a large number of CPUs are vertically connected with node controllers to achieve 2X performance and more.

SCALE-OUT

Scale-Out or Horizontal Scaling (Distributed connection of servers over many compute nodes)

Scale-Out is a distributed computer architecture. Typically, a compute node consists of 2-4 CPU sockets. A number of such nodes are then horizontally connected in non-cache coherent way through a network fabric. The memories are not visible across nodes. It is not possible to obtain a single memory image (across sockets) as we get in Scale-Up systems. This will limit throughput and capacity for all applications that are memory “hungry” and compute intensive.