Cortex-A75 and Cortex-A55 DynamIQ processors Powering applications from mobile to autonomous driving

Similar documents
Cortex-A75 and Cortex-A55 DynamIQ processors Powering applications from mobile to autonomous driving

A Secure and Connected Intelligent Future. Ian Smythe Senior Director Marketing, Client Business Arm Tech Symposia 2017

Advanced IP solutions enabling the autonomous driving revolution

Arm s Latest CPU for Laptop-Class Performance

ARM instruction sets and CPUs for wide-ranging applications

ARM processors driving automotive innovation

The Changing Face of Edge Compute

Maximizing heterogeneous system performance with ARM interconnect and CCIX

DynamIQ Processor Designs Using Cortex-A75 & Cortex-A55 for 5G Networks

Comprehensive Arm Solutions for Innovative Machine Learning (ML) and Computer Vision (CV) Applications

Accelerating intelligence at the edge for embedded and IoT applications

New ARMv8-R technology for real-time control in safetyrelated

Exploring System Coherency and Maximizing Performance of Mobile Memory Systems

Beyond TrustZone PSA Reed Hinkel Senior Manager Embedded Security Market Development

Each Milliwatt Matters

Compute solutions for mass deployment of autonomy

ECE 571 Advanced Microprocessor-Based Design Lecture 22

DynamIQ Processor Designs Using Cortex-A75 & Cortex- A55 for 5G Networks

WAVE ONE MAINFRAME WAVE THREE INTERNET WAVE FOUR MOBILE & CLOUD WAVE TWO PERSONAL COMPUTING & SOFTWARE Arm Limited

A Developer's Guide to Security on Cortex-M based MCUs

CCIX: a new coherent multichip interconnect for accelerated use cases

Building blocks for 64-bit Systems Development of System IP in ARM

Building High Performance, Power Efficient Cortex and Mali systems with ARM CoreLink. Robert Kaye

2017 Arm Limited. How to design an IoT SoC and get Arm CPU IP for no upfront license fee

Enable AI on Mobile Devices

Mobile & IoT Market Trends and Memory Requirements

Artificial Intelligence Enriched User Experience with ARM Technologies

Using Virtual Platforms To Improve Software Verification and Validation Efficiency

Mobile & IoT Market Trends and Memory Requirements

Mobile & IoT Market Trends and Memory Requirements

ARM big.little Technology Unleashed An Improved User Experience Delivered

Bringing Intelligence to Enterprise Storage Drives

Designing, developing, debugging ARM Cortex-A and Cortex-M heterogeneous multi-processor systems

3D Graphics in Future Mobile Devices. Steve Steele, ARM

Next Generation Enterprise Solutions from ARM

Beyond TrustZone Security Enclaves Reed Hinkel Senior Manager Embedded Security Market Develop

Beyond TrustZone PSA. Rob Coombs Security Director. Part1 - PSA Tech Seminars Arm Limited

The Next Steps in the Evolution of ARM Cortex-M

Addressing 7nm Arm DynamIQ Cluster Design Challenges Using the Cadence Digital Implementation Flow

Connect your IoT device: Bluetooth 5, , NB-IoT

Analyzing and Debugging Performance Issues with Advanced ARM CoreLink System IP Components

Unleash the DSP performance of Arm Cortex processors

Optimizing Cache Coherent Subsystem Architecture for Heterogeneous Multicore SoCs

Negotiating the Maze Getting the most out of memory systems today and tomorrow. Robert Kaye

Implementing debug. and trace access. through functional I/O. Alvin Yang Staff FAE. Arm Tech Symposia Arm Limited

Designing Security & Trust into Connected Devices

Connect Your IoT Device: Bluetooth 5, , NB-IoT

Beyond TrustZone Part 1 - PSA

Optimizing ARM SoC s with Carbon Performance Analysis Kits. ARM Technical Symposia, Fall 2014 Andy Ladd

Modeling Performance Use Cases with Traffic Profiles Over ARM AMBA Interfaces

Integrating CPU and GPU, The ARM Methodology. Edvard Sørgård, Senior Principal Graphics Architect, ARM Ian Rickards, Senior Product Manager, ARM

New Approaches to Connected Device Security

On-chip Networks Enable the Dark Silicon Advantage. Drew Wingard CTO & Co-founder Sonics, Inc.

