SlideShare a Scribd company logo
AMD EMBEDDED
SOLUTIONS (AES)

Cots moves to multi-core
Cameron Swen
TYPES OF MULTI-CORE PROCESSORS




      Homogeneous                Heterogeneous
    Multi-core Processor       Multi-core Processor




2   COTS Moves to Multi-core
MULTI-CORE PROCESSORS TUNED FOR TARGET MARKETS



                                  “Bulldozer”
                                  Performance &
                                    Scalability
                                                     High Performance and Multi-processor Applications




                                                                 Low Power Applications
                                  “Bobcat”
                               Flexible, Low Power
                                     & Small




3   COTS Moves to Multi-core
MULTI-CORE BENEFITS




                                                         Core 1   Core 2

 Performance per Watt
 Deterministic behavior
    – Real-time software is not interrupted by GUI operations
 Reliability
    – Isolate critical system functions from non-critical functions
 Security
    – Isolate communications or user interface from sensitive data

4    COTS Moves to Multi-core
A NEW ERA OF MULTI-CORE PROCESSOR DESIGN


                                                                                                                       Heterogeneous
                   Single-Core Era                                  Multi-Core Era
                                                                                                                        Systems Era


      Constrained by:                                            Constrained by:                                        Enabled by:
            Power                                                         Power                                     Abundant data parallelism
          Complexity                                             Parallel SW availability                            Power efficient GPUs
                                                                       Scalability
                                                                                                                       Constrained by:
                                                                                                                       Programming models




                                                                                            Targeted Application
Single-thread
Performance




                                                   Performance




                                                                                                Performance
                                                   Throughput



                                               ?
                                                                                we are
                                                                                 here
                                      we are
                                       here
                                                                                                                          we are
                                                                                                                           here


                              Time                                           Time                                               Time
                                                                      (# of processors)                               (Data-parallel exploitation)




5          COTS Moves to Multi-core
A NEW ERA OF PROCESSOR DESIGN & PERFORMANCE

                          Microprocessor Advancement
    CPU



                      Single-Core         Multi-Core      Heterogeneous
                          Era                Era           Systems Era




                                                              Heterogeneous    System-level
                                                                Computing     programmable
    Programmability




                                                                               OpenCL/DX
                                    Homogeneous                                driver-based
                                                                                programs
                                     Computing




                                                                                              Advancement
                                                                                                  GPU
                                                                                 Graphics
                                                                               driver-based
                                                                                programs




                                     Throughput Performance                      GPU



6       COTS Moves to Multi-core
TRADITIONAL X86 ARCHITECTURE




                               01010101010101                  010101010101010
                               10101010101010 NORTH            101010101010101
                               01010101010101 BRIDGE           010101010101010
                               10101010101010                  101010101010101




CPUs are designed for:                  A Northbridge is designed for:   GPUs are designed for:
• General purpose tasks                 • Controlling communications     • Graphics tasks
  (e.g. primary PC usage, calculations)    among the CPU, GPU, RAM,        (e.g. video rendering, display
• Common applications (Windows®,           BIOS and the Southbridge        output)
    Spreadsheets, Word processing,…)                                     •   Most visual applications (3D-
•   Serial data processing                                                   rendering, HD Video playback)
                                                                         •   Parallel data processing




7   COTS Moves to Multi-core
INTRODUCING THE AMD ACCELERATED PROCESSING UNIT (APU)




                                  APUs are the next generation of AMD processors, with the
                                combined power of AMD CPU technologies and discrete-class,
                                       DirectX®11 capable, AMD Radeon™ graphics.




 8   COTS Moves to Multi-core
MULTI-CORE APU BENEFITS



     Performance Per Watt                                                      Platform Scalability                                             Parallel Processing


 •       Take full advantage of                                    •           A single platform can scale                                       •   Leverage parallel
          parallel processing                                                   from 1 to multiple cores                                           processing to get the
                                                                                                                                                  maximum performance
                                                                           •     Scale x86 and graphics                                                from the APU
                                                                                      performance
                                                                                                                                              • Increase performance

     Gflops/Watt
                                                                       •        Scale APU power from 5-                                       without adding cost or power
                                                                                         18W                                                          to the system

     5                                                                                                                                           • OpenCL™ allows
                                                                                                                                                 programmers to preserve
     0                                                                                                                                         their expensive source code
                 Athlon™ II
                                      G-Series
                                                                                                                                                investment across multiple
                    P320
                                                                                                                                                   product generations.




