Sie sind auf Seite 1von 1

CPUs are general purpose integrated circuit / IC that are designed to perform a

variety of operations. For eg. Watching


movie, listening music , making spread-sheets , creating word documents, playing
games etc.
On the other hand, GPUs are specific purpose circuits that are designed to perfo
rm specific tasks like rendering videos,
playing high-end graphics games (which needs parallel processing).
While a CPU consists of a few cores optimized for sequential serial processing
, a GPU has a massively parallel
architecture consisting of thousands of smaller, more efficient cores designed f
or handling multiple tasks simultaneously.
A CPU commonly has 4 to 8 fast, flexible cores clocked at 2-3Ghz, whereas a GPU
have thousands of relatively simple
cores clocked at about 1Ghz.
i.e, CPUs have a few but powerful cores whereas GPUs have less powerful but hund
reds of cores
This makes GPUs very good at extremely parallel tasks, but less good at processi
ng only a single data stream.
Therefore, tasks that can be efficiently divided into many parallel parts will s
ee enormous benefits when running on
a GPU.
This highly parallel architecture is the reason that a GPU can process such larg
e batches of data so quickly.
But if a single serial operation is given to both CPU and GPU, then CPU will pro
cess the data earlier than GPU because
CPU operates at frequency higher than that of GPU.
So , Serial portions of the code run on the CPU while parallel portions run on t
he GPU.
So, in our pc, for high performance , we need both GPU as well as CPU.

Das könnte Ihnen auch gefallen