Chapter 3 - Types of processor Flashcards
What is the stored program memory concept?
» Machine code instructions are fetched and executed serially by a processor that performs arithmetic and logical operations
How is the machine code fetched in Von Neumann’s architecture?
1 Mark
» Are fetched from memory one at a time, decoded and executed in the processor
Name 5 descriptions of Von Neumann’s architecture.
5 Marks
» Shared memory space for instructions and data
» Can access instructions and data required using a dedicated bus
» Single Processor CU manages program control and single ALU
» Same bus used for transferring both instructions and data
» Uses FDE cycle to excute one isntruction at a time in a linear sequence
What is the one problem of Von Neumann’s architecture and the 2 advantages?
» The data bus is slower than the rate at which the CPU can carry out instructions - Von Neumann Bottleneck - CPU has to wait for the data transfer as it is much faster
» Simpler operating system
» Easy to program
What is Harvard architecture extensively used in?
» With embedded Digital Signal Processing (DSP)
What is DSP?
» The DSP applications include audio and speech signal processing, sonar and radar signal processing and digital image processing
What are embedded systems?
1 Mark
» Special purpose computers built into devices often operating in real-time, such as those used in navigation systems
Name 4 descriptions of Harvard architecture.
4 Marks
» Memory is split into 2 parts
» One part for data and another part for program instructions
» Each part is accessed with a different bus
» Allows for pipelining
Why is data and instructions being fetched with a different bus an advantage?
1 Mark
» Allows the CPU to pipeline as the CPU can fetch both data and instructions at the same time
Can Harvard architecture be faster than von Neumann’s architecture?
» Yes because data and instructions can be fetched in parallel instead of competing for the same bus and avoids von Neumann bottlenecks
What is the one comparison between the Harvard and Von Neumann architecture in terms of where they are conventionally used?
1 Mark
» Von Neumann - used in conventional processors in PC’s
» Harvard used in DSP
What is the one comparison between Harvard and Von Neumann’s architecture in terms of memory?
1 Mark
» Von Neumann - data and programs share the same memory
» Harvard - the instructions and data are held in separate memories
What is the one comparison between the Harvard and Von Neumann architecture in terms of buses?
1 Mark
» One bus is used to transfer data and instructions - Von Neumann
» Parallel data and instruction buses may be used - Harvard
What is the one comparison between Harvard and Von Neumann’s architecture in terms of size?
1 Mark
» Programs can be optimised in size - Neumann
» Programs tend to be large - Harvard
What does CISC stand for?
1 Mark
» Complex Instruction Set Computer
What is the instruction set?
1 Mark
» Set of all instructions written in machine code, that can be recognised and executed by a given CPU
What is the aim of a CISC processor?
1 Mark
» Aims to complete the task in few lines of code as possible
What is complex instruction?
» Combines a load/store instruction with the instruction that carries out the actual calculation
With a CISC processor, what is so special about the hardware?
» Processor hardware and circuitry has to be more complicated so it can understand and execute a series of operations within one instruction
» Therefore more expensive
How many clock cycles do complex instructions take to execute and why?
» Takes more than one clock cycle, because its performing multiple operation
What are the three advantages of a CISC processor?
3 Marks
» Quicker to code programs
» Compiler has very little work to do to translate a high-level language statement into machine code
» Since code is short - requires less RAM is required to store the instructions
What are the three disadvantages of a CISC processor?
3 Marks
» They are larger as they require more transistors, making them more expensive
» Take multiple cycles per line of code, therefore programs run more slowly due to comploicated circuit
» Pipelining is not possible
What does RISC stand for?
1 Mark
» Reduced Instruction Set Computer
What does RISC aim to do?
1 Mark
» Aim to use simple instructions that will be executed within a single clock cycle
How are complex commands such as MULT represented in a RISC processor?
» Separated into a number of simpler commands
Why does the compiler have to do more work in a RISC processor?
1 Mark
» Since there are more lines of code
Why do the RISC instructions require fewer transistors?
1 Mark
» Because they have less complex hardware
What are the 4 advantages of a RISC processor?
4 Marks
» Simpler processor designe
» Easier to pipeline - As fixed length of codes
» Decode unit has a lower power consumption
» Faster than CISC as programs run faster due to simpler instructions
What are the 2 disadvantages of a RISC processor?
2 Marks
» Compiler has to do more work to translate high-level code into machine code
» More RAM is required to store the machine code
Where are RISC and CISC architectures used?
2 Marks
» RISC is used mainly in portable devices
» CISC although less common, are used in Desktops and PC’s
What is a co-processor?
1 Mark
» Is an extra processor used to supplement the functions of the primary processor for a specialised task, and carries out only a limited range of functions
What is a Multi-core processor?
