[SOLVED] how to understand the specs of Graphics card?

Sep 7, 2020
37
3
35
I don't exactly understand the specifications of these graphics cards,

in Nvidia it is :
CUDA Cores, ray tracing cores, tensor cores.

in Amd it's :
Compute Units, Steam processors,Ray Accelerators,Infinity Cache.

I kinda understand some of the concepts but not all.
so please explain me guys.

And is it possible to compare a graphics card from AMD with a Nvidia Graphics card?

Thanks a lot for your time guys
 
Solution
Thanks a lot Zer0123, I'm actually trying to understand all the basics of gpu and cpu, like what cuda does and what is the work of tensor cores, what is L1 Cache and stuffs like that
Well that's good! Knowledge like that can come in handy. What you can do is read up articles on those matters or watch some youtube videos that explain such things (I know I know, I'm being the captain obvious here) and then ask about any specific part you don't understand.

I suggest starting from CPUs, there's a plethora of interesting facts and history associated with CPUs.
Even if I explained how the different components of the two different types of cards were similar, ultimately it would be a poor way of comparing the two cards, as the architectures are fundamentally different and it is generally not an accurate reflection of the performance you'll get in a particular game. I recommend you read performance comparisons between different games at different resolutions for the cards you are interested in in order to get an idea of which card performs better for your needs.
 
  • Like
Reactions: Barry Cupas

Zer0123

Reputable
Aug 8, 2017
34
5
4,545
You can read up on each of the specific features/technologies, but the overall bottomline is that both AMD and Nvidia technologies aim to achieve the same thing, albeit sometimes through different means and different names.

For example, Nvidia Cuda core technology is the equivalent to AMD's Stream processor, or Nvidia ray tracing technology kind of does the same thing as AMD's Ray accelerators etc.

So while the individual cards themselves can be compared, it wouldn't be definitive at all. Since each program handles or supports different technologies differently. One might take advantage of the Ray accelerator more than it would with a ray tracing unit etc. So the comparison would be between the overall performance advantage of each technology and their respective cards in a lot of different programs.
 
  • Like
Reactions: Barry Cupas
Sep 7, 2020
37
3
35
Thanks a lot Zer0123, I'm actually trying to understand all the basics of gpu and cpu, like what cuda does and what is the work of tensor cores, what is L1 Cache and stuffs like that
 

Zer0123

Reputable
Aug 8, 2017
34
5
4,545
Thanks a lot Zer0123, I'm actually trying to understand all the basics of gpu and cpu, like what cuda does and what is the work of tensor cores, what is L1 Cache and stuffs like that
Well that's good! Knowledge like that can come in handy. What you can do is read up articles on those matters or watch some youtube videos that explain such things (I know I know, I'm being the captain obvious here) and then ask about any specific part you don't understand.

I suggest starting from CPUs, there's a plethora of interesting facts and history associated with CPUs.
 
  • Like
Reactions: Barry Cupas
Solution