Small-Scale Reconfigurability for Improved Performance and Double-Precision

[Jawed]

http://www.cs.virginia.edu/~ktd3q/docs/ssr_gpu_ije.pdf

We explore the application of Small-Scale Reconfigurability (SSR) to graphics hardware. SSR is an architectural technique wherein functionality common to multiple subunits is reused rather than replicated, yielding high-performance reconfigurable hardware with reduced area requirements (Vijay Kumar and Lach 2003). We show that SSR can be used effectively in programmable graphics architectures to allow double-precision computation without affecting the performance of single-precision calculations and to increase fragment shader performance with a minimal impact on chip area.

Chocks away!

Interestingly, something called Qsilver is used

Qsilver is a simulation framework for graphics architectures that can simulate low-level GPU activity for any existing OpenGL application (Sheaffer et al. 2004). Qsilver uses Chromium (Humphreys et al. 2002) to intercept and transform an OpenGL application’s API calls and create an annotated trace that encapsulates geometry, timing, and state information. This trace serves as input to the Qsilver simulator core, which performs an accurate timing simulation of the graphics hardware and produces detailed statistics.

Jawed

 

[Anteru]

Nice finding, thanks ... and nVidia is involved, too ...

 

[3dcgi]

I haven't read the paper yet so I can't comment on it, but I believe Qsilver was presented as a poster at Siggraph a couple years back. I recall thinking it looked interesting, but I haven't followed it since then.