Designing Security & Trust into Connected Devices

Profiling and Debugging OpenCL Applications with ARM Development Tools. October 2014

The Next Steps in the Evolution of Embedded Processors

Arm TrustZone Armv8-M Primer

Deep Learning on Arm Cortex-M Microcontrollers. Rod Crawford Director Software Technologies, Arm

ARM the Company ARM the Research Collaborator

Growth outside Cell Phone Applications

Making progress vs strategy

Performance Optimization for an ARM Cortex-A53 System Using Software Workloads and Cycle Accurate Models. Jason Andrews

Bringing the benefits of Cortex-M processors to FPGA

Big.LITTLE Processing with ARM Cortex -A15 & Cortex-A7

How to Build Optimized ML Applications with Arm Software

A New Security Platform for High Performance Client SoCs

Bringing Intelligence to Enterprise Storage Drives

SIMPLIFYING THE CAR. Helix chassis. Helix chassis. Helix chassis WIND RIVER HELIX CHASSIS WIND RIVER HELIX DRIVE WIND RIVER HELIX CARSYNC

Building Ultra-Low Power Wearable SoCs

Designing Security & Trust into Connected Devices

ARM: Investing for future growth

Innovation is Thriving in Semiconductors

The Challenges of System Design. Raising Performance and Reducing Power Consumption

Multicore and MIPS: Creating the next generation of SoCs. Jim Whittaker EVP MIPS Business Unit

Evolving IP configurability and the need for intelligent IP configuration

How to Build Optimized ML Applications with Arm Software

Optimizing Hardware/Software Development for Arm-Based Embedded Designs

Machine learning for the Internet of Things

Embedded segment market update

Bifrost - The GPU architecture for next five billion

Enabling Arm DynamIQ support. Dan Handley (Arm) Ionela Voinescu (Arm) Vincent Guittot (Linaro)

Arm crossplatform. VI-HPS platform October 16, Arm Limited

How to protect Automotive systems with ARM Security Architecture

ARM mbed Towards Secure, Scalable, Efficient IoT of Scale

SYSTEMS ON CHIP (SOC) FOR EMBEDDED APPLICATIONS

ARM Multimedia IP: working together to drive down system power and bandwidth

Toward a Memory-centric Architecture

Multicore platform towards automotive safety challenges

Trustzone Security IP for IoT

The Bifrost GPU architecture and the ARM Mali-G71 GPU

Communication Patterns in Safety Critical Systems for ADAS & Autonomous Vehicles Thorsten Wilmer Tech AD Berlin, 5. March 2018

Diversity of. connectivity required for scalable IoT devices. Sam Grove Principal Software Engineer Arm. Arm TechCon 2017.

Zynq-7000 All Programmable SoC Product Overview

Building firmware update: The devil is in the details

S32K Microcontroller Press Pack

OCP Engineering Workshop - Telco

Getting the Most out of Advanced ARM IP. ARM Technology Symposia November 2013

Fast, Scalable and Energy Efficient IO Solutions: Accelerating infrastructure SoC time-to-market

The ARM Cortex-A9 Processors

Formal for Everyone Challenges in Achievable Multicore Design and Verification. FMCAD 25 Oct 2012 Daryl Stewart

Transcription:

Cortex-A75 and Cortex-A55 DynamIQ processors Powering applications from mobile to autonomous driving Stefan Rosinger Director, Product Management Arm Arm TechCon 2017

Agenda Market growth and trends DynamIQ technology Arm Cortex-A75 and Cortex-A55 processors 2

Arm: The industry s architecture of choice Extraordinary growth from sensors to server 50 billion 100 billion 50 billion chips shipped chips shipped chips expected to ship 4 years 4 years 22 years 2013 1991 3 2017 2021

Arm focused on key growth market segments Mobile Client Infrastructure Embedded & Automotive 4

Looking ahead from edge to cloud Cortex beyond mobile Mixed reality Safe and autonomous Hyper-efficient Secure private compute 5

Arm architecture for total computing Cortex-A Highest performance Cortex-R Faster responsiveness Cortex-M Smallest/lowest power SecurCore Tamper resistant Designed for high-level operating systems Designed for high performance, hard realtime applications Designed for discrete processing and microcontrollers Designed for physical security 6

Innovating for the scalable future Key Arm technologies Expanding Arm technology processor architecture for broad market Up to 8 CPUs Octacore smartphones Dual cluster Heterogeneous processing Nearly Unlimited design spectrum Covers all existing use cases DynamIQ cluster Dynamic flexibility 2013 2017 7

