Computers and Architectures Flashcards
How are applications written
Applications are written using millions of lines of complex instructions
What can hardware execute
Hardware can execute only a simple set of predefined instructions
How many layers are needed to go from complex instructions to simple instructions
7
What is the order of operations from high-level language to hardware
High level language,
Assembly language program (for MIPS)
Binary Machine language program (for MIPS)
In the mid-1980s what was processor performance growth largely driven by
largely driven by technology
What rate was processor performance growing in 1986
52%, or doubling every 2 years
What was the growth in 1986 caused by
more advanced architectural and organizational ideas typified in RISC architectures
What happened between 2003 and 2011 to processor performance
In 2003 the limits of power due to the end of Dennard scaling and the available instruction-level parallelism slowed uniprocessor performance to 23% per year until 2011
What was the growth between 2011 and 2015 in processor performance
less than 12% or doubling every 8 years
What caused the slowdown in growth between 2011 and 2015
in part due to the limits of parallelism of Amdahl’s Law
Define Amdahl’s law
In other words, the performance improvement of a system as a whole is limited by its bottlenecks.
Define Moore’s Law
Moore’s law is the observation that the number of transistors in an integrated circuit (IC) doubles about every two years.
What is the growth currently
since 2015 just 3.5% per year or doubling every 20 years
What are Personal Mobile Devices
Wireless devices with multimedia interface such as Smart phones and tablet computers
What are the cons of personal mobile devices
o Cost
o Energy efficiency
o Size requirement
o Responsiveness & Predictability
What are desktop computers
Low end notebooks to high end workstations,
battery operated laptops etc.
What are the cons to desktop computers
Cost vs performance trade-offs mainly
WHat are servers
Large scale and more reliable computing services
o Key features: Availability, Scalability, Throughput
What is throughput
the amount of work done per unit of time
What are cluster/wareshouse scale computers
SoftwareasaService(SaaS)likesocialnetworking,search,
multiplayer games, video sharing… has led to clusters.
o Largest clusters are termed as warehouse scale computers
What are the cons of cluster/wareshouse scale computers
o Price-performance
o Power
o Availability just like servers
o Scalability through LANs
What are embedded Computers
- Found in everyday machines; microwave ovens, washing machines,
printers, ….
o AlsofoundinNOTsocommonmachines;Aerospaceapplications, Nuclear, power plants,…
What are the cons of embedded computers
o Cost
o Application-Specific Performance
o Energy Efficiency
o Reliability in remote operating conditions
o Temporal Correctness (sometimes)
What did Micheal Flynn in 1966 propose
Flynn’s Taxonomy: a classification for exploitation techniques of application parallelism by the hardware.
What were the 4 classes Micheal created
- Single instruction stream, single data stream (SISD)
- Single instruction stream, multiple data streams (SIMD)
- Multiple instruction streams, single data stream (MISD)
- Multiple instruction streams, multiple data streams (MIMD)
What is Single instruction stream, single data stream (SISD)
- Uniprocessor category
o From programmer’s perspective, it is the standard sequential computer
o Can exploit instruction-level parallelism
o Use ILP techniques such as superscalar and speculative execution
What is Single instruction stream, multiple data streams (SIMD)
Same instruction is executed by multiple processors
o Used data streams are different
o Exploits data-level parallelism by applying the same operations to multiple items of data in parallel
o Each processor has its own data memory (hence the M D of SIMD)
What is Multiple Instruction streams, Single Data stream (MISD)
No commercial multiprocessor of this type has been built to date, but it rounds out this simple
classification
What is Multiple instruction streams, multiple data streams (MIMD)
Each processor fetches its own instructions and operates on its own data
o Targets task-level parallelism
o More flexible than SIMD and more generally applicable
o Inherently more expensive than SIMD
What is RISC architecture
RISC architecture reduces the cycles per instruction while increasing the number of instructions per program.
What is CISC architecture
CISC architecture reduces the number of instructions per program while increasing the number of cycles per instruction.
What does it mean by RISC has a smaller and simple instruction set
- Fewer instructions
- Fixed instruction length
- Simple Operations
What does it mean by RISC single cycle execution
Fast execution
SImple control logic
Why does RISK have fast execution
EachinstructioninaRISCprocessorisdesignedto execute in a single clock cycle, which leads to faster processing.
Why does RISC have simple control logic
Theuniformityofinstructiontimingsimplifies the control unit of the CPU, reducing the complexity and improving speed.
What is meant by load/store architecture
Only load and store instructions can access memory; all other instructions operate on registers
What are the key charateristics of RISC
Small and simple instruction set
in RISC what does its large register set do
reduces the need to access slower main memory
What does RISC efficient pipelining mean
Instructions are optimized for pipelining, allowing high throughput.
What does RISC simple addressing modes mean
Limited and straightforward memory addressing modes.
What does RISC Compiler Optimization mean
Relies on compilers for efficient instruction scheduling and optimization.
What Does RISC no microcode allow for
Executes instructions directly in hardware without needing microcode.
What does RISC high instruction throughput allow for
High-speed instruction execution due to pipelining and load/store architecture.
What does RISC uniform instruction format allow for
Fixed-length instructions make decoding simpler and faster.
What are the classes of computers classified by functionality
- Personal Mobile Devices (PMDs)
- Desktop computers
- Servers
- Cluster/warehouse scale computers
- Embedded computers
What does RISC stand for
Reduced Instruction Set Computing
What does CISC stand for
Complex Instruction Set Computing
What does pipelining refer to in RISC architecture
A technique that allows multiple instructions to be processed simultaneously at different stages
True or False: RISC architectures typically avoid complex, specialized instructions.
True
What role do compliers play in RISC architectures
Relies on compilers for efficient instruction scheduling and optimization
What is the function of Microcode in RISC
Executes instructions directly in hardware without needing microcode
