Karteikarten
What is Computer Hardware?
Karten in diesem Stapel (25)
- 1In a PC, there tend to be two chip-sets: the northbridge, and the southbridge. What is the essential difference between t he two?The northbridge handles fast communications amongst the CPU, the memory and the GPU, whereas the southbridge handles slower communications involving external hard disks, the mouse, the keyboard, the internet, the printer and other such devices.
- 2What are the two main components of an integrated circuit?Transistors interconnected by microscopic wires.
- 3What is the current trend in microprocessor design so as to overcome difficulties with power dissipation as we build faster and faster single processors?Multi-core processors where one CPU with a high clock speed is replaced by a number of CPUs with lower clock speeds but which, when working together, can give better computational power.
- 4What is Moore's law?A rough description that long-term transitor capacity doubles every 18-24 months.
- 5What is a Boolean function? In relation to Boolean functions, what is so special about NOT, AND and OR gates?A boolean function is a function f : {0,1}^n -> {0,1}. We can build a circuit to compute any Boolean function using only NOT, ANd and OR gates. ie they are a functionally complete set.
- 6What is a NAND gate? Show how a NOT-gate, AND-gate and an OR-gate can be constructed using just NAND gatesA NAND gate takes two outputs and outputs 0 iff both inputs are 1.
- 7Buil a circuit using NOT, AND and OR gates that computers the function with 3 boolean inputs and where the output is 1 iff there are an even number of 1s in the input.
- 8What are a half adder and a full adder? Show how a full adder can be built using 2 half adders.A half adder takes two inputs x and y and computes the boolean sum of x and y, and also the resulting carry bit (0+0=0,0+1=1,1+0=1,1+1=1+C). A full adder takes x, y and z as inputs and computes the boolean sum, resulting in the sum bit and the carry bit.
- 9Explain the basic principle behind using half adders and full adders to build a circuit that computes the sum of two n-bit strings.Let the two n-bit strings be X1X2...Xn and Y1Y2...Yn. A half adder takes X1 and Y1 as input and outputs the sum, with the carry fed into a full-adder. This full adder also has inputs X2 and Y2, and its sum as the output, with its carry fed into a full-adder. This full adder also has inputs X3 and Y3, its sum as the output, with its carry fed into a full adder and so on. BOth the sum and the carry of the final full adder are output.
- 10What are the three fundamental phases of integrated circuit design? And what does each phase do?A functional specification of the chip is derived (this describes exactly what the IC is supposed to do - involves chip aera, power, speed, and cost). The register transfer level design is then undertaken (the FS is used to describe the exact behaviour of the digital circuits on the chip, as well as the interconnections to inputs and outputs, with this description being in terms of logic gates and interconnecting wires).
Then the RTL is mapped to a physical layout in silicon (specific physical attributes must be respected, eg, it is crucial that appropriate spacing between transistors is maintained in the physical layout). - 11What is the ultimate aim of research in formal methods?The ultimate aim is to enable us to mathematically prove properties of designs, programs, and so on, so as not to just rely on empirical testing.
- 12A hardware description language is used so as to better enable the design of integrated circuits. Name two distinctive attributes of computer hardware that are normally expressible in a hardware description languages.Unlike a normal programming language, an HDL includes explicit notations for expressing time and concurrency.
- 13Name three key components in a CPU microarchitecture and briefly describe the purpose of each.The datapath: performs all the data processing operations and includes the ALU, which is the part of the CPU that performs all the arithmetic and logical operations on data, and a limited number of memory locations called registers.
The control: tells the datapath, memory, and I/O devices what to do (conduit bet ween datapath and main memory)
A cache: consists of small, fast, expensive on-chip memory and is used to store memory items that need to be regularly accessed. - 14Name two types of bus within a CPU and explain the general purpose of each. What is the width of a bus? Explain how the width of a bus imposes memory or data limigations within a CPU.Data bus: carries contents of memory locations between processor and main memory
Address bus: holds addresses of locations in main memory.
The width of a bus is hte number of parallel wires in it. The width of a data bus determines the word size of the computer; the width of the address bus determines the size of addressable memory. - 15What is the von Neumann bottleneck? Which component of a modern CPU did the von Neumann bottleneck give rise to?vN bottleneck is a limitation of the rate of data transfer between teh CPU and memory (data and instructions have to be fetched in sequential order and idle time is wasted while waiting for data titems or instructions to be fetched from memory). the vN bottleneck gave rise to the use of caches.
- 16how does the Harvard architecture differ from the von Neumann architecture?The Harvard architecture has memory that is partitioned into data memory and instruction memory, with dedicated buses for each. vN has unpartitioned memory.
- 17Explain the difference between a byte and a word in relation to computer architecture. How is the size of a word related to the width of a bus?1 byte = 8 bits. A word, being a number of bytes, is the amount of data that can be handled by a processor as one unit. A word size is the same as the width of the bus that carries it eg the data bus.
- 18What is a basic principle as regards different types of memory and its physical distance from the CPU?The cost and performance of memory is generally proportional to its physical distance from the CPU. (inversely surely?)
- 19What is the difference between static RAM and dynamic RAM?DRAM is where a bit of data is stored using a transistor/capacitor combination. SRAM is where a bit of data is stored by a flip-flop, which incorporates 4-6 transistors. SRAM is stable and fast but takes up more memory. DRAM is cheap, slow, and needs to be refreshed because of "leaky" capacitors.
- 20What is the purpose of cache memory in a CPU?Caches are expensive memory that are used to store rapidly accessed items.
- 21What are registers in the CPU?Registers are on-chip memory locations that are limited in number. They provide the fastest way to access data.
- 22Explain carefully the four phases of the fetch-decode-fetch-execute processor cycle. Explain the purpose of any CPU components you happen to mention.The instruction fetch stage involes the supply of the instruction address (via the address bus) and the return from memory (via the data bus) of the instruction.
The instruction decode phase involves interpreting the stored instruction within the CPU
The operand fetch phase involves the supply of the address of any required data (via the address bus) and the return from memory (via data bus) of this data.
The execute instruction phase involves the CPU performing the necessary actions (this phase is sometimes split into two with the execute instruction phase followed by a write-back phase where data is written back to memory if need be). - 23Show different layers of abstraction in a modern computer system.Software: { applications
algorithms and programs
programming languages
operating systems }
instruction set architecture
hardware: { microarchitecture
gates/register level transfer}
electrical engineering: {circuits
devices
physics} - 24Modern parallel computing can come in many shapes and forms. Briefly explain the fundamental difference between multi core computing and GPGPU computingA multi core computer is a computer where more than one processor is integrated on one IC. GPGPU computing uses the processors within the GPU to compute with rather than deal with the screen pixels.
- 25Give two illustratios of principles of CT in the context of hardwareProcessing in parallel in multi-core computing. Prefetching and caching in anticipation of future use in caching within CPUs.
Informationen
- Beschreibung
- Karteikarten am What is Computer Hardware?, erstellt von SEZ McKelvey am 26/03/2018.
- Erstellt von
- SEZ McKelvey
- Aufrufe
- 2
- Bewertung