DynamIQ technology

Imagine the possibilities Video 9

Arm DynamIQ Rearchitecting the compute experience Designed from the ground up for AI Large system performance uplift More intelligent systems 10

Accelerating AI adoption everywhere DynamIQ boosting AI/ML performance with Arm Compute Library, new CPUs and GPU Mali GPU Cortex-A CPUs Dedicated processor instructions and optimized libraries for AI 11 Accelerator Improved access to acceleration

Uncompromised performance at all tiers DynamIQ increases big.little performance levels High end Mid Entry level 4b+4L Laptop-ready compute performance 1b+7L 2x single-thread performance (vs. today s octacore) 4L Elevating your user experience 12 Example configurations shown

Intelligent power savings C P U C P U CPU Finer-grained speed control Power States Faster on/sleep/off CPU memory Autonomous power management 13

DynamIQ: New single cluster design for new cores DynamIQ big.little systems Greater product differentiation and scalability Improved energy efficiency and performance SW compatibility with Energy Aware Scheduling (EAS) Private L2 and shared L3 caches SCU Cortex-A75 32b/64b Core Private L2 cache.. Peripheral port Cortex-A55 32b/64b Core Private L2 cache Async bridges Local cache close to processors L3 cache shared between all cores ACP AMBA4 ACE DynamIQ Shared Unit (DSU) Shared L3 cache DynamIQ Shared Unit (DSU) Contains L3, Snoop Control Unit (SCU) and all cluster interfaces 1b+7L 2b+6L 4b+4L 1b+2L 1b+3L 1b+4L 14 Example: DynamIQ big.little configurations

Cortex-A75 and Cortex-A55 CPUs

New DynamIQ-based CPUs for new possibilities Cortex-A75 processor >50% more performance compared to current devices Cortex-A55 processor 2.5x higher power efficiency compared to current devices Estimated device performance using SPECINT2006, final device results may vary Comparison using Cortex-A73 at 2.4GHz vs Cortex-A75 at 3GHz Comparison using Cortex-A53 in 28nm devices vs Cortex-A55 in 16nm devices 16

New DynamIQ-based CPUs for new possibilities Cortex-A75 Cortex-A55 Geekbench v4 1.34x Geekbench v4 1.22x Octane 2.0 1.48x Octane 2.0 1.14x LMBench memcpy 1.16x LMBench memcpy 1.97x SPECFP2006 1.33x SPECFP2006 1.42x SPECINT2006 1.22x SPECINT2006 1.21x Baseline to Cortex-A73 Baseline to Cortex-A53 All comparisons at iso process and frequency All comparisons at iso process and frequency 17

L1 memory system Common features 4-way set associative VIPT with PIPT programmer s view Improved prefetchers Cortex-A75 64KB Wider load-store than Cortex-A73 Support Read-after-Write OoO with filtering Cortex-A55 L1 TLB L1 Data Cache Store Buffer Prefetcher L2 TLB L2 Cache 16KB / 32KB / 64KB Improved store buffer bandwidth to L1 Larger 16-entry L1-TLB 18

L2 memory system Common features Private L2 cache in each core Running at core speed Exclusive data cache Cache stashing into the L2 Non-blocking 1024-entry TLB for hit-under-miss Cortex-A75 L1 TLB L1 Data Cache Store Buffer Prefetcher L2 TLB L2 Cache 256KB / 512KB Cortex-A55 0KB / 64KB / 128KB / 256KB 19

DynamIQ Shared Unit (DSU) Support for multiple performance domains Core 0 DynamIQ cluster 0-7 Cores Core 7 Latency and bandwidth optimizations DynamIQ Shared Unit (DSU) Streamlines traffic across bridges Snoop filter L3 Cache Asynchronous bridges Bus I/F ACP and peripheral port I/F Power Management Advanced power management features Supports large amounts of local memory Scalable interfaces for edge to cloud applications Low latency interfaces for closely coupled accelerators 20

System solutions

Premium mobile system Complete system IP portfolio from Arm GIC-600 I/O Coherent Masters Built for premium NIC-450 Mali G72 Mali- V61 Mali- Display Scaled to mid-range and entry level Cortex-A75 Cortex-A55 MMU-500 MMU-500 Incorporates the latest features Security with Arm TrustZone Sensor hub and acceleration DMC- 500 LPDDR4 DMC- 500 Memory system, integrated TrustZone CoreLink CCI-550 Sensor fusion hub Cortex-M7 Crypto Cell- 712 Interconnect NIC-450 Peripherals 22

