Problem 5 - PowerPoint PPT Presentation

1 / 2
About This Presentation
Title:

Problem 5

Description:

List scheduling (assume 2 multipliers and 1 adder) (NOTE1 : for ALAP and force-directed scheduling, set ... Clique partitioning (consider the interconnect cost) ... – PowerPoint PPT presentation

Number of Views:33
Avg rating:3.0/5.0
Slides: 3
Provided by: vlsiSsT
Category:
Tags: clique | problem

less

Transcript and Presenter's Notes

Title: Problem 5


1
Problem 5
v0
v1
v2
I
I
I
  • Consider the data-flow graph in Fig.5. Assume
    that multiplication takes 2 clock cycles and
    addition takes 1 clock cycle to execute.
  • Compute the scheduling based on following
    scheduling algorithms
  • ASAP scheduling
  • ALAP scheduling
  • Force-directed scheduling
  • List scheduling (assume 2 multipliers and 1
    adder)
  • (NOTE1 for ALAP and force-directed scheduling,
    set Tmax 10)
  • (NOTE2 for list scheduling, use mobility as the
    priority function)

v6
v3
v4
v5
v7
v8
v13
O
v9
v10
O
v14
v11
v12
O
Fig.5 Data-Flow Graph
2
Problem 5
v0
v1
v2
I
I
I
  • Do register binding on the following binding
    algorithms
  • Left-edge algorithm
  • Clique partitioning (consider the interconnect
    cost)
  • (NOTE1 use the operation scheduling result of
    either force-directed scheduling or list
    scheduling)
  • (NOTE2 schedule the input node v0, v3, v2 at t
    0, and schedule the output nodes v12, v13, v14
    at t Tmax 1)
  • Do functional unit binding on the multiplications
    (consider the interconnect cost)
  • Do port binding on the adder inputs.
  • Draw the diagram of the synthesized datapath
    based on the above results.

v6
v3
v4
v5
v7
v8
v13
O
v9
v10
O
v14
v11
v12
O
Fig.5 Data-Flow Graph
Write a Comment
User Comments (0)
About PowerShow.com