Based on performance per watt comparisons between AMD Fusion APUs and the AMD Athlon™ II P320 CPU combined with the AMD Mobility Radeon™ HD 4250 GPU. In testing conducted by
AMD performance labs, AMD Fusion APUs demonstrated the following: A-Series-up to approximately 500 GFLOPS; E-Series/C-Series-up to approximately 90 GFLOPS at 18/9 W. In
comparison, the AMD Athlon™ II P320 CPU and AMD Mobility Radeon HD 4250 GPU deliver a combined total of 74 GLOPS at 38 W. Requires application support for AMD Accelerated Parallel
Processing (APP) technology. AMD Accelerated Parallel Processing technology works with applications designed to take advantage of GPU acceleration capabilities.


 9       COTS Moves to Multi-core
AMD EMBEDDED G-SERIES PLATFORM
FUELING THE INNOVATION FOR TOMORROW’S
TECHNOLOGY…TODAY!

                                THANK YOU!



                          Industry           Embedded
                        Standards            Innovation




10   COTS Moves to Multi-core
Disclaimer
     The information presented in this document is for informational purposes only and may contain technical inaccuracies,
     omissions and typographical errors.

     The information contained herein is subject to change and may be rendered inaccurate for many reasons, including but
     not limited to product and roadmap changes, component and motherboard version changes, new model and/or
     product releases, product differences between differing manufacturers, software changes, BIOS flashes, firmware
     upgrades, or the like. AMD assumes no obligation to update or otherwise correct or revise this information. However,
     AMD reserves the right to revise this information and to make changes from time to time to the content hereof without
     obligation of AMD to notify any person of such revisions or changes.

     AMD MAKES NO REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE CONTENTS HEREOF AND ASSUMES NO
     RESPONSIBILITY FOR ANY INACCURACIES, ERRORS OR OMISSIONS THAT MAY APPEAR IN THIS INFORMATION.
     AMD SPECIFICALLY DISCLAIMS ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR
     PURPOSE. IN NO EVENT WILL AMD BE LIABLE TO ANY PERSON FOR ANY DIRECT, INDIRECT, SPECIAL OR OTHER
     CONSEQUENTIAL DAMAGES ARISING FROM THE USE OF ANY INFORMATION CONTAINED HEREIN, EVEN IF AMD IS
     EXPRESSLY ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.




     Trademark Attribution
     AMD, the AMD Arrow logo, AMD Virtualization, AMD-V, Radeon, and combinations thereof are trademarks of Advanced Micro
     Devices, Inc. HDMI is a trademark of HDMI Licensing, LLC. DirectX and Windows are registered trademarks of Microsoft
     Corporation in the United States and/or other jurisdictions. OpenCL is a trademark of Apple Inc. used by permission by
     Khronos. PCIe and PCI Express are registered trademarks of PCI-SIG. Other names used in this presentation are for
     identification purposes only and may be trademarks of their respective owners.

     ©2011 Advanced Micro Devices, Inc. All rights reserved.




11      COTS Moves to Multi-core

More Related Content

PDF
An FPGA-based Scalable Simulation Accelerator for Tile Architectures @HEART2011
PDF
Nvidia Cuda Apps Jun27 11
PDF
Track A-Shmuel Panijel, Windriver
PDF
2013 02 08 annunci power 7 plus sito cta
PDF
05 2012 power_roadshow_software_on_power
PPT
Modeling System Behaviors: A Better Paradigm on Prototyping
PDF
Dme presentation-feb2013v2-1
PDF
Toward a practical “HPC Cloud”: Performance tuning of a virtualized HPC cluster
An FPGA-based Scalable Simulation Accelerator for Tile Architectures @HEART2011
Nvidia Cuda Apps Jun27 11
Track A-Shmuel Panijel, Windriver
2013 02 08 annunci power 7 plus sito cta
05 2012 power_roadshow_software_on_power
Modeling System Behaviors: A Better Paradigm on Prototyping
Dme presentation-feb2013v2-1
Toward a practical “HPC Cloud”: Performance tuning of a virtualized HPC cluster

What's hot (18)

PDF
Ibm power7
PDF
2D Games to HPC
PDF
3 d to_hpc
PDF
3 d to _hpc
PDF
Windows Server 2012 Hyper-V Networking Evolved
PDF
Shunra VE Network Appliance
PDF
Xen summit spring2010_tom_woller_amd
PDF
KM-H3000E
PDF
Toward a practical “HPC Cloud”: Performance tuning of a virtualized HPC cluster
PDF
AMD Opteron 6200 and 4200 Series Presentation
 
PDF
Infrastruttura Efficiente Di Sun E Amd -Virtualise with Confidence
PDF
My Ph.D. Research
PDF
AMD Analyst Day 2009: Rick Bergman
 
