Stanford Smart Memories Project has Tensilica cores.

[Brimstone]

Neat project. They are trying to fuse together Multi-threading and Data-streaming in a single architechture. They mapped the Stanford Imagine Stream Processor and Hydra to it with decent results. The link below is from when they used a Mips64 core, and I haven't seen any papers with benchmarks with Tensilca being used. The switch to Tensilica is relativley recent in terms of the projects life span.


http://mos.stanford.edu/papers/km_isca_00.pdf

The Stanford Smart Memories Project is a research effort to design a single-chip computing element which provides configurable hardware support for diverse computing models and maps efficiently to future wire-limited VLSI technologies. The project involves researchers in VLSI circuits, computer architecture, compilers, operating systems, computer graphics, and computer networking.

http://www-vlsi.stanford.edu/smart_memories/

The slides they have are worth a look. Here are a few.



http://www-vlsi.stanford.edu/smart_memories/stuff/posters/SM HW Poster 03-09.pdf

http://www-vlsi.stanford.edu/smart_memories/stuff/posters/SM SW Poster 03-09.pdf

http://www-vlsi.stanford.edu/smart_memories/stuff/posters/SM HW Poster 04-03.pdf

 

[Entropy]

36x30 inch pdfs.
Hard to make an intelligent comment without plowing in some decent amount of time. They point to the same trends other people do, and some of the conclusions seem universal. These guys are at silicon, which is nice. Silicon is a very long way from "market" when it comes to academic research project though. They did mention 3D-gfx as one of the application areas they analyzed beforehand.

 

[flick556]

If I understood everything correctly it should be good for consoles where memory is shared between gpu and cpu.

 

[Nick]

36x30 inch pdfs.

Quite normal for academic posters.

 

[Entropy]

36x30 inch pdfs.

Quite normal for academic posters.

I know - it's just such a tell tale sign of academia.
And, of course, the relative abundance of 44 inch wide printers - now when I started out in post graduate research, we didn't even have laser printers, no, we had a daisy writer connected to a PDP-11, and.....