GPU Code-Generation Extensions to the SCORE Compiler

Add to your list(s) Download to your calendar using vCal

• Cody Huang
• Wednesday 01 June 2011, 13:00-14:00
• Level 9 Mahanakorn Lab., EE Dept..

If you have a question about this talk, please contact .

Streaming languages and frameworks offer a natural way to express parallelism for many computing substrates. We show how to extend the SCORE framework to support NVIDIA GP Us for simple streaming functions. The SCORE framework now supports a range of targets including spatial FPGA Verilog, VLIW FPGA -overlays, Microblaze C along with the newly developed CUDA backend. Our generated code is comparable in performance to hand-written code for simple computational kernels. Moving forward, we expect this backend to be a platform for followon research: shape analysis, auto-tuning support, automated kernel-decomposition, nested data-parallel mapping among others.

This talk is part of the CAS Talks series.

This talk is included in these lists:

• Andrea Picciau's list
• CAS Talks
• Circuits and Systems Group: Internal Seminars
• Level 9 Mahanakorn Lab., EE Dept.

Note that ex-directory lists are not shown.