CUDA Core Vs CPU Cores

[djoshi]

Hi,

 

What is the difference between CUDA Core and CPU Cores.

 

As a end user my query is:

 

If I am rendering on i7 CPU I get 8 cores(Rendering Squares) are rendering my image.

 

Now if I render using 128 core GPU that means would it render the image with 128 cores(Rendering Squares)??

[Dimitris Tolios]

Hi,

 

What is the difference between CUDA Core and CPU Cores.

 

As a end user my query is:

 

If I am rendering on i7 CPU I get 8 cores(Rendering Squares) are rendering my image.

 

Now if I render using 128 core GPU that means would it render the image with 128 cores(Rendering Squares)??

 

Not exactly.

Yes CPU threads are represented as "buckets" of 64x64 pixels by default - your little squares in the rendering window.

 

GPU rendering with "unbiased rendering engines" like VRay RT / Octane / Iray etc use brute force calculation and usually do not look like a blurry image resolved by buckets. You can "youtube" videos that show how the rendering window looks like, but the closest equivalent would be the light cache pre-pass (with show calc phase option enabled), where little pixel dots resolve more and more detail out of a dark frame, as more and more light ray bounces are calculated on the fly by the GPU.

 

So no, you will see no buckets (Squares) while rendering with your GPU.

 

CUDA cores are not the same as CPU cores. AMD and Nvidia has broken down their graphic calculation process using hundreds of small shading calculation hardware units. Each one of those gets a small portion of the image and dedicates to it, working in parallel with the rest of the "shaders" or "cores" for a complete final result.

 

Just like with CPU cores, all CUDA or GPU cores are also not equal, thus unless we are talking the same generation or architecture, actual output is not scalable with the number of cores: newer Kepler GPUs like the GTX 680 have 3x the CUDA cores the card it replaces had (512 in the Fermi based GTX 580) but it is nowhere near 3x as fast - it is actually slower in many tasks, especially for GPU rendering, and with 3x the cores it barely breaks even in performance with its predecessor - that's with the current generation of software. The Fermi cores were "bigger" and more complex, the Kepler ones are smaller, simpler yet faster. This suits some tasks and boosts performance , while penalizes other tasks.

 

So beware when comparing, keep it apples to apples, as oranges are present even within the same company's lineup.

[djoshi]

Thanks..Dimitrios..Very clear explanation.