What is thread and working in Cpus

Solution
You've only hit the tip of the iceberg. There's huge differences in architecture. Cores aren't the same so you go deeper down and threads are going through the same process. And parallelism is a completely other topic in talking about thread handling. Gpus and cpus can't process the same languages. All gpus, intel, nvidia, and amd can be used for certain workloads, it's not limited to cuda. There are many apis for such work. Ghz are not created equal, even for cpus, 2ghz can get more work done than 4 ghz. You cannot go by speed to know which is faster. But this doesn't explain threads, just that a gpu can handle more at once. The post just isn't asking enough to know what exact topic to go into.

zjuventus14

Honorable
Sep 14, 2013
222
0
10,760
The only real difference between these two is that a GPU has over 100 cores, if it is NVIDIA they are CUDA Cores, which can be used by applications under certain workloads where more cores are better. The drawback is that those cores are at the base clock, which is usually between 1 and 1.4 Ghz, while CPU's have between 2(low-end) and 18($3,000 Xeons) cores, which are faster, usually between 2 and 4 Ghz. If that is what you meant, that is the only real difference. GPUs are made this way because they need to do many calculations to orient and shade things, for example, at 1920 by 1080p, 2073600 pixels are bing rendered, and although one pixel may be quick to render, with so many, hundreds of cores are needed to do this in a reasonable amount of time, especially since after being rendered, shaders must be applied. CPUs, on the other hand, are optimized to perform large calculations, but without as many variables, which benefits from fewer, fast cores.
 
You've only hit the tip of the iceberg. There's huge differences in architecture. Cores aren't the same so you go deeper down and threads are going through the same process. And parallelism is a completely other topic in talking about thread handling. Gpus and cpus can't process the same languages. All gpus, intel, nvidia, and amd can be used for certain workloads, it's not limited to cuda. There are many apis for such work. Ghz are not created equal, even for cpus, 2ghz can get more work done than 4 ghz. You cannot go by speed to know which is faster. But this doesn't explain threads, just that a gpu can handle more at once. The post just isn't asking enough to know what exact topic to go into.
 
Solution

TRENDING THREADS