128 vs 256 memory interface

[phillyman36]

Can someone explain the difference in 128 vs 256 mem interface in video cards? i know that the 256 is better can can someone please explain why and how it works?

 

[Jeff7181]

It's in terms of bus width. A 256-bit bus is twice as wide as a 128-bit bus, so it has twice as much bandwidth at the same speed.

 

[Mem]

Good article on 256mb video card memory via 512mb card memory below,read that.

To answer this we first need to understand exactly what happens to a 256MB card when you max out its memory (VRAM).

Simply put, the data that cannot be stored locally in VRAM is transferred to system memory where you will incur large latency penalties across the board.

Technically the only penalty incurred should be fps losses, but a typical symptom you'll probably encounter is stuttering caused by texture swapping from the system RAM to the VRAM, generally known as cache thrashing.

This takes place because textures needing to be displayed on screen obtain absolute priority, so if a texture needing to go on screen is currently being held in system RAM it will need to be moved to the video card ram before it can be displayed.

Link.

 

[phillyman36]

Thanks. Under what conditions does the 128 bit interface start to get bottlenecked? I assume his res video on modern cards dont pass enough data to cause problems. ie the slower the card runs the more likely you will notice problems vs a faster card when you crank the res up

 

[Jeff7181]

Anti-aliasing and anisotropic filtering are pretty bandwidth intense I believe. High resolutions also are, obviously.

 

[kylebisme]

Originally posted by: Mem
Good article on 256mb video card memory via 512mb card memory below,read that.

To answer this we first need to understand exactly what happens to a 256MB card when you max out its memory (VRAM).

Simply put, the data that cannot be stored locally in VRAM is transferred to system memory where you will incur large latency penalties across the board.

Technically the only penalty incurred should be fps losses, but a typical symptom you'll probably encounter is stuttering caused by texture swapping from the system RAM to the VRAM, generally known as cache thrashing.

This takes place because textures needing to be displayed on screen obtain absolute priority, so if a texture needing to go on screen is currently being held in system RAM it will need to be moved to the video card ram before it can be displayed.

Link.

Heh, we wasn't asking about memory denseties though, but rather memory interfaces.


As for bus widths, they are relitive to bus speeds. For instance; a 500mhz 128bit bus is equivalent to a 250mhz 256bit bus. So if you understand what faster ram does, then you understand what a wider bus does as they both have the same effect.

 

[Mem]

Heh, we wasn't asking about memory denseties though, but rather memory interfaces.

Yep I know,but I know somebody will ask that sooner or later .

 

[Jeff7181]

Originally posted by: TheSnowman

Originally posted by: Mem
Good article on 256mb video card memory via 512mb card memory below,read that.

To answer this we first need to understand exactly what happens to a 256MB card when you max out its memory (VRAM).

Simply put, the data that cannot be stored locally in VRAM is transferred to system memory where you will incur large latency penalties across the board.

Technically the only penalty incurred should be fps losses, but a typical symptom you'll probably encounter is stuttering caused by texture swapping from the system RAM to the VRAM, generally known as cache thrashing.

This takes place because textures needing to be displayed on screen obtain absolute priority, so if a texture needing to go on screen is currently being held in system RAM it will need to be moved to the video card ram before it can be displayed.

Link.

Heh, we wasn't asking about memory denseties though, but rather memory interfaces.


As for bus widths, they are relitive to bus speeds. For instance; a 500mhz 128bit bus is equivalent to a 250mhz 256bit bus. So if you understand what faster ram does, then you understand what a wider bus does as they both have the same effect.

Almost... they both provide the same bandwidth, but the 500 MHz bus would typically have lower latency. That's the reason AMD went with Hyper Transport, which is a high speed (2000 MHz) and thus low latency bus. But it's only 16-bits wide (relatively small considering FSB's have been 64-bits wide for ten years, and dual chanel memory interfaces are effectively 128-bits wide). So it provides enough bandwidth, and the latency is very low since clock cycles are ticking by 2,000,000,000 times per second.