PDF
AJA IO Express
PDF
GPU - HD Video White Paper
PDF
V Evohd Intel
PDF
B.Tech Final Project
Ibm power7
2D Games to HPC
3 d to_hpc
3 d to _hpc
Windows Server 2012 Hyper-V Networking Evolved
Shunra VE Network Appliance
Xen summit spring2010_tom_woller_amd
KM-H3000E
Toward a practical “HPC Cloud”: Performance tuning of a virtualized HPC cluster
AMD Opteron 6200 and 4200 Series Presentation
 
Infrastruttura Efficiente Di Sun E Amd -Virtualise with Confidence
My Ph.D. Research
AMD Analyst Day 2009: Rick Bergman
 
AJA IO Express
GPU - HD Video White Paper
V Evohd Intel
B.Tech Final Project
Ad

Similar to Cots moves to multicore: AMD (20)

PDF
AFDS 2011 Phil Rogers Keynote: “The Programmer’s Guide to the APU Galaxy.”
PDF
ARM Techcon Keynote 2012: Sensor Integration and Improved User Experiences at...
PDF
Ximea - the pc camera, 90 gflps smart camera
PDF
Performance and scalability of Informix ultimate warehouse edtion on Intel Xe...
PDF
Engineered Systems: Oracle’s Vision for the Future
PPTX
Software Parallelisation & Platform Generation for Heterogeneous Multicore Ar...
PDF
Fremtidens platform til koncernsystemer (IBM System z)
PDF
IBM System z - zEnterprise a future platform for enterprise systems
PPTX
Tech Ed09 India Ver M New
PDF
Introduction to GPU Programming
PDF
Os Madsen Block
PPTX
Veloxum corporate introduction for crowdfunder may 29 2012
PDF
Enabling Cognitive Workloads on the Cloud: GPUs with Mesos, Docker and Marath...
PDF
Enea OSE Datasheet
PDF
Heterogeneous Systems Architecture: The Next Area of Computing Innovation
 
PDF
Simulation Directed Co-Design from Smartphones to Supercomputers
PDF
Http Jaoo.Com.Au Sydney 2008 File Path= Jaoo Aus2008 Slides Dave Thomas Lif...
PDF
Mellanox hpc day 2011 kiev
PDF
A series presentation
PDF
Talk on Parallel Computing at IGWA
AFDS 2011 Phil Rogers Keynote: “The Programmer’s Guide to the APU Galaxy.”
ARM Techcon Keynote 2012: Sensor Integration and Improved User Experiences at...
Ximea - the pc camera, 90 gflps smart camera
Performance and scalability of Informix ultimate warehouse edtion on Intel Xe...
Engineered Systems: Oracle’s Vision for the Future
Software Parallelisation & Platform Generation for Heterogeneous Multicore Ar...
Fremtidens platform til koncernsystemer (IBM System z)
IBM System z - zEnterprise a future platform for enterprise systems
Tech Ed09 India Ver M New
Introduction to GPU Programming
Os Madsen Block
Veloxum corporate introduction for crowdfunder may 29 2012
Enabling Cognitive Workloads on the Cloud: GPUs with Mesos, Docker and Marath...
Enea OSE Datasheet
Heterogeneous Systems Architecture: The Next Area of Computing Innovation
 
Simulation Directed Co-Design from Smartphones to Supercomputers
Http Jaoo.Com.Au Sydney 2008 File Path= Jaoo Aus2008 Slides Dave Thomas Lif...
Mellanox hpc day 2011 kiev
A series presentation
Talk on Parallel Computing at IGWA
Ad

