· Daily Matrices
· DAC Pavilion Panels
· Management Day@DAC
· Wireless Wednesday
· Search the Program

· Keynotes
· Papers
· Panels
· Special Sessions
· Monday Tutorial
· Friday Tutorials

· Intro to EDA
· Integrated Design Systems Workshop
· UML for SoC Design
· Women's Workshop

· RTL Handoff
· Core-based SoC Design






















WEDNESDAY, June 15, 2005, 02:00 PM - 05:00 PM | Room: 211AB
TRACK:SYSTEM-LEVEL DESIGN AND VERIFICATION

  HoT Core-based SoC Design
  Using Configurable Processors to Replace RTL Blocks (Tensilica, Inc. and Virage Logic Corp.)

    Configurable processors can be tailored to an exact application and are often used to replace blocks of RTL. Because configurable processors employ firmware instead of RTL-defined state machines for their control algorithms, it's easier and faster to develop and verify processor-based task engines for many embedded SoC tasks than to develop and verify RTL-based hardware to perform the same tasks. Tensilica lets designers add specialized functions right into the processor's execution units without requiring that the designers understand the processor architecture. Configurable processors are delivered as synthesizable RTL code, ready for placement into an FPGA or SoC design.

This 3-hour tutorial will show how to design a configurable processor with the equivalent performance to hand-coded RTL using Tensilica's XPRES compiler, which automatically analyzes C code to determine the best processor configuration and extensions. This tutorial will also show how to evaluate and select the optimal embedded memory IP for their Xtensa processor configurations using a web-based portal interface from Virage Logic. This tutorial will show the performance impact of various different memory configurations, and how important it is to tightly match the memory to the processor configuration.

The tutorial will start with a very short description of the overall design flow, then get right into hands-on use of the XPRES compiler. Software will be loaded with a complex C-based algorithm, which the student can review and run through the XPRES compiler to identify critical inner loops and other tuning opportunities. The XPRES compiler will allow the student to create several trial processor configurations that boost performance, and the student will be able to see the area and performance trade-offs of the various configuration options.