TYPES OF PROCESSOR Flashcards
(15 cards)
Describe how parallel processing would increase the performance of the computer system.
- Parallel processing allow multiple program instructions to be executed concurrently
- Different CPU cores can tackle different sections of a job simultaneously
- Longer rendering jobs can be split, increasing the speed of execution
Describe what is meant by the term ‘RAM’
- Random Access Memory
- A form of primary memory
- Used to hold data and / or programs in use
- Volatile / loses its contents when power is lost
Explain why GPUs are more suitable than CPUs for graphics processing
- CPUs are general purpose processors
- Whereas GPUs are coprocessors designed specifically for graphics
- They are parallel processors which are not well suited to general multitasking duties performed by the CPU
- And so likely to have built in circuitry / instructions for common graphics operations
- GPUs are able to perform and instruction on multiple pieces of data at one time
- Often we want to do this when processing graphics (e.g. transforming points in a polygon or shading pixels)
- Which means it can perform transformation to onscreen graphics quicker than a CPU
Describe the differences between CISC and RISC architectures
- CISC has a larger instruction set
- RISC has a smaller instruction set
- CISC is difficult to pipeline
- RISC is easier to pipeline
- CISC tends to have more addressing modes
- RISC tends to have fewer addressing modes
- CISC instructions may take multiple clock cycles to execute
- RISC instructions take one clock cycle to execute
- RISC can only do complex tasks by combining multiple instructions
- CISC can do complex tasks in one line
- In assembly for CISC, a statement takes one mnemonic
- May take multiple mnemonics in RISC
- Some instructions in CISC will rarely get used
- In RISC instructions get used regularly
- CISC has more complex circuitry / more transistors
- RISC has less complex / simple circuitry
- CISC uses less RAM
- RISC uses more RAM as programs tend to be larger
Describe the suitability of CISC and RISC for different devices
- RISC has smaller instruction set
- Requires fewer transistors / less complex circuitry
- Means less power is required
- No cooling
- Smaller battery
- Longer battery life
Describe non-graphical uses for GPUs
- Performing complex numerical calculations
- Calculations on matrices / vectors / multiple data at the same time
Give examples of uses for the GPU
- To render models
- Run CAD software
- Run modelling calculations
Explain the effect of using a co-processor for:
Complex calculations for scientific research
Printing personalised letters
- Calculations are done by the maths co-processor
- So processing is faster
- When using floating point arithmetic
- No increase in speed
- As co-processor not suitable for task / as there are no calculations
State the differences between RAM and ROM
- RAM is volatile / ROM is not volatile
- RAM is read and write (editable) / ROM is read only (cannot be altered)
- RAM is larger / ROM is smaller
State what is stored in RAM and give a reason why it is used
- User files / programs / OS currently in use
- User must be able to alter contents of the file / computer needs access to the software but needs to be able to replace it
- RAM offers direct access
- RAM operates at a much faster speed than most secondary storage devices
State what is stored in ROM and give reason why ROM is necessary
- Bootstrap / BIOS
- Must be available when the computer switched on (therefore must be stored on medium which is non-volatile)
- The bootstrap / BIOS must not be deleted / unintentionally amended ( and therefore is best stored on a read-only medium)
Explain the advantages of storing an operating system on ROM
- ROM is quick to start up so the system can be started up quickly
- ROM cannot be altered so there is no change of the OS being accidentally or maliciously changed
Discuss the different approaches that can be used to provide increasingly larger amounts of computing power and the types of problem they are suited to
- Processors have increasingly large clock speeds and can be overclocked
- Processors can have multiple cores
- Supercomputers can have multiple processors (and GPUs)
- GPUs can be applied to problems other than graphics processing
- Problems can be distributed across a number of computers working together
- Having multiple cores can speed up smaller problems but this will not be enough for larger problems
- Supercomputers are prohibitively exceptionally expensive to buy and run for all but large organisations
- GPUs are becoming a cost efficient way of tackling problems
- GPUs tend to have large number of cores so can run on highly parallelisable problems
- But only where the same instruction is being applied to multiple pieces of data (SIMD)
- Increased clock speed is limited to smaller problems
- Even doubling the clock speed would only halve the time taken
- Parallel Processing isn’t suited to all problems
- Most problems are only partially parallelisable
- Writing algorithms for parallel processing is more challenging than GPUs
- GPUs suited to a subset of science / engineering problems where the same calculation is repeated on multiple data sets
Other than the maths co-processor, state other types of coprocessors and describe why they may no longer be used
- I/O coprocessors in mainframe systems
- Coprocessors managing network operations (NPUs)
- Many of the examples are now obsolete as CPUs have become more comprehensive and functionally more comprehensive
What is a coprocessor
Any additional processor used for a specialised task