The best upgrade path for mid-range devices Power, performance and area of octa-core CPU systems Area Integer (SPECINT2000) Floating Point (SPECFP2000) 2.34 Memory Streaming (JMCStream) 1.46 1.21 1.00 1.02 1.08 2.25 3.58 4.06 2-4x performance (compared to LITTLE only) Cortex-A53 Octa A53 A53 L2 A53 A53 A53 A53 L2 Coherent Interconnect A53 A53 Cortex-A55 Octa A55 A55 A55 A55 A55 A55 A55 A55 L3 Interconnect Cortex-A75/Cortex-A55 Octa A75 A55 A55 A55 A55 A55 A55 A55 L3 Interconnect Same or similar area (compared to Cortex-A53 LITTLE only) 23

Safer autonomous systems DynamIQ supports safety critical industrial and automotive systems Resilient systems Allowing systems to operate safely under failure Adaptive compute Uncompromised performance for autonomous systems Faster responsiveness Quicker safety critical decision-making 24

Diverse range of automotive compute solutions Powertrain Cortex-R52 Real time Homogeneous multi-core Vision ADAS Cortex-A55 + Cortex-R52 Heterogeneous multi-core Computer vision Control Autonomous driving Cortex-A75 + Cortex-R52 High performance multi-cluster Machine learning Functional safety Infotainment Cortex-A75 + Cortex-A55 Energy-aware scheduling Rich OS Security Central body control Low power Efficient performance Scalable Other modules V2X Chassis Security Radar Sensor Audio Cortex-M7, Cortex-M0+ 25

Safety and reliability with DynamIQ Compute performance for ADAS and IVI Higher performance for autonomous cars Faster responsiveness for safety critical tasks Functional safety Autonomous system Sense Perceive Decide Actuate SoC ASIL D systematic capability Advanced RAS architectural features Industry s broadest functional safety capable CPU IP portfolio Cortex-M Sensors Cortex-A CPU CPU L2 L2 Cache L2 L2 Cache Cache L3 Cache Application cores Cortex-R Lock-step core Safety Island 26

Power-constrained ADAS platform big GIC-600 ELA LITTLE ELA Cortex-A75 Cortex-A55 DSU DSU CoreLink CCI-550 Mali-G72 GPU Safety island DCLS Cortex-R52 CoreLink NIC-450 Security Enclave CryptoCell 312 SRAM DMA I/O Isolated for highest reliability & integrity System monitoring, diagnostics & recovery Failsafe communication Mali- V61 MMU-500 Assertive Display Mali-Display CoreLink NIC-450 Mali-C71 CoreLink NIC-450 Peripherals DMC LPDDR4x 27

Summary DynamIQ technology is enabling more capable SoCs for key growth markets Brings new device solutions for all markets & tiers from mobile to automotive, and beyond Cortex-A75: Breakthrough performance Cortex-A55: Efficiency redefined Elevating the user experience for the premium doubling the performance levels for the mid-range and mass-market New single cluster approach for better scalability and flexibility New DynamIQ Shared Unit, new memory hierarchy, new advanced power management features 28

Arm architecture for total computing SoC IP Ecosystem Security Intelligence Widest, most proven choice of IP to meet diverse PPA needs Worlds #1 embedded software ecosystem Support across all Cortex-A and some Cortex-M CPUs Arm Compute Library & DynamIQ accelerate intelligence at the edge 28 billion Arm-based embedded chips shipped* 29 *As of CY Q1 2017

For further information Find demos and more information at the Arm booth (402) and Mbed booth (712) At TechCon After TechCon Advanced IP solutions to enable the autonomous automotive revolution James Scobie Functional Safety: What is Arm doing to support this critical capability? Neil Stroud Cortex-A75 and Cortex-A55 DynamIQ processors powering applications from mobile to autonomous drive Stefan Rosinger Panel: What does functional safety mean for the automotive supply chain? Moderator Andrew Moore https://developer.arm.com stefan.rosinger@arm.com 30

Thank You! Danke! Merci! 谢谢! ありがとう! Gracias! Kiitos! 31

The Arm trademarks featured in this presentation are registered trademarks or trademarks of Arm Limited (or its subsidiaries) in the US and/or elsewhere. All rights reserved. All other marks featured may be trademarks of their respective owners. www.arm.com/company/policies/trademarks 32