A Survey on Reconfigurable Accelerators for Cloud Computing

Slides:



Advertisements
Similar presentations
INTRODUCTION TO CLOUD COMPUTING CS 595 LECTURE 6 2/13/2015.
Advertisements

What is Cloud Computing? o Cloud computing:- is a style of computing in which dynamically scalable and often virtualized resources are provided as a service.
M.A.Doman Model for enabling the delivery of computing as a SERVICE.
© Hitachi Data Systems Corporation All rights reserved. 1 1 Det går pænt stærkt! Tony Franck Senior Solution Manager.
Presenter MaxAcademy Lecture Series – V1.0, September 2011 Introduction and Motivation.
Cloud Computing – The Cloud Dr. Jie Liu. Definition  Cloud computing is Web-based processing, whereby shared resources, software, and information are.
Oricane AB Breakthrough in Green Software Technology.
Networking Virtualization Using FPGAs Russell Tessier, Deepak Unnikrishnan, Dong Yin, and Lixin Gao Reconfigurable Computing Group Department of Electrical.
PhD course - Milan, March /09/ Some additional words about cloud computing Lionel Brunie National Institute of Applied Science (INSA) LIRIS.
Workload Optimized Processor
Windows Azure Conference 2014 Deploy your Java workloads on Windows Azure.
Advanced Computer Architecture, CSE 520 Generating FPGA-Accelerated DFT Libraries Chi-Li Yu Nov. 13, 2007.
For Testbeds TM. Secure, multi-tenant cloud orchestration platform –Turnkey platform for delivering IaaS clouds –Hypervisor agnostic –Massively scalable,
March 9, 2015 San Jose Compute Engineering Workshop.
Heterogeneous CPU/GPU co- processor clusters Michael Fruchtman.
Next Generation Operating Systems Zeljko Susnjar, Cisco CTG June 2015.
Motivations for Innovations in Operational Excellence Bruce Rodin VP – Wireless Technology Bell Canada.
Cloud Architecture. SPI Model Cloud Computing Classification Model – SPI Cloud Computing Classification Model – SPI - SaaS: (Software as a Service) -
Philipp Gysel ECE Department University of California, Davis
1 EIT 2.2 Is your company missing out on the cost-savings opportunities offered by data center consolidations? Andy Abbas Co-Founder and Vice President.
© 2007 IBM Corporation IBM Software Strategy Group IBM Google Announcement on Internet-Scale Computing (“Cloud Computing Model”) Oct 8, 2007 IBM Confidential.
Software architectures and tools for highly distributed applications Voldemaras Žitkus.
Cloud Agility with Performance Bridging the Performance Gap for Virtual Network Infrastructure Paul Andersen Sr. Marketing Director.
Data Management Strategies Technology acquisition planning Business Continuity consulting IDS Analytics Pro assessment software EMPOWER yourself Keith.
GIS IN THE CLOUD Cloud computing furnishes scalable GIS technology that is maintained off premises and delivered on demand as services via the Internet.
Hadoop Javad Azimi May What is Hadoop? Software platform that lets one easily write and run applications that process vast amounts of data. It includes:
Prof. Jong-Moon Chung’s Lecture Notes at Yonsei University
The Holmes Platform and Applications
SAS® Viya™ Overview ANDRĖ DE WAAL, GLOBAL ACADEMIC PROGRAM
CLOUD ARCHITECTURE Many organizations and researchers have defined the architecture for cloud computing. Basically the whole system can be divided into.
Performance Assurance for Large Scale Big Data Systems
Connected Infrastructure
Organizations Are Embracing New Opportunities
Summary Remaining Challenges The Future Messages to Take Home.
FPGAs for next gen DAQ and Computing systems at CERN
Introduction to Distributed Platforms
Dynamo: A Runtime Codesign Environment
Initial Experiences with Deploying FPGA Accelerators in Datacenters
Big Data A Quick Review on Analytical Tools
OSS: Java, Open Source Data Infra
IOT Critical Impact on DC Design
Status and Challenges: January 2017
Astronomical Data Processing & Workflow Scheduling in cloud
The Multikernel: A New OS Architecture for Scalable Multicore Systems
Windows Server 2016 Platform for Modern Apps Microsoft Build 2016
Couchbase Server is a NoSQL Database with a SQL-Based Query Language
HPE Persistent Memory Microsoft Ignite 2017
Recap: introduction to e-science
Connected Infrastructure
The Improvement of PaaS Platform ZENG Shu-Qing, Xu Jie-Bin 2010 First International Conference on Networking and Distributed Computing SQUARE.
OSS: Java, Open Source Data Infra
Cloud Computing By P.Mahesh
Low Latency Analytics HPC Clusters
NSF : CIF21 DIBBs: Middleware and High Performance Analytics Libraries for Scalable Data Science PI: Geoffrey C. Fox Software: MIDAS HPC-ABDS.
Xuechao Wei, Peng Zhang, Cody Hao Yu, and Jim Wu
Cloud Computing Dr. Sharad Saxena.
Highly Efficient and Flexible Video Encoder on CPU+FPGA Platform
Your Next LIMS: SaaS or On-Premise? Presented by:
SaaS Developer Helps Small North American Oil and Gas Firms Reduce Costs with Azure Power “The latest generation of our oil and gas reporting software.
Chapter 1 Introduction.
Technical Capabilities
IBM Power Systems.
Greg Stitt ECE Department University of Florida
2/25/2019.
Agenda Need of Cloud Computing What is Cloud Computing
Salesforce.com Salesforce.com is the world leader in on-demand customer relationship management (CRM) services Manages sales, marketing, customer service,
ANTOL, HOLIČ, KEDA PA195 SPRING 2016
CS 239 – Big Data Systems Fall 2018
Copyright © JanBask Training. All rights reserved Get Started with Hadoop Hive HiveQL Languages.
Presentation transcript:

A Survey on Reconfigurable Accelerators for Cloud Computing Dr. Christoforos Kachris, Prof. Dimitrios Soudris ICCS/NTUA, Greece FPL 2016 1 September 2016

Accelerators in data centers By 2020, Intel predicts a third of cloud providers will use FPGAs, analysts noted in a keynote at their annual data center event… FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016 FPGA 2014:   FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

Data Center Requirements Traffic requirements increase significantly in the data centers but the power budget remains the same (Source: ITRS, HiPEAC, Cisco) FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

Hardware accelerators HW acceleration can be used to reduce significantly the execution time and the energy consumption of several applications (10x-100x) A solution that can be used to overcome this problem is the use of application-specific accelerators. Specialized multicore processors with application-specific acceleration modules can leverage the underutilized die area to overcome the initial power barrier, delivering significantly higher performance for the same power envelope. The main idea is to use the abundant die area by implementing application-specific accelerators and dynamically powering up only those accelerators suitable for a given workload. [Source: Xilinx, 2016] FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

Google application Specific Accelerators deployed in DC Google Has Built A Custom Chip For Machine Learning The result is called a Tensor Processing Unit (TPU), a custom ASIC we built specifically for machine learning — and tailored for TensorFlow. Google has been running TPUs inside the data centers for more than a year, and have found them to deliver an order of magnitude better-optimized performance per watt for machine learning. This is roughly equivalent to fast-forwarding technology about seven years into the future (three generations of Moore’s Law). FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

A survey on HW accelerator for Cloud computing HW accelerators Search engine and Page ranking MapReduce Spark Memcached Databases FPGAs in the cloud framework FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

Web search and Page Ranking MS Catapult: Bing web search engine 95% higher throughput per server Or, (while maintaining equivalent throughput) Tail latency: reduced by 29%. FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

