written 5.3 years ago by
teamques10
★ 64k
|
• modified 5.0 years ago |
Mumbai University > Computer Engineering > Sem 8 > parallel and distributed systems
|
written 5.0 years ago by
teamques10
★ 64k
|
modified 5.0 years ago
by
ankitpandey
• 100
|
| Distributed System | Parallel System |
|---|---|
| High level of abstraction. The programmer generally do not think about low level details. | Both low and high level on implementation but at a lower level then distributed frame work. |
| Memory usage is comparatively higher due to constraints such as data immutability. | Memory usage is comparatively lower as it allows in place data structure changes. |
| Simple to develop applications due to simple API's. | Comparatively complex to develop applications. |
| Fault tolerance is built-in. | Fault tolerance is to be handled by programmer through additional code CPU usage is comparatively lower due to very over heads and optimize collective fractions. |
| Run time speed is slower. | Run time speed is higher. |
| Examples : Spark, Flink | Example : MPI, HPX, Charm, etc. |
and 2 others joined a min ago.