PARALLELISED APPLICATIONS ON
MULTICORE PROCESSORS
Vishnu.G (CB107CS070), Baskar.R (CB107CS110),
Avinash Kumar.M.K (CB107CS209)
DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING
AMRITA VISHWA VIDYAPEETHAM
ETTIMADAI,COIMBATORE-641105
ggvishnu29@gmail.com
ABSTRACT:
This Project is aimed at analysing and multiple CPU's in one chip. A multi-core
comparing the performance of selected processor is quite different from separate
applications which are parallelised and made to processors work together, with each processor
run on a multicore platform. Tools like Vtune having its own dedicated memory, cache and
performance analyser, Intel C++ compiler, other hardware whereas in a multi-core
OpenMP Concurrency platform are used for processor, the multiple execution cores are
collecting thread data and program execution placed on the same chip and resources such as
time. Using the data collected from the tools, a cache and memory are common to all the
graph is drawn showing the speed up attained execution cores. When we run two applications
when the applications are parallelised simultaneously in a CPU, what actually happens
is that the CPU time is shared between the two
Introduction: applications which are run simultaneously, this
CPU time slice for each application would be in
The Digital age paves way for an the order of microseconds. In a multi-core
undeterred growth in all fields of science, system there are say n number of cores, so it is
especially in the field of microprocessors, as the feasible to run n number of applications
number of transistors on a microprocessor chip simultaneously. This Paper is aimed at
has grown exponentially. All of this was in envisaging the need for parallelizing present day
accordance with Moore's law which proposed software for better adaptation to multi-core
that the number of transistors in Intel's chips architectures.
would be doubled every Eighteen months, and
so, as such the number of transistors on each Our project mainly deals with analyzing
chip reached millions and there was no stopping the performance of applications which are
this colossal growth. The clock frequency of parallelized using a concurrency platform. We
operation rose from a few megahertz to a few used various tools provided by Intel for testing
gigahertz . Later, it became evident that the heat and analysing the application. They are as
generated was so intense that any more over follows:
clocking of the micro processor would result in
the chip melting to the intense heat generated 1. Intel C++ compiler
during its operation. The frequency of operation 2. OpenMP concurrency platform
could not be increased further without excessive 3. Vtune performance analyzer
heat generation. This paved way for the multi-
core processors to come to the fore. A multi-
core processor consists of multiple cores or
Thread 0:
Execution time: 2.4465558 s