LogP machine

The LogP machine is a model for parallel computation. [Culler et al. 1993] It aims at being more practical than the PRAM model while still allowing for easy analysis of computation.The name is not related to the mathematical logarithmic function: Instead, the machine is described by the four parameters $L$, $o$, $g$ and $P$.

The LogP machine consists of arbitrarily many processing units with distributed memory.The processing units are connected through an abstract communication medium which allows point-to-point communication.

The machine is described by the four parameters:

* $L$, the latency of the communication medium.
* $o$, the overhead of sending and receiving a message.
* $g$, the gap required between two send/receive operations.
* $P$, the number of processing units. Each local operation on each machine takes the same time (``unit time"). This time is called a processor cycle.

The units of the parameters $L$, $o$ and $g$ are measured in multiples of processor cycles.

See Also

* Bulk Synchronous Parallel

Notes

References

Citation
last = Culler
first = David
last2 = Karp
first2 = Richard
last = Patterson
first = David
last = Sahay
first = Abhijit
last = Schauser
first = Klaus Erik
last = Subramonian
first = Eunice
last = Von Eicken
first = Thorsten
title = LogP: Towards a realistic model of parallel computation
journal = ACM SIGPLAN Notices
volume = 28
issue = 7
pages = 1-12
date = July
year = 1993
url = http://www-static.cc.gatech.edu/~bader/COURSES/UNM/ece638-Fall2004/papers/CKP93.pdf
doi = 10.1145/173284.155333