Accelerating a Smart and Interconnected World. Stefano Zammattio Product Marketing Manager April 2018

Transcription

1 Accelerating a Smart and Interconnected World Stefano Zammattio Product Marketing Manager April 2018

2 By 2020 Avg. Internet user Autonomous vehicles Smart Factory Connected airplane Source: Amalgamation of analyst data and Intel analysis. 1.5 GB of traffic / Day 4 TB of data / Day 1 PB of Data / Day 5 TB of data / Day The Coming Flood of Data 50 Billion connected devices!!

3 a Smart and Connected world Cloud & DATA Center MEMORY FPGA Things & Devices

4 4 At Intel We re powering the future of computing and communications delivering experiences once thought to be impossible Autonomous Driving 5g networks Artificial intelligence Virtual Worlds

5 Acceleration choices Acceleration of compute means HETEROGENEOUS COMPUTE Host CPU DEDICATED ACCELERATORS for maximum compute efficiency of specific, stable functions ASSP/ ASIC FPGA/ GPU VERSATILE ACCELERATORS for customized and changing workloads in networking, storage, and compute

6 Intel delivers end to end Platforms Devices/edge NETWORK Cloud/Data center CUSTOM Memory & Storage 4g/5G Software

7 Acceleration with FPGA - Use cases Lookaside Acceleration Inline Acceleration Network Device Intel Xeon Processor Intel FPGA Devices Storage Memory Memory FPGA devices support direct connection to data without loading the processor or CPU memory

8 DEPLOYING FPGA Acceleration in the CLOUD Public and Private Cloud Users Workload End-User Developed Functions Intel- Developed Functions 3 rd -Party Developed Functions Examples: Machine Learning Encryption, Compression Big Data Analytics NFV, vswitch Launch Workload Orchestration Software (FPGA-Enabled) Resource Pool Open Programmable Acceleration Engine (OPAE) Storage Network Compute VM AF VM= Virtual Machine AF = Accelerator Function

9 Why Intel FPGAs are Critical Delivering the performance of hardware with the programmability of software Flexible Reprogrammable Inherently Parallel Low Latency High Performance Energy Efficient

10 Artificial Intelligence is Transforming Industries Consumer Health Finance Retail Government Energy Transport Industrial Other Smart Assistants Chatbots Search Augmented Reality Robotics Enhanced Diagnostics Drug Discovery Patient Care Research Sensory Aids Algorithmic Trading Fraud Detection Research Personal Finance Risk Mitigation Support Experience Marketing Merchandising Loyalty Supply Chain Security Defense Data Insights Safety & Security Resident Engagement Smarter Cities Oil & Gas Exploration Smart Grid Operational Improvement Conservation Automated Cars Automated Trucking Aerospace Shipping Search & Rescue Factory Automation Predictive Maintenance Precision Agriculture Field Automation Advertising Education Gaming Professional & IT Services Telco/Media Sports

11 Intel Xeon Processor Scalable Family Foundation for AI Begin your journey with AI on the chip you know Intel Stratix 10 FPGA Flexible acceleration Accelerate the widest range of AI and other workloads & configurations Intel Movidius Myriad VPU Low power Neural networks and computer vision applications at ultra-low power Intel Nervana Neural Network Processor* Deep learning by design Accelerate the most intensive deep learning deployments with this custom-built processor All performance positioning claims are relative to other processor technologies in Intel s AI portfolio

12 Intel s AI Ecosystem Enabled For FPGA Solutions Tools/SDK Intel Deep Learning Studio Intel Computer Vision SDK Intel Deep Learning Toolkit for FPGA Frameworks BIGDL Performance Libraries Intel Math Kernal Library (Intel MKL) Intel Math Kernel Library for Deep Neural Networks (Intel MKL-DNN) Intel ngraph hardware +

13 Intel FPGA Focuses on Inferencing Step 1: TRAINING (In Data Center Over Hours/Days/Weeks) Step 2: INFERENCING (At the Edge or in the Data Center - Instantaneous) Massive data sets: labeled or tagged input data New input from camera and sensors Create Deep neural net math model Trained model Trained neural network model Output classification 90% person 8% traffic light 97% person Output classification

