Download presentation
Presentation is loading. Please wait.
Published byOswald Rice Modified over 9 years ago
1
A SEMINAR ON 1 www.engineersportal.in
2
CONTENT 2 The Stream Programming Model The Stream Programming Model-II Advantage of Stream Processor Imagine’s Performance Implementation on the Stream Processor www.engineersportal.in
3
The Stream Programming Model Programmable Kernel Stream 4 data Stream 3 data Stream 2 data Stream 1 data The Main Idea 3 www.engineersportal.in
4
The Stream Programming Model Programmable Kernel Stream 4 data Stream 3 data Stream 2 data Stream 1 transformed data The Main Idea 4 www.engineersportal.in
5
The Stream Programming Model Programmable Kernel Stream 4 data Stream 3 data Stream 2 data Stream 1 transformed data The Main Idea 5 www.engineersportal.in
6
The Stream Programming Model Programmable Kernel Stream 4 data Stream 3 data Stream 2 data Stream 1 transformed data The Main Idea 6 www.engineersportal.in
7
The Stream Programming Model Programmable Kernel Stream 4 data Stream 3 data Stream 2 data Stream 1 transformed data The Main Idea 7 www.engineersportal.in
8
The Stream Programming Model Transform Chaining Kernels Example: The Geometry Stage of the OpenGL Pipeline Input Vertexes ShadeAssemble CullProject Toward Rasterization Stage 8 www.engineersportal.in
9
The Stream Programming Model Hardware Implementation: the Imagine Stream Processor Communicate with host and issue operations. 9 www.engineersportal.in
10
The Stream Programming Model Hardware Implementation: the Imagine Stream Processor Transfer data between parts of the chip. 10 www.engineersportal.in
11
The Stream Programming Model Hardware Implementation: the Imagine Stream Processor Local storage and reuse of intermediate streams. 11 www.engineersportal.in
12
The Stream Programming Model Hardware Implementation: the Imagine Stream Processor Store kernel code. 12 www.engineersportal.in
13
The Stream Programming Model Hardware Implementation: the Imagine Stream Processor Execute one kernel at a time. 13 www.engineersportal.in
14
The Stream Programming Model Hardware Implementation: the Imagine Stream Processor Connection with other Imagine chips. 14 www.engineersportal.in
15
Advantages of a Stream Processor Programmability Efficient Shading Example: OpenGL Inefficiency 15 www.engineersportal.in
16
Advantages of a Stream Processor Programmability Efficient Shading Example: OpenGL Inefficiency 1. Draw the plane. 16 www.engineersportal.in
17
Advantages of a Stream Processor Programmability Efficient Shading Example: OpenGL Inefficiency 1. Draw the plane. 2. Draw the cube. 17 www.engineersportal.in
18
Advantages of a Stream Processor Programmability Efficient Shading Example: OpenGL Inefficiency 1. Draw the plane. 2. Draw the cube. 3. Redraw the cube. 18 www.engineersportal.in
19
Advantages of a Stream Processor Programmability Efficient Shading Example: OpenGL Inefficiency 1. Draw the plane. 2. Draw the cube. 3. Redraw the cube. Redraw the complete scene to obtain correct shadow on one object. 19 www.engineersportal.in
20
Advantages of a Stream Processor Programmability Efficient Shading Hardware Implementation of New API API Example: Pixar’s Renderman (Reyes Image Rendering Architecture) 20 www.engineersportal.in
21
Advantages of a Stream Processor Producer - Consumer Locality Capture Example: OpenGL Pipeline Inefficiency Geometry Stage Rasterization Stage Composite Stage Vertexes 21 www.engineersportal.in
22
Advantages of a Stream Processor Producer - Consumer Locality Capture Example: OpenGL Pipeline Inefficiency Geometry Stage Rasterization Stage Composite Stage Vertexes Assembled Triangles Fragments Pixels 22 www.engineersportal.in
23
Advantages of a Stream Processor Producer - Consumer Locality Capture Example: OpenGL Pipeline Inefficiency Geometry Stall Rasterization Stage Composite Stage Vertexes Assembled Triangles Fragments Pixels 23 www.engineersportal.in
24
