Scribd
Upload
4 views2 pages

GPUs

GPUs, or graphics processing units, are specialized co-processors designed for parallel processing, allowing them to handle multiple data streams simultaneously, making them ideal for tasks like rendering graphics and machine learning. While CPUs excel at complex operations on small data sets, GPUs are optimized for performing numerous calculations concurrently, particularly in matrix and vector operations. The architecture of GPUs, with thousands of smaller cores, enables them to efficiently manage large amounts of data in real-time.

Uploaded by

shaambavy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
4 views2 pages

GPUs

GPUs, or graphics processing units, are specialized co-processors designed for parallel processing, allowing them to handle multiple data streams simultaneously, making them ideal for tasks like rendering graphics and machine learning. While CPUs excel at complex operations on small data sets, GPUs are optimized for performing numerous calculations concurrently, particularly in matrix and vector operations. The architecture of GPUs, with thousands of smaller cores, enables them to efficiently manage large amounts of data in real-time.

Uploaded by

shaambavy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

GPUs

CPU considered the brain of the computer carrying out and controlling all
instructions

As technology advanced lots of demand was placed on this CPU to carry


out high performance tasks

Co processorany additional processor used for a specialised task

MOST WELL KNOWN type is the GPU initially used for rendering
graphics ,the large number of cores in CPUs have been exploited to
process many parallel streams of data at the same time no matter what
the data might be-no longer limited to graphics processing

TYPICALLY slower than CPU core, CPU are more general function GPU are
highly specialised so ae quicker for CERTAIN given tasks not all

CPUs excel at performing complex operations on small data sets CPUs can
handle simple problems with large data

GPUs CAN do complex calculations and do however excel at performing


matrices/vector calculations and multiple data at the same time

WHY are GPUS good at rendering graphics lots of calculations(coordinates


on images calculations, floating point arithmetic, generating
representation etc)

Huge number of tasks like that can be done parallelly as they do not
require each other eg each RGB calculation is independent per pixel

GPUs are designed for efficient parallel processing so despite each task
being carried out slower vs a CPU still more tasks in given time period

Graphics cards contain their own sets of RAM as well as a processor. The processor
on a graphics card is referred to as a GPU (graphics processing unit).
It is not unusual for a GPU to contain 1,000+ cores. Operations such as 3D graphics
rendering are divided across multiple cores, as each core performs a small part of
the whole operation.
optimised to parallel process calculations, which are required for machine learning
algorithms
A GPU works by parallelizing tasks to perform many operations
simultaneously. Unlike a CPU (Central Processing Unit), which focuses on
executing sequential instructions quickly, a GPU consists of thousands of
smaller cores that work together to process multiple pieces of data
concurrently. This parallel processing capability allows GPUs to handle
large amounts of data in real-time, making them particularly suited for
tasks that require massive computations, such as rendering graphics,
processing videos, and performing complex calculations.

You might also like