HPC (High Performance Computing) is data processing with very high performance which means they can do a high volume of calculations in a short amount of time. Actually more than one million times faster than the fastest commodity desktop, laptop or server systems according to IBM.
An HPC data center has powerful processors, a high density of servers and large cooling requirements. It has been specially developed to handle this enormous power capacity, which in one rack today can be over 50 Kilowatts and that increases in line with better equipment.
Today, many HPC data centers are connected to large private and public companies that need a lot of computing power. One example is “supercomputers”, which among other things, are used for simulations, research and other data processing. However, there are several private HPC data centers under development – like GreenBox™. Especially in Nordic countries such as Norway due to more favorable climatic conditions and more sustainable electricity generation.
Read more about our HPC solution here
What is high performance computing (HPC)?
HPC is the ability to process data and perform complex calculations at a very high speed. The combined computing power makes it possible to solve problems that would otherwise not be possible with ordinary computing systems.
For comparison, an average computer can perform 3 billion calculations per second. HPC solutions are capable of performing quadrillion calculations in one second.
How does HPC work?
HPC mainly consist of three components:
- Compute
- Network
- Storage
In HPC solutions, several data servers are connected together in a cluster. Compared to regular computing systems which run one task after another, HPC is able to run multiple tasks simultaneously through software and algorithms seamlessly.
What is an HPC cluster?
An HPC cluster is a group of servers connected in a network that works together as a single system. It’s usually hundreds or thousands of servers in the cluster that are connected together using high-bandwidth networking technologies. The nodes in each cluster work in parallel with each other and increase processing speed to deliver very high performance computing.