Cots moves to multicore: AMD

  • 1. AMD EMBEDDED SOLUTIONS (AES) Cots moves to multi-core Cameron Swen
  • 2. TYPES OF MULTI-CORE PROCESSORS Homogeneous Heterogeneous Multi-core Processor Multi-core Processor 2 COTS Moves to Multi-core
  • 3. MULTI-CORE PROCESSORS TUNED FOR TARGET MARKETS “Bulldozer” Performance & Scalability High Performance and Multi-processor Applications Low Power Applications “Bobcat” Flexible, Low Power & Small 3 COTS Moves to Multi-core
  • 4. MULTI-CORE BENEFITS Core 1 Core 2  Performance per Watt  Deterministic behavior – Real-time software is not interrupted by GUI operations  Reliability – Isolate critical system functions from non-critical functions  Security – Isolate communications or user interface from sensitive data 4 COTS Moves to Multi-core
  • 5. A NEW ERA OF MULTI-CORE PROCESSOR DESIGN Heterogeneous Single-Core Era Multi-Core Era Systems Era Constrained by: Constrained by: Enabled by: Power Power  Abundant data parallelism Complexity Parallel SW availability  Power efficient GPUs Scalability Constrained by: Programming models Targeted Application Single-thread Performance Performance Performance Throughput ? we are here we are here we are here Time Time Time (# of processors) (Data-parallel exploitation) 5 COTS Moves to Multi-core
  • 6. A NEW ERA OF PROCESSOR DESIGN & PERFORMANCE Microprocessor Advancement CPU Single-Core Multi-Core Heterogeneous Era Era Systems Era Heterogeneous System-level Computing programmable Programmability OpenCL/DX Homogeneous driver-based programs Computing Advancement GPU Graphics driver-based programs Throughput Performance GPU 6 COTS Moves to Multi-core
  • 7. TRADITIONAL X86 ARCHITECTURE 01010101010101 010101010101010 10101010101010 NORTH 101010101010101 01010101010101 BRIDGE 010101010101010 10101010101010 101010101010101 CPUs are designed for: A Northbridge is designed for: GPUs are designed for: • General purpose tasks • Controlling communications • Graphics tasks (e.g. primary PC usage, calculations) among the CPU, GPU, RAM, (e.g. video rendering, display • Common applications (Windows®, BIOS and the Southbridge output) Spreadsheets, Word processing,…) • Most visual applications (3D- • Serial data processing rendering, HD Video playback) • Parallel data processing 7 COTS Moves to Multi-core
  • 8. INTRODUCING THE AMD ACCELERATED PROCESSING UNIT (APU) APUs are the next generation of AMD processors, with the combined power of AMD CPU technologies and discrete-class, DirectX®11 capable, AMD Radeon™ graphics. 8 COTS Moves to Multi-core
  • 9. MULTI-CORE APU BENEFITS Performance Per Watt Platform Scalability Parallel Processing • Take full advantage of • A single platform can scale • Leverage parallel parallel processing from 1 to multiple cores processing to get the maximum performance • Scale x86 and graphics from the APU performance • Increase performance Gflops/Watt • Scale APU power from 5- without adding cost or power 18W to the system 5 • OpenCL™ allows programmers to preserve 0 their expensive source code Athlon™ II G-Series investment across multiple P320 product generations. Based on performance per watt comparisons between AMD Fusion APUs and the AMD Athlon™ II P320 CPU combined with the AMD Mobility Radeon™ HD 4250 GPU. In testing conducted by AMD performance labs, AMD Fusion APUs demonstrated the following: A-Series-up to approximately 500 GFLOPS; E-Series/C-Series-up to approximately 90 GFLOPS at 18/9 W. In comparison, the AMD Athlon™ II P320 CPU and AMD Mobility Radeon HD 4250 GPU deliver a combined total of 74 GLOPS at 38 W. Requires application support for AMD Accelerated Parallel Processing (APP) technology. AMD Accelerated Parallel Processing technology works with applications designed to take advantage of GPU acceleration capabilities. 9 COTS Moves to Multi-core
  • 10. AMD EMBEDDED G-SERIES PLATFORM FUELING THE INNOVATION FOR TOMORROW’S TECHNOLOGY…TODAY! THANK YOU! Industry Embedded Standards Innovation 10 COTS Moves to Multi-core
  • 11. Disclaimer The information presented in this document is for informational purposes only and may contain technical inaccuracies, omissions and typographical errors. The information contained herein is subject to change and may be rendered inaccurate for many reasons, including but not limited to product and roadmap changes, component and motherboard version changes, new model and/or product releases, product differences between differing manufacturers, software changes, BIOS flashes, firmware upgrades, or the like. AMD assumes no obligation to update or otherwise correct or revise this information. However, AMD reserves the right to revise this information and to make changes from time to time to the content hereof without obligation of AMD to notify any person of such revisions or changes. AMD MAKES NO REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE CONTENTS HEREOF AND ASSUMES NO RESPONSIBILITY FOR ANY INACCURACIES, ERRORS OR OMISSIONS THAT MAY APPEAR IN THIS INFORMATION. AMD SPECIFICALLY DISCLAIMS ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE. IN NO EVENT WILL AMD BE LIABLE TO ANY PERSON FOR ANY DIRECT, INDIRECT, SPECIAL OR OTHER CONSEQUENTIAL DAMAGES ARISING FROM THE USE OF ANY INFORMATION CONTAINED HEREIN, EVEN IF AMD IS EXPRESSLY ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. Trademark Attribution AMD, the AMD Arrow logo, AMD Virtualization, AMD-V, Radeon, and combinations thereof are trademarks of Advanced Micro Devices, Inc. HDMI is a trademark of HDMI Licensing, LLC. DirectX and Windows are registered trademarks of Microsoft Corporation in the United States and/or other jurisdictions. OpenCL is a trademark of Apple Inc. used by permission by Khronos. PCIe and PCI Express are registered trademarks of PCI-SIG. Other names used in this presentation are for identification purposes only and may be trademarks of their respective owners. ©2011 Advanced Micro Devices, Inc. All rights reserved. 11 COTS Moves to Multi-core