Axa Assurance Maroc - Insurer Innovation Award 2024
AMD Vega Presentation - GPU Memory Architecture
1. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Cautionary Statement
This presentation contains forward-looking statements concerning Advanced Micro Devices, Inc. (AMD) including the features, functionality, expectations, timing and availability of AMD's future products, which are made pursuant to the Safe
Harbor provisions of the Private Securities Litigation Reform Act of 1995. Forward-looking statements are commonly identified by words such as "would," "may," "expects," "believes," "plans," "intends," "projects" and other terms with
similar meaning. Investors are cautioned that the forward-looking statements in this document are based on current beliefs, assumptions and expectations, speak only as of the date of this presentation and involve risks and uncertainties that
could cause actual results to differ materially from current expectations. Such statements are subject to certain known and unknown risks and uncertainties, many of which are difficult to predict and generally beyond AMD's control, that
could cause actual results and other future events to differ materially from those expressed in, or implied or projected by, the forward-looking information and statements. Material factors that could cause actual results to differ materially
from current expectations include, without limitation, the following: Intel Corporation’s dominance of the microprocessor market and its aggressive business practices may limit AMD’s ability to compete effectively; AMD is party to a wafer
supply agreement with GLOBALFOUNDRIES Inc. (GF) with obligations to manufacture products at GF with certain exceptions. If GF is not able to satisfy AMD’s manufacturing requirements, its business could be adversely impacted; AMD relies
on third parties to manufacture its products, and if they are unable to do so on a timely basis in sufficient quantities and using competitive technologies, AMD’s business could be materially adversely affected; failure to achieve expected
manufacturing yields for AMD’s products could negatively impact its financial results; the success of AMD’s business is dependent upon its ability to introduce products on a timely basis with features and performance levels that provide
value to its customers while supporting and coinciding with significant industry transitions; if AMD cannot generate sufficient revenue and operating cash flow or obtain external financing, it may face a cash shortfall and be unable to make all
of its planned investments in research and development or other strategic investments; the loss of a significant customer may have a material adverse effect on AMD; AMD’s receipt of revenue from its semi-custom SoC products is
dependent upon its technology being designed into third-party products and the success of those products; global economic uncertainty may adversely impact AMD’s business and operating results; the markets in which AMD’s products are
sold are highly competitive; AMD may not be able to generate sufficient cash to service its debt obligations or meet its working capital requirements; AMD has a substantial amount of indebtedness which could adversely affect its financial
position and prevent it from implementing its strategy or fulfilling its contractual obligations; the agreements governing AMD’s notes and the secured revolving line of credit (Secured Revolving Line of Credit) impose restrictions on AMD that
may adversely affect its ability to operate its business; uncertainties involving the ordering and shipment of AMD’s products could materially adversely affect it; the demand for AMD’s products depends in part on the market conditions in the
industries into which they are sold. Fluctuations in demand for AMD’s products or a market decline in any of these industries could have a material adverse effect on its results of operations; AMD’s ability to design and introduce new
products in a timely manner is dependent upon third-party intellectual property; AMD depends on third-party companies for the design, manufacture and supply of motherboards, software and other computer platform components to
support its business; if AMD loses Microsoft Corporation’s support for its products or other software vendors do not design and develop software to run on AMD’s products, its ability to sell its products could be materially adversely affected;
AMD’s reliance on third-party distributors and AIB partners subjects it to certain risks; AMD’s inability to continue to attract and retain qualified personnel may hinder its product development programs; our issuance to West Coast Hitech L.P
(WCH) of warrants to purchase 75 million shares of our common stock, if and when exercised, will dilute the ownership interests of our existing stockholders, and the conversion of the 2.125% Notes may dilute the ownership interest of our
existing stockholders, or may otherwise depress the price of our common stock; in the event of a change of control, AMD may not be able to repurchase its outstanding debt as required by the applicable indentures and its Secured Revolving
Line of Credit, which would result in a default under the indentures and its Secured Revolving Line of Credit; the semiconductor industry is highly cyclical and has experienced severe downturns that have materially adversely affected, and
may continue to materially adversely affect its business in the future; acquisitions, divestitures and/or joint ventures could disrupt its business, harm its financial condition and operating results or dilute, or adversely affect the price of, its
common stock; AMD’s business is dependent upon the proper functioning of its internal business processes and information systems and modification or interruption of such systems may disrupt its business, processes and internal controls;
data breaches and cyber-attacks could compromise AMD’s intellectual property or other sensitive information, be costly to remediate and cause significant damage to its business and reputation; AMD’s operating results are subject to
quarterly and seasonal sales patterns; if essential equipment, materials or manufacturing processes are not available to manufacture its products, AMD could be materially adversely affected; if AMD’s products are not compatible with some
or all industry-standard software and hardware, it could be materially adversely affected; costs related to defective products could have a material adverse effect on AMD; if AMD fails to maintain the efficiency of its supply chain as it
responds to changes in customer demand for its products, its business could be materially adversely affected; AMD outsources to third parties certain supply-chain logistics functions, including portions of its product distribution,
transportation management and information technology support services; the completion and impact of the 2015 restructuring plan, its transformation initiatives and any future restructuring actions could adversely affect it; AMD may incur
future impairments of goodwill; AMD’s worldwide operations are subject to political, legal and economic risks and natural disasters, which could have a material adverse effect on it; worldwide political conditions may adversely affect
demand for AMD’s products; unfavorable currency exchange rate fluctuations could adversely affect AMD; AMD’s inability to effectively control the sales of its products on the gray market could have a material adverse effect on it; if AMD
cannot adequately protect its technology or other intellectual property in the United States and abroad, through patents, copyrights, trade secrets, trademarks and other measures, it may lose a competitive advantage and incur significant
expenses; AMD is a party to litigation and may become a party to other claims or litigation that could cause it to incur substantial costs or pay substantial damages or prohibit it from selling its products; AMD’s business is subject to potential
tax liabilities; and AMD is subject to environmental laws, conflict minerals-related provisions of the Dodd-Frank Wall Street Reform and Consumer Protection Act as well as a variety of other laws or regulations that could result in additional
costs and liabilities. Investors are urged to review in detail the risks and uncertainties in AMD's Securities and Exchange Commission filings, including but not limited to AMD's Quarterly Report on Form 10-Q for the quarter ended September
24, 2016.
2. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Vega
AMD’s Next-Generation GPU Architecture
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
3. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM ESTCONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
We want rich, lavish
virtual worlds.
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
4. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
We want to create with limitless detail,
in real time.
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
5. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM ESTCONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
We want to make decisions
based on exabytes of data
in an instant.
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
6. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM ESTCONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
GPUs are taking on more
diverse workloads
WORKSTATION
Physically Based Rendering
Physics Modeling
Loom (VR)
Hi-Res HDR Content Creation
GAMING
4K VR
Consoles
New Rendering Pipelines
New APIs
eSports
COMPUTE
Machine Learning
Image Recognition / Computer Vision
Natural Data Processing
GPU
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
7. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Conventional architectures are
not scaling to meet needs
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
8. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Game install sizes are expanding exponentially
Gigabytes
RelativeDataSize
Deus Ex Series Install Disk Size (source: Steam)
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Chart for illustrative purposes
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
9. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
The Lord of
the Rings
Fellowship of the
Ring
The Hobbit
An Unexpected
Journey
Pro graphics data sets are well into the petabytes
Petabytes
RelativeAssetSize
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Avatar
The Hobbit
The Desolation of
Smaug
The Hobbit
Battle of the Five
Armies
The BFG
See endnote for details
10. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Compute workloads have shot into the exabytes
Character Recognition Object Detection
Image Recognition
Image/Video Recognition
1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017
See endnote for details
Data Point Too Big to
Illustrate
Exabytes
RelativeTrainingSetSize
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
11. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Growth in processing power is outpacing
growth in memory capacity
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Relative GPU Compute (GFLOPS)
Relative GPU Storage Capacity
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST See endnote for details
12. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Introducing the world’s
most scalable GPU
memory architecture
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
13. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
High-Bandwidth Cache
High-Bandwidth
Cache
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
14. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
bandwidth per pin*
2XHBM2
*vs HBM
15. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
HBM2
Over 50% smaller footprint
HBM2 vs. GDDR5
8X Capacity/stack*
*vs HBM
16. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
High-Bandwidth
Cache Controller
High-Bandwidth
Cache
HBCC
NV RAM
Network
Storage
System DRAM
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
17. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
virtual address space
512 TBHigh-Bandwidth
Cache Controller
18. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Adaptive, fine-grained
data movement
High-Bandwidth
Cache Controller
19. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Total Allocations Accessed
(Ultra 4K)
Time
(Ultra 4K)
See endnotes for details
Time
GraphicsMemory
GraphicsMemory
20. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM ESTCONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Image from Deus Ex: Mankind Divided™ courtesy of Eidos Montreal
21. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM ESTCONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST Image from Deus Ex: Mankind Divided™ courtesy of Eidos Montreal
22. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
New Programmable
Geometry Pipeline
High-Bandwidth
Cache
HBCC
NV RAM
Network
Storage
System DRAM
Geometry
Pipeline
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
23. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
peak throughput
per clock
2XOver
New Programmable
Geometry Pipeline
See endnotes for details
26. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Next-Generation
Compute Engine
High-Bandwidth
Cache
HBCC
NV RAM
Network
Storage
System DRAM
Geometry
Pipeline
Compute
Engine
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
27. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
I N T R O D U C I N G
VegaNCU
Next-GenerationComputeUnit
512
8-bit ops
per clock
256
16-bit ops
per clock
128
32-bit ops
per clock
Double Precision Rate is Configurable
*See endnotes for details
28. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Rapid Packed Math
Supercharges performance of emerging workloads
29. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
NCU is optimized for higher clock
speeds and higher IPC
CU*
NCU
*See endnotes for details
30. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
We’ve been working on reducing memory
bandwidth consumption for many years
Texture & Color CompressionFastZ ClearHiZ
31. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Next Generation
Pixel Engine
High-Bandwidth
Cache
HBCC
NV RAM
Network
Storage
System DRAM
Geometry
Pipeline
Compute
Engine
Pixel
Engine
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
32. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Fetch once enabled by smart primitive
rasterization with on-chip bin cache
Shade once enabled by culling of pixels
invisible to final scene
Draw Stream Binning Rasterizer
Designed to improve performance and saves power
33. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
L1
Compute
Engine
Pixel
Engine
Geometry
Engine
L1
L1
L2
Memory
Controller
GDDR5
GDDR5
GDDR5
GDDR5
GDDR5
GDDR5
GDDR5
GDDR5
Legacy Architecture –
Non-coherent Pixel
and Texture Memory
Access
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
34. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Render back-ends
are now clients of
the L2 cache.