MapReduce Accelerator C. Kachris, D. Diamantopoulos, G. C. Sirakoulis, and D. Soudris, “An fpga-based integrated mapreduce accelerator platform,” Journal of Signal Processing Systems, pp. 1–13, 2016. C. Kachris, G. C. Sirakoulis, and D. Soudris, “A reconfigurable mapreduce accelerator for multi-core all-programmable socs,” in System-on-Chip (SoC), 2014 International Symposium on, Oct 2014, pp. 1–6 FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016 Spark Accelerator J. Cong, M. Huang, D. Wu, and C. H. Yu, “Invited – heterogeneous datacenters: Options and opportunities,” in Proceedings of the 53rd Annual Design Automation Conference, ser. DAC ’16. New York, NY, USA: ACM, 2016, pp. 16:1–16:6 When Apache Spark Meets FPGAs: A Case Study for Next-Generation DNA Sequencing Acceleration Deploying Accelerators At Datacenter Scale Using Spark, Spark Summit FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

Memcached accelerator 36x in RPS/Watt with low variation M. Blott, L. Liu, K. Karras, and K. Vissers, “Scaling out to a single-node 80gbps memcached server with 40terabytes of memory,” in Proceedings of the 7th USENIX Conference on Hot Topics in Storage and File Systems, ser. HotStorage’15. Berkeley, CA, USA: USENIX Association, 2015 FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016 In-memory Databases 7x to 14x speedup for most queries Source: [B. Sukhwani, H. Min, M. Thoennes, P. Dube, B. Brezzo, S. Asaad, and D. E. Dillenberger, “Database analytics: A reconfigurable-computing approach,” IEEE Micro, vol. 34, no. 1, pp. 19–29, Jan 2014.] FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016 SQL Databases [Source:  Jian Ouyang, Baidu, Hot Chips 2016] Baidu has recently presented an FPGA-based acceleration for data centers for the SQL databases FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

A survey on HW accelerator for Cloud computing HW accelerators Search engine and Page ranking MapReduce Spark Memcached Databases FPGAs in the cloud framework FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

IBM’s OpenPower IP Store FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

Intel’s vision on IP Store FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

RC3E, Dresden University Source: [O. Knodel and R. G. Spallek, “RC3E: provision and management of reconfigurable hardware accelerators in a cloud environment,” in 2nd International Workshop on FPGAs for Software Programmers, 2015] FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016 The VINEYARD approach An App-store for Hardware accelerators as IPs Foster the development of an eco-system with Hardware accelerators as IPs in the same way as software packages. Load the required functions based on the application requirements [ www.vineyard-h2020.eu ] FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

HW Accelerators for Cloud Computing FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

Speedup vs Energy efficiency FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

Batch vs Streaming applications FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016 Speedup per category Page Rank applications achieve the higher speedup Memcached application achieve higher energy efficiency FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

Communication Interface Designs with PCIe offers the higher speedup But due to communication overhead offers low energy efficiency FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016 HDL vs HLL HDL and HLLs offer almost the same speedup! HDL: Higher energy efficiency (but this may depend also on the application) FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

FPGAs in HyperScale Data Centers The ecosystem of Hardware IPs in the embedded system world can be adopted in the data centers. Accelerators IPs can foster the innovation of IPs in the domain of cloud computing and big data analytics FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016 Roadmap Paradigm shift (From Homogeneous Data Centers to Heterogeneous Data Centers) IaaS, PaaS, SaaS for accelerators 3rd party Hardware IP developers contribute to a common market place for Hardware Accelerators in the same way as Embedded systems FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016 Convergence on Os Vendor Specific OS in mobiles Vendor Agnostic OS Vendor Agnostic OS, Architecture specific Vendor Specific OS in PCs FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

Convergence on FPGA AppStore Vendor-specific accelerator Accelerator1 FPGA VendorA VendorB GPU VendorC VendorD Accelerator2 … Vendor-agnostic Platform-specific Platform-agnostic IP Store Options FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

Roadmap on FPGAs in the Cloud Compress FPGA Xilinx (a,b,…) Altera (a,b,..) Compress FPGA Xilinx Altera Special HW accel Compress Compress FPGA GPU Xeon Phi Vendor-specific AppStore Platform-agnostic AppStore Vendor-agnostic Platform-specific AppStore FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016

FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016 Thank you for your time Questions? More info: kachris@microlab.ntua.gr www.vineyard-h2020.eu This work has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 687628 - VINEYARD FPL 2016, Christoforos Kachris, ICCS/NTUA, September 2016