1.3 Computer-System Architecture Flashcards
A computer system can be organized according to the number of general-purpose _______.
processors
Most computer systems used to have a single processor containing one CPU with a single _______.
processing core
The core is the component that executes _______.
instructions
The main CPU is capable of executing a general-purpose _______.
instruction set
Special-purpose processors run a limited _______.
instruction set
Examples of special-purpose processors include disk, keyboard, and _______.
graphics controllers
The operating system sends information about tasks to special-purpose processors and monitors their _______.
status
A disk-controller microprocessor implements its own disk queue and _______.
scheduling algorithm
The use of special-purpose microprocessors does not turn a single-processor system into a _______.
multiprocessor
If there is only one general-purpose CPU with a single processing core, the system is a _______.
single-processor system
Very few contemporary computer systems are defined as _______ systems.
single-processor
Multiprocessor systems have ___ or more processors.
two
The primary advantage of multiprocessor systems is increased _______.
throughput
The speed-up ratio with N processors is _______ than N.
less
In symmetric multiprocessing (SMP), each peer CPU processor performs all tasks, including _______ functions.
operating-system
In SMP architecture, each CPU processor has its own set of _______.
registers
All processors in a multiprocessor system share _______ over the system bus.
physical memory
A multiprocessor system allows processes and resources to be _______ dynamically among the various processors.
shared
The definition of multiprocessor now includes _______ systems.
multicore
On-chip communication is _______ than between-chip communication.
faster
A dual-core design features two cores on the same _______.
processor chip
The local cache specific to each core is often known as a level 1 or _______ cache.
L1
The level 2 cache is local to the chip but is _______ by the two processing cores.
shared
A multicore processor with N cores appears to the operating system as _______ standard CPUs.
N
Adding additional CPUs increases computing power, but too many will lead to contention for the _______.
system bus
The approach where each CPU has its own local memory is known as _______.
non-uniform memory access (NUMA)
NUMA systems can scale more effectively as more _______ are added.
processors
A potential drawback of NUMA systems is increased _______ when accessing remote memory.
latency
Blade servers consist of multiple independent multiprocessor _______.
systems
Each blade processor board boots _______ and runs its own operating system.
independently
A clustered system gathers together multiple _____.
CPUs
Clustered systems are composed of two or more individual systems, also known as _____.
nodes
Clustered systems are considered _____.
loosely coupled
The generally accepted definition of clustered computers is that they share _____ and are closely linked via a local-area network.
storage
Clustering is usually used to provide high-_____ service.
availability
High availability is generally obtained by adding a level of _____.
redundancy
A layer of cluster software runs on the cluster _____.
nodes
The ability to continue providing service proportional to the level of surviving hardware is called _____ degradation.
graceful
Systems that can suffer a failure of any single component and still continue operation are called _____.
fault tolerant
Fault tolerance requires a mechanism to allow the failure to be detected, diagnosed, and, if possible, _____.
corrected
In asymmetric clustering, one machine is in _____ mode while the other is running applications.
hot-standby
In symmetric clustering, two or more hosts are running applications and are _____.
monitoring each other
Clusters can provide high-_____ computing environments.
performance
The technique known as _____ divides a program into separate components that run in parallel.
parallelization
Parallel clusters allow multiple hosts to access the same data on shared _____.
storage
The function that ensures no conflicting operations occur in a parallel cluster is commonly known as a _____ manager.
distributed lock
Cluster technology is changing rapidly, with some products supporting thousands of systems in a cluster, including those separated by _____.
miles
Storage-area networks (SANs) allow many systems to attach to a pool of _____.
storage
In a database cluster, dozens of hosts can share the same database, greatly increasing _____ and reliability.
performance
