SLR2- 1.1: GPUs and their uses
No description yet. This curated deck is ready for daily study.
Card preview (22)
What is a co-processor?
A co-processor is any additional processor used for a specialised task
What is the purpose of a co-processor?
To improve the overall speed of the computer by executing concurrently (at the same time)
What does GPU stand for?
Graphical Processing Unit
What is a GPU?
A type of co-processor that is responsible for processing graphics within the computer to reduce the load on the CPU
What are the key differences between GPU’s and CPU’s
GPUs: * Slower processing power per core * Thousands & smaller number of cores * Simple ,repetitive operations CPUs * Faster processing power per core *Fewer & powerful number of cores * Complex operations
What can a GPU be used for besides graphics?
1) 3D modelling 2) Data modelling 3) Financial modelling 4) Data mining 5) Performing complex numerical calculations 6) Machine learning 7) Calculations of multiped data at the same time
How is a GPU used in 3D modelling?
To render lighting effects, textures and shadows
How is a GPU used in data modelling?
To handle many calculations at the same time, they can handle large data sets and complex operations (sorting + filtering data)
How is a GPU used in financial modelling?
To simulate different scenarios in risk modelling and option pricing * Lots of simulations can run in parallel
What is meant by data mining?
The process of analysis large amounts of data to find patterns
How is a GPU used in data mining?
It can carry out main computational tasks (.e.g. sorting, searching, pattern recognition)
What is meant by machine learning?
It is training a computer on a massive amount of data which can be done in parallel
How is a GPU used in machine learning?
To speed up the process of making new predictions on new data
How is a GPU used in calculations on multiple data at the same time?
It is done by GPUs instead of CPUs as GPUs are set up for parallel processing
What type of task are GPU’s suited for?
• Specialist instructions • Multiple cores • SIMD processing
Why are GPUs suited for specialist instructions?
• GPUs are designed to execute specialists instructions like in 3D rendering • These capabilities have been expanded and generalised over time so makes GPUs suitable for a wide range of complex calculations
Why are GPUs suited for multiple cores?
• GPUs have smaller cores optimised for parallel processing • GPUs can perform tasks at the same time • Useful in machine learning when large amounts of data need to be processes
Why are GPUs suited for SIMD processing?
GPUs support SIMD processing because they were originally designed to perform the same operations on multiple pixels/vertices at the same time
What are some of the benefits of GPUs?
• Parallel processing • Speed • Efficiency
Why is parallel processing a benefit of GPUs?
GPUs can handle many tasks at the same time as they are multiple core processors
Why is speed a benefit of GPUs?
GPUs use parallel processing which speeds up tasks * Specifically those with large amounts of data
Why is efficiency a benefit of GPUs?
GPUs can perform more calculations per unit of power than CPUS * Makes them more energy efficient when it comes to tasks