High-Bandwidth
Cache
HBCC
NV RAM
Network
Storage
System DRAM
Geometry
Pipeline
Compute
Engine
Pixel
Engine L1
L1
L1
L2
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
35. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
Helps improve
performance with
applications that use
deferred shading.
High-Bandwidth
Cache
HBCC
NV RAM
Network
Storage
System DRAM
Geometry
Pipeline
Compute
Engine
Pixel
Engine L1
L1
L1
L2
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
36. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM ESTCONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
L1
Compute Engine
Pixel Engine
Geometry Pipeline
L1
L1
L2
High Bandwidth
Cache Controller
High-Bandwidth
Cache
NVRAM
Network Storage
System DRAM
CPU MM Display XDMA PCIe®
CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
37. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM ESTCONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
New Programmable Geometry
Pipeline
Revolutionary
High Bandwidth Cache
Advanced
Pixel Engine
Vega
GPU Architecturefor the Immersive and Instinctive Computing Era
Next-Gen Compute Unit
38. CONFIDENTIAL | UNDER NDA UNTIL JANUARY 5, 2017, 9AM EST
ENDNOTES
Pro graphics data set slide: Data provided by a third party studio and not verified by AMD. Data is historic - cinema asset data sizes for Lord of the Rings (2001) @.15PB, Avatar (2009) @ 1 PB, The
Hobbit –Part1 (2012) @ 1.4 PB, The Hobbit –Part2 (2013) @ 1.8 PB, The Hobbit –Part3 (2014) @ 2.3 PB, and The BFG (2016) @ 3 PB. VG-6
Compute workload data set slide: Typical word character recognition data set defined as 18.3 MB (http://wordnet.princeton.edu/wordnet/download/old-versions/ ). Object Identification datasets
defined as 490MB (http://host.robots.ox.ac.uk/pascal/VOC/databases.html#VOC2005_1). Image recognition defined as 144 GB (http://www.image-net.org/challenges/LSVRC/2010/download-all-
nonpub). Image and video recognition datasets defined as 144 GB (http://image-net.org/challenges/LSVRC/2015/) Natural Data Analysis datasets defined as 2.5QB(http://www.vcloudnews.com/every-
day-big-data-statistics-2-5-quintillion-bytes-of-data-created-daily/). VG-7
Growth in processing power slide: Data based on historic product specs; GPU relative frame buffer size vs relative TFLOP capability. The ATI Radeon 9700 Pro was 0.026 TFLOPs with 128 MB
framebuffer. The ATI Radeon X950 XT was 0.08 TFLOPs with 256 MB framebuffer. The ATI Radeon X1900 XT was 0.375 TFLOPs with 512 MB framebuffer. The ATI Radeon HD 2900 XT was 0.4755
TFLOPs with 512 MB framebuffer. The ATI Radeon HD 4870 XT was 1.2 TFLOPs with 512 MB framebuffer. The ATI Radeon HD 5870 was 1.2 TFLOPs with 512 MB framebuffer. The AMD Radeon HD
7970 was 3.79 TFLOPs with 3 GB framebuffer. The AMD Radeon R9 290X was 5.63 TFLOPs with 4 GB framebuffer. The AMD Radeon R9 Fury X was 8.6 TFLOPs with 4 GB framebuffer. VG-5
Witcher 3 and Fallout 4 data slide: Data based on AMD Internal testing of an early Vega sample using an AMD Summit Ridge pre-release CPU with 8GB DDR4 RAM, Vega GPU, Windows 10 64 bit, AMD
test driver as of Dec 5, 2016. Results may vary for final product, and performance may vary based on use of latest available drivers. VG-4
Geometry throughput slide: Data based on AMD Engineering design of Vega. Radeon R9 Fury X has 4 geometry engines and a peak of 4 polygons per clock. Vega is designed to handle up to 11 polygons
per clock with 4 geometry engines. This represents an increase of 2.6x. VG-3
CU vs Vega NCU slide: Discrete AMD Radeon™ and FirePro™ GPUs based on the Graphics Core Next architecture consist of multiple discrete execution engines known as a Compute Unit (“CU”). Each CU
contains 64 shaders (“Stream Processors”) working together. GD-78