Geforce GTX 260 Core 216

The 4870 wins 3 of 5 1920x1200 benches.

The 4870's 3-2 lead over the 260 216, at 1920x1200, does not continue onto 2560x1600 benches. That's all I'm saying.

True, and we do have to take into account the different memory algorithms that Nvidia and AMD employ.

 

so theyve added more shaders is that the general opinion the 260 was lacking shader power compared to the 4870?

 

Essentially. GT200 in all forms has superior pixel fillrates and texturing capabilities compared to RV770, and bandwidth is basically a wash so that really just leaves triangle setup rates and shader compute rates.

 

Thanks Chris!

 

I hate Hybrid SLI too.


I do like the prices of the original GTX260 more, now though. Kinda hard to see how the 9800GTX(+) will survive.

 

Though I got it late. I am working on an investigation to how it will SLI with the normal 260 right now.

 

Granted, average framerate figures can easly hide the fact that the framerate will plumit whenever your card with runs out of VRAM and thus haves to swap data with main RAM before it can draw the next frame.

 

So after messing with the 216 GTX 260 for a while and SLIIng it with a standard 260. I cant say that inter-frame delay is any different than SLI. It varies from game to game like any other AFR configuration. One card does seem to have slightly faster frame time rendering but nothing that was immediately noticable compared to a standard 260 configuration. In "Raw" FPS its actually a little faster to mix them than to go with a normal 260 SLI configuration. But nothing that would make or break it. I did a few performance benches if anyone is interested,


http://forums.slizone.com/index.php?showtopic=28579

 

a have a quick question about hybrid sli thats a igp helping out a standalone card - yes
can you have it with 2 standalone cards eg: gt260 + 8800gts ?

 

Geforce Boost only works with the low end spectrum ((Geforce 8400 type cards)). So no. You wouldnt want to do that anyway. You'd get huge micro stuttering due to very large inconsistencies in frame distribution.

Chris