This is the second book in a three-part series that traces the development of the GPU, which is defined as a single chip with an integrated transform and lighting (T&L) capability. This feature previously was found in workstations as a stand-alone chip that only performed geometry functions. Enabled by Moore’s law, the first era of GPUs began in the late 1990s.
Silicon Graphics (SGI) introduced T&L first in 1996 with the Nintendo 64 chipset with integrated T&L but didn’t follow through. ArtX developed a chipset with integrated T&L but didn’t bring it to market until November 1999.
The need to integrate the transform and lighting functions in the graphics controller was well understood and strongly desired by dozens of companies. Nvidia was the first to produce a PC consumer level single chip with T&L in October 1999. All in all, fifteen companies came close, they had designs and experience, but one thing or another got in their way to prevent them succeeding.
All the forces and technology were converging; the GPU was ready to emerge. Several of the companies involved did produce an integrated GPU, but not until early 2000.
This is the account of those companies, the GPU and the environment needed to support it. The GPU has become ubiquitous and can be found in every platform that involves a computer and a user interface.
Author(s): Jon Peddie
Publisher: Springer
Year: 2023
Language: English
Pages: 354
City: Cham
Foreword
Preface
The Author
A Lifetime of Chasing Pixels
Acknowledgments and Contributors
Contents
List of Figures
List of Tables
1 Introduction
1.1 Nvidia’s NV10—First Integrated PC GPU (September 1999)
1.1.1 Nvidia’s Non GPU GPU
1.1.2 The S3 Nvidia Patent Suit and Silicon Valley Gossip
1.2 Summary and Conclusion
1.3 Epilog—All the Others
References
2 The GPUs’ Functions
2.1 The Pipeline
2.1.1 The Meaning of Real Time
2.1.2 Binning and Branding
2.1.3 The Frame Buffer
2.1.4 Object-Level Clipping
2.2 The Rendering Equation
2.3 The Geometry Creation
2.4 The Software Story: The All-Important APIs
2.4.1 The Triangle Setup
2.4.2 Drawing and Shading
2.4.3 Triangle Sorting
2.4.4 Texture Mapping
2.5 Fill Rate, Rendering Pipelines, and Triangle Size
2.5.1 Rendering Techniques
2.5.2 Instancing
2.5.3 Aliasing
2.5.4 Scaling
2.5.5 Environment Mapping
2.6 Generating the Image: Hardware Issues
2.6.1 VPU—Visual Processing Unit
2.6.2 Multi-display
2.7 Crypto GPU
2.8 Audio, In, Out, In Again
2.9 Conclusions
References
3 The Major GPU Eras
3.1 The First Era—Transform and Lighting—DirectX 7 (1999)
3.1.1 Shading and Shaders
3.1.2 Geometry Processing
3.2 The Second Era—Programmable Shaders—DirectX 8 and 9 (2001–2006)
3.2.1 Pixel (Fragment) Shader Stage
3.2.2 How Many Shaders? Is There a Limit?
3.3 The Third Era—The Unified Shader—DirectX 10 and 11 (2006–2009)
3.3.1 Geometry Shader (2006)
3.4 The Fourth Era—Compute Shaders—DirectX 11 (2009–2015)
3.4.1 Tessellation Shader (October 2009)
3.4.2 Summary
3.5 The Fifth Era—Ray Tracing and AI—DirectX 12 (2015–2020)
3.5.1 Ray Tracing Shaders
3.5.2 Real-Time Ray Tracing with AI
3.5.3 Variable Rate Shading—2019
3.6 The Sixth Era—Mesh Shaders—DirectX 12 Ultimate (2020)
3.6.1 Primitive and Mesh Shaders—2017–2020
3.6.2 Sampler Feedback
3.7 Summary on Shading
3.7.1 Mobile
3.7.2 GPU-Compute
3.8 FLOPS Versus Fraps: Cars and GPUs
3.8.1 Why Good Enough is Not
3.9 Conclusion
References
4 The First Era of GPUs
4.1 The Golden Age—Transform, and Lighting Changes the Industry (1999–2001)
4.1.1 On Being First
4.2 First Era Discrete PC-Based GPUs
4.2.1 Glaze3D Bitboys 2.0 (1999–2001)
4.2.2 S3 Savage 2000 (November 1999)
4.2.3 ATI and Nvidia: First Era GPUs (1999–2002)
4.2.4 ATI Radeon R100—256 (April 2000)
4.2.5 Nvidia’s NV15—GeForce 2 GTS (April 2000)
4.2.6 STMicroelectronics—Imagination Technologies Kyro II (2001–2002)
4.3 The Development and History of the Integrated GPU (1999–)
4.3.1 ArtX
4.3.2 ATI’s First IGP (March 2000)
4.3.3 SiS’ First PC-Based IGP (December 2000)
4.3.4 Nvidia’s nForce 220 IGP (June 2001)
4.3.5 ATI’s IGP 320 (2002)
4.4 IGP Conclusion
4.5 The Expansion Years (2001–2016)
4.5.1 The Collapse and the Rise of Graphics Chip Companies
4.6 Conclusion
References
5 The GPU Environment—Hardware
5.1 It Takes a Village to Build a GPU
5.2 Semiconductor Technology
5.2.1 Intel Introduces Angstroms
5.2.2 Chiplets
5.3 PC Bus Architectures
5.3.1 Industry Standard Architecture: 1981
5.3.2 Micro Channel Architecture: 1987
5.3.3 Extended ISA: 1988
5.3.4 VESA Local Bus: 1992
5.3.5 Peripheral Component Interconnect: 1992
5.3.6 Accelerated Graphics Port: 1997
5.3.7 Peripheral Component Interconnect Express: 2003
5.3.8 Other I/O
5.4 GPU Video Outputs
5.4.1 VGA: 1987
5.4.2 DVI (1999–)
5.4.3 HDMI (2002–)
5.4.4 High Dynamic Range (2015)
5.4.5 DisplayPort
5.4.6 Seeing More
5.4.7 Virtual Reality Headsets
5.4.8 Augmented Reality Glasses
5.4.9 Mixed Reality Headsets
5.4.10 Monitor Synchronization: 2013–2015
5.5 Multiple AIBs in a System
5.5.1 Multi-GPIs (1996)
5.6 Conclusion
References
6 Application Program Interface (API)
6.1 Application Program Interface
6.1.1 APIs and OSs
6.1.2 History of DirectX
6.1.3 The History of OpenGL
6.1.4 The Fahrenheit Project
6.1.5 Low-Level APIs
6.1.6 WebGPU
6.1.7 DirectX 12
6.1.8 Microsoft DirectX Shader Model 4.0: Enhancements
6.2 Conclusion
References
7 The GPU Environment—Software Extensions and Custom Features
7.1 Software Libraries and Tools
7.1.1 Ambient Occlusion
7.1.2 Nvidia’s DLSS (February 2019)
7.1.3 AMD’s Fidelity FX Super Resolution (May 2021)
7.1.4 Intel’s XeSS (March 2022)
7.2 More Than a Driver
7.2.1 SYCL
7.2.2 GLSL
7.2.3 HLSL
7.2.4 SPIR-V
7.2.5 Textures
7.3 Summary
7.4 Software Development Kits for Developers
7.4.1 Nvidia’s GameWorks (2014-)
7.4.2 AMD’s FX Library (2014)
7.4.3 AMD’s GPUOpen (2015–)
7.4.4 Application Enhancement Software
7.4.5 AMD’s Gaming Evolved
7.4.6 Nvidia’s GeForce Experience
7.5 Conclusion
7.6 Summary
References
Appendix A Acronyms
Appendix B Definitions
Index