2 Marks
» Single-chip containing more than one independent processing units
» Each core can fetch decode and execute an instruction independently
» Focuses efforts of multiple CPUs into 1 task
What is parallel processing?
1 Mark
» Multiple processors are executing more than one task at the same time by dividing the program instruction between the multiple cores
What is a GPU?
» A specialized electronic circuit that is very efficient at manipulating computer graphics and image - processing
» Contains specialised processing cores, created to render images
» Can work quickly over a large block of similar data - Due to having large number of cores - which are higly paralleiziable
What is a GPU a form of?
» A Co-processor which can be used with a CPU to acclerate performance
What is one difference between multi-core systems and parallel processing?
1 Mark
» They accomplish a similar task however instead of requiring multiple cores they can complete task with a single core by using threading
Compare RISC and CISC
5 Marks
» CISC - Each instruction takes more than one clock cycle to execute - RISC - Each instruction executes in one clock cycle
» CISC - High power consumption - RISC - Low power consumption
» RISC - Simple processor design - CISC - Complicated processor design
» RISC - Complex task can only be performed by combining multiple instructions - CISC - Complex instruction can be performed using less instructions
» RISC - Pipelining is possible - CISC - Pipelining is not possible
What is concurrent processing?
2 Marks
» One processor is executing more than one task seemingly at the same
» With each processing in turn being given a slice of processor time
What are some limitations of parallel processing?
3 Marks
» Algorithm must be suitable for parallel processing
» Increased complexity of the control unit, which means increased time spent managing processes instead of executing them
» Potential for deadlock, where one process requires the output of another
What are 2 ways parallel processing can be achieved in modern computer systems?
2 Marks
» Multiple processors within a single computer, each dedicated to a specific task, such as a GPU
» Multiple cores inside a single CPU, running multiple process simultaneously
Why is a GPU more suited for processing graphics than a CPU?
4 Marks
» CPUs are general purpose processors, whereas GPUS are designed specifically for graphics, so are likely to have built-in circuitry for graphic operations
» GPUS are able to perform an instruction on multiple data locations at the same time
» We want to do this when processing graphics
» Which means it can process and perform transformations to onscreen graphics faster than the CPU
What is SIMD?
2 Marks
» Single instruction, multiple data, is a type of parallel processing
» Hardware components that perform the same operation on multiple data operands concurrently
What is an array processor architecture?
» Single instruction multiple data
» Allows same instruction to operate simultaneously on multiple data locations
Other than graphics, state 2 additional uses of a GPU, and why is it more effective than a CPU at the specific task?
» Cryptocurrency mining, Audit processing,machine learning
» Process involved calculations on large data sets
» Single instruction multiple data
» Specialist hardware/ array processing than a CPU
Why does simply not doubling the number of cores not double the performance?
» Some programs cant make maximum use of all cores
» Overheads involved with inter-core communication
What is MMID?
2 Marks
» Multiple instruction multiple data
» Different instructions carried out at the same time on different pieces of data
What does a GPU Process?
» Involved processing many calculation on large data sets SMID
What are 2 benefits of a multicore system and one disadvantage?
3 Marks
» More jobs can be carried out in a short time because they are executed simultaneously
» Task can be shared between processors to reduce the load on individual processor and avoid bottlenecks
» Hard to program code to decompose problems efficiently for multicore processing
What is a CPU bottle neck?
» A bottleneck occurs when a PC is running a demanding application and its performance begins to stall.
Why do supercomputers have multiple GPUs and cores?
» Cost-effective method
» GPU tend to have a large number of cores so can run highly parallel problems
Give one advantage and one disadvantage, other than cost, of using Von Neumann compared with array processor architectures?
2 Marks
» Simpler operating system
» Slower than array processing on large data sets
What are the 2 different approches to parallel processing?
2 Marks
» SIMD
» MIMD
Why is a supercomputer useful?
» CPUs can work in parallel
» On the same problem
Which part of the FDE cycle will the accumulator recieve a value from the ALU?
1 Mark
» Execute
What might a cotemporary processor contain?
» Both a harvard architecture and a Von Neumann architecture
What is mean by the term Von Neumann architecuture?
2 Marks
» Uses same memory for data and instructions
» Uses same bus for data and isntructions
Where would an interrupt be handled?
1 Mark
» After the decode phase
Describe one difference between CISC and RISC
2 Mark
» CISC architecture has more complex circuitry, whilst RISC has simple circuitry minimising manufacture cost
What does BRA do?
» Accumulator checked to see if value held is positive or zero
» If so BRANCH carried out/jumps to specificed location
What is a thread?
» A sequence of instructions that have been sent to the CPU to be processed
What is the main benefit of multithreading?
» Allows 2 threads to be run on each core
What is one use of a GPU?
» Modelling
