1.1.2 Types of Processor

instructions that a computer is able to carry out

• larger instruction set with more complex instructions • instructions take more than 1 clock cycle to execute • more general purpose registers • instructions take up less memory

• smaller instruction set with simpler instructions • instructions take 1 clock cycle to execute (suitable for pipelining) • has fewer addressing modes • fewer registers

• uses less power • costs less to manufacture

• takes up less space in memory

graphics processing unit

any additional processor used for a specific task

to improve the overall speed of the comptuer by excecuting concurrently with the CPU

performing simple instructions on large data sets

performing complex instructions on small data sets

• CPUs are more general purpose (run faster generally) • GPUs are speciallised for performing instructions on large data sets eg. rendering graphics (faster for specific tasks)

• simple instruction multiple data • having multiple processing elements which perform the same operation on multiple data points simultaneously

• oil exploration • weather modelling • machine learning • linear algebra • statistical analysis • stock option predictions • gravitational field theory

a single chip that has two or more independent processing units

• multitasking - each core can process something different • background tasks - multicore processor can assign background tasks to one core which reduces their impact of speed on the other tasks • improved responsiveness - if a program becomes unresponsive, another core can continue to run the program to stop it slowing the computer down

the processing of program instructions by dividing them between multiple processors or processor cores

• increase speed and performance • problem solving - large problems can be solved more easily • real time applications eg. graphics rendering are improved

• limit on maximum speed - not all parts of a program can be parallelised • complex programs - harder to write code for parallel processing as data has to be synced and shared correctly • communication between processors takes time • limited applicability - not all programs can use the parallel processing • more difficult to debug because of the precise timing of events