Introduction Flashcards
What does High-Performance Computing (HPC) usually refer to?
The practice of aggregating computing power to deliver much higher performance than a typical desktop computer for solving large problems in science, engineering, or business.
Why is HPC needed in physical sciences and engineering?
Many realistic problems generate complex systems of equations that cannot be solved analytically, requiring numerical solutions.
Name an example of an organization that heavily uses HPC in the UK
The Met Office (for weather forecasting, climate change studies, and oceanography)
How is HPC used in astrophysics?
To simulate star formation and analyze telescope data.
What is a “digital twin” in computational engineering?
A virtual representation of a physical system that is continuously updated with real-time data for simulations and analysis.
How is HPC applied in life sciences?
Genomics, including DNA sequencing and protein structure analysis.
What is the Top500 List?
A ranking of the world’s fastest HPC systems based on High-Performance Linpack (HPL) benchmark performance.
How often is the Top500 List published?
Twice a year.
What does the Linpack benchmark measure?
Floating point operations per second (FLOPS) using a dense linear algebra workload.
What is HPL?
High-Performance Linpack: a parallel version of Linpack used for ranking HPC systems.
How has HPC performance changed over time?
It has grown exponentially since the 1990s
Why is parallelism essential in HPC?
HPC calculations require a vast number of floating-point operations, and running them in parallel makes the solution feasible.
Why is parallel programming harder than serial programming?
It requires ensuring both correctness and efficiency while managing synchronisation and communication between tasks.
Why can’t we rely on just one really fast processor?
Due to physical limits: power consumption, heat dissipation, and the breakdown of Dennard scaling.
What is Moore’s Law?
The number of transistors on a chip doubles approximately every 24 months.
What is Dennard Scaling?
A principle stating that as transistors get smaller, power density remains constant, allowing chips to become faster and more energy efficient—though this has now broken down.
What is a commodity cluster?
An HPC system built using commercially available components rather than specialized hardware.
How does a massively parallel processing (MPP) system differ from a commodity cluster?
MPP systems use more specialised hardware and are less reliant on off-the-shelf components.
What is the most common HPC architecture today?
Commodity clusters
Name an example of an MPP system
- Fugaku
- IBM BlueGene
- Cray XC Series
What is the University of Exeter HPCL Isca?
A commodity cluster using Intel x86 compute nodes, Infiniband interconnect, and a GPFS parallel filesystem.
What are the four main components of an HPC cluster?
- Compute nodes
- Interconnect
- Mass Storage
- Login Nodes
What is the function of compute nodes?
They provide the processor cores and memory required to execute workloads.
What is the purpose of an interconnect in an HPC cluster?
It enables compute nodes to communicate and access storage.
