Parallel Rendering

1

2

Introduction

• In many situations, standard rendering pipeline not sufficient

Need higher resolution display

More primitives than one pipeline can handle

•Want to use commodity components to build system that can render in parallel

•Use standard network to connect

3

Power Walls

• Where display large data sets? Need resolution comparable to data set to see detail

• Medical: CT / MRI

• Ocean / atmospheric

• Solutions? Multi LCD / Plasma panels

Multiple projectors• Commodity

• High-end

4

Tiled Display

5

CS Power Wall

•6 dual processor Intellestations•G Force 3 Graphics cards•6 commodity projectors (1024 x 768)•Gigabit ethernet•Back projected screen•Shared facility with scalable system group

Investigate OS and network issues

6

CS Power Wall

7

CS Power Wall

8

Power Wall

• Inexpensive but some problems Color matching

Vignetting

Alignment • Overlap areas

Synching

Dark field

9

Graphics Architectures

•Pipeline Architecture SGI Geometry Engine

Geometry passes through pipeline

Hardware for• clipping • transformations• texture mapping

Project/SortClipTransform Rasterize Screen

10

Building Blocks

•Graphics processors consist of geometric blocks and rasterizers

•Geometric units: transformations, clipping, lighting

•Rasterization: scan conversion, shading

•Parallelize by using mutiple blocks

•Where to do depth check?

R

G G G

R R

11

Sorting Paradigm

•Can categorize different ways of interconnecting blocks using sorting paradigm:

•each projector responsible for one area of screen

must sort primitives and assign to proper projector

•Algorithms categorized by where sorting occurs

12

Three Rendering Methods

Sort-First Rendering Sort-Middle Rendering Sort-Last Rendering

R

G G G

R R

Sort G G G

R R R

Sort R

G G G

R R

Composite

13

Sort First

•Each R assigned to area of screen•Each G coupled to own R•Must sort primitives first•Can use commodity cards

R

G G G

R R

Sort

14

Sort-First Rendering: Random Triangles Application

15

Sort Middle

• Gs and Rs decoupled

• Each G can be assigned any group of objects

• Each R assigned to area of screen

• Must sort between stages

G G G

R R R

Sort

16

Sort Last

•Couple Rs and Gs•Assign objects to Gs to load balance or via application

•Composite results at end

R

G G G

R R

Composite

17

Tree Compositing

•Composite in pairs•Send color and depth buffers•Each time half processors become idle

18

Binary Swap Compositing

•Each processor responsible for one part of display

•Pass data to right n times

19

Sort-Last Rendering: Random Triangles Application

20

Comparison

•Sort first Appealing but hard to implement

•Sort middle Used in hardware pipelines

More difficult to implement with add-on commodity cards

•Sort last Easy to implement with compositing stage

High network traffic

21

Mapping to Clusters

•Different architectures Shared vs distributed memory

Communication overhead

Parallel vs distributed algorithms

•Easy to do sort last•Must evaluate communication cost•Standard visualization strategies incorrect if transparency used

22

Vista Azul

•Experimental architecture from IBM donated to AHPCC

•Half Intel nodes, half AIX nodes•Only one (PCI) graphics card per four processors

•Contained Scalable Graphics Engine (SGE):

•high-speed high-resolution color buffer accessible by all processors

23

Vista Azul

24

Comparison Between Sort-First and Sort-Last

Sort Last Rendering vs. Sort First Rendering

0

5

10

15

20

25

30

35

0 2 4 6 8 10 12

Number of Processors

CP

U T

ime

(sec

on

ds) Sort Last Rendering

Sort First Rendering

25

Performance on PC Cluster

•Following experiments done by Ye Cong on CS cluster

6 Intellestations

Gigabit Ethernet

GForce 3 graphics

•Show effect of network

26

Sort-First vs Sort LastRandom Triangles

27

Sort First vs Sort LastTeapot

28

Azul vs Intellistations

29

Software for Parallel Rendering

•Write your own sort-first sort-last•WireGL/Chromium (Stanford)•Embed inside package (VTK)

30

WireGL: A Distributed Graphics System

• SW-based parallel rendering system unifies rendering power of collection of cluster

nodes

• Scalability achieved by integrating parallel applications into sort-first parallel Rs

• Each node in cluster: either rendering client or rendering server

• Clients submit OpenGL commands concurrently to servers

• Servers render final physical image

31

Chromium

•Successor to WireGL•Allows both sort first and sort last rendering

• Implemented on CS cluster•Most of gain in performance because Chromium and WireGL can group state-changing commands separately from rendering commands

32

Chromium vs Sort First

MRI rotation