14 What s Inside the Intel Computer Vision SDK Beta Deep Learning Deployment Toolkit OpenCV* Optimized Libraries & OpenVX* Cross-platform approach to deep learning inference Model Optimizer Convert & optimize trained models Inference Engine Run optimized inferences Optimized functions for Intel Processors Create own customer kernels or use a library of functions Runtimes, emulator, kernels, workload samples Enhanced, graphical development using Vision Algorithm Designer Hardware Support GPU FPGA CPU GPU CPU GPU IPU CPU Deep Learning Frameworks OpenCL Driver for Intel Architecture Deep Learning Tools from Intel OpenVX and the OpenVX logo are trademarks of the Khronos Group Inc. OpenCL and the OpenCL logo are trademarks of Apple Inc. used by permission by Khronos

15 Intel FPGA Deep Learning Acceleration Suite Caffe TensorFlow Standard ML Frameworks Supports common software frameworks (Caffe, Tensorflow) Intel DL software stack provides graph optimizations Intel FPGA Deep Learning Acceleration Suite provides turnkey or customized CNN acceleration for common topologies Model Optimizer Intel Deep Learning Deployment Toolkit Optimized Acceleration Engine Inference Engine Pre-compiled Graph Architectures GoogleNet Optimized Template ResNet Optimized Template SqueezeNet Optimized Template VGG Optimized Template Additional, Generic CNN Templates Feature Map Cache Conv PE Array Crossbar Memory Reader /Writer Config Engine DDR DDR DDR DDR Intel Xeon Processor Hardware Customization Supported

16 Intel FPGA Deep Learning Acceleration Suite Caffe TensorFlow Standard ML Frameworks Supports common software frameworks (Caffe, Tensorflow) Intel DL software stack provides graph optimizations Intel FPGA Deep Learning Acceleration Suite provides turnkey or customized CNN acceleration for common topologies Model Optimizer Intel Deep Learning Deployment Toolkit Optimized Acceleration Engine Inference Engine DLA SW API Intel Xeon Processor Heterogenous CPU/FPGA Deployment Intel Arria FPGA Pre-compiled Graph Architectures GoogleNet Optimized Template ResNet Optimized Template SqueezeNet Optimized Template VGG Optimized Template Additional, Generic CNN Templates Hardware Customization Supported Feature Map Cache Conv PE Array Crossbar Memory Reader /Writer Config Engine DDR DDR DDR DDR

17 Introducing : Intel Programmable Acceleration Card with Intel Arria 10 GX FPGA Enabled by Acceleration Stack for Intel Xeon CPU with FPGAs For broad adoption in the data center

18 Intel portal for all things related to FPGA acceleration * 01.org is an open source community site Acceleration Stack for Intel Xeon with FPGAs FPGA Acceleration Platforms Acceleration Solutions & Ecosystem Knowledge Center FPGA as a Service Academia 01.org *

19 Intel technology leadership Make the world s best semiconductors Be the Leading end to end platform provider for the new data world Lead the Artificial intelligence & Autonomous revolution Deliver the best customer experiences on the planet

20 Inventing the Future Others predict the future. At Intel, we're building it.

21 21 Workshop: Thomas Jakobsen & Yasser Bajwa: Grazper Technologies Intro and tutorial on how to create an artificial brain and build it into a product Part 1: Introduction to embedded AI using deep learning How to create artificial intelligence in a form that can be used in real-time in a physical product Part 2: Optimized deep learning on FPGA How to make the artificial brain function in a compact product, for example, an intelligent camera.

22

23 Legal Notices and Disclaimers Intel technologies features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration. No computer system can be absolutely secure. Check with your system manufacturer or retailer or learn more at intel.com. Tests document performance of components on a particular test, in specific systems. Differences in hardware, software, or configuration will affect actual performance. Consult other sources of information to evaluate performance as you consider your purchase. For more complete information about performance and benchmark results, visit Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit Cost reduction scenarios described are intended as examples of how a given Intel-based product, in the specified circumstances and configurations, may affect future costs and provide cost savings. Circumstances will vary. Intel does not guarantee any costs or cost reduction. This document contains information on products, services and/or processes in development. All information provided here is subject to change without notice. Contact your Intel representative to obtain the latest forecast, schedule, specifications and roadmaps. No license (express or implied, by estoppel or otherwise) to any intellectual property rights is granted by this document. Intel does not control or audit third-party benchmark data or the web sites referenced in this document. You should visit the referenced web site and confirm whether referenced data are accurate. Intel Corporation. Intel, the Intel logo, and others are trademarks of Intel Corporation or its subsidiaries in the U.S. and/or other countries. *Other names and brands may be claimed as the property of others.