Advantages of a Stream Processor Producer - Consumer Locality Capture Example: OpenGL Stream Inplementation Vertex Streams Fragment Streams Pixel Streams Rasterization Kernels Composite Kernels Geometry Kernels Triangle Streams 24 www.engineersportal.in
25
Advantages of a Stream Processor Producer - Consumer Locality Capture Example: OpenGL Stream Inplementation Vertex Streams Fragment Streams Pixel Streams Rasterization Kernels Composite Kernels Geometry Kernels Triangle Streams 25 www.engineersportal.in
26
Advantages of a Stream Processor Flexible Resource Allocation Example: OpenGL Pipeline Inefficiency Geometry Stage Rasterization Stall Composite Stall Vertexes Waste of hardware capacity. 26 www.engineersportal.in
27
Advantages of a Stream Processor Flexible Resource Allocation Example: OpenGL Stream Implementation Vertex Streams Rasterization Kernels Composite Kernels Geometry Kernels No waste: kernels are pieces of code running on the same hardware! 27 www.engineersportal.in
28
Advantages of a Stream Processor Pipeline Reordering Example: Blending off in the OpenGL Pipeline Part of Rasterization - Composite Stage Texture Kernel Blending Kernel Depth Kernel Fragments 28 www.engineersportal.in
29
Advantages of a Stream Processor Pipeline Reordering Example: Blending off in the OpenGL Pipeline Part of Rasterization - Composite Stage Texture Kernel Blending Kernel Depth Kernel Fragments Many fragments are needlessly textured 29 www.engineersportal.in
30
Advantages of a Stream Processor Pipeline Reordering Example: Blending off in the OpenGL Pipeline Part of the Rasterization/Composite Stage Texture Kernel Depth Kernel Fragments We can reorder the pipeline. 30 www.engineersportal.in
31
Advantages of a Stream Processor Obvious Scalability Data Level Parallelism Texture Kernel Texture Kernel Texture Kernel Fragments 31 www.engineersportal.in
32
Advantages of a Stream Processor Obvious Scalability Functional Parallelism Texture Kernel Blending Kernel Depth Kernel 32 www.engineersportal.in
33
Imagine’s Performance That looks great! 33 www.engineersportal.in
34
Imagine’s Performance “Interaction between host processor and graphics subsystem not modeled” in Imagine. “Many hardware-accelerated systems are limited by the bus between the processor and the graphics subsystem”. 34 www.engineersportal.in
35
Imagine’s Performance “Imagine clocks rate is also significantly higher (500MHz vs. 120 MHz)”. 35 www.engineersportal.in
36
Imagine’s Performance 36 www.engineersportal.in
37
Imagine’s Performance But the comparison is still “instructive”. “Running our tests on commercial systems gives a sens of relative complexity”. Frame Rate Normalized to the Sphere Test NVIDIA Quadro and Imagine Relative Performance 37 www.engineersportal.in
38
Implementation on the Stream Processor Frame Speed Frame Complexity/ Quality OpenGLReyes 38 www.engineersportal.in
39
Implementation on the Stream Processor Frame Speed Frame Complexity/ Quality Enhanced OpenGL Implementation Degraded Reyes Implementation 39 www.engineersportal.in
40
Implementation on the Stream Processor OpenGL Implementation Reyes Implementation Isim Simulator Models complete Imagine architecture. Idebug Simulator Do not model kernel stalls Do not model cluster occupancy effects Increased size of dynamically addressable memory How to compare the results? 40 www.engineersportal.in
41
Results 41 www.engineersportal.in
42
Conclusion “When comparing graphics algorithms, [the lack of specialization] does make Imagine performance- neutral to the algorithms employed”. “Our Reyes implementation made slight changes to the simulated Imagine hardware [...] Having a larger [size of addressable memory] was vital for kernel efficiency”. “Imagine is an appropriate platform for comparing different rendering algorithms toward an eventual goal of high-performance hardware implementation.” 42 www.engineersportal.in
43
43 www.engineersportal.in
44
? 44 www.engineersportal.in
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.