OpenPOWER Summit North America 2019: Full Schedule

A strong and sustainable ecosystem with diverse OpenPOWER technology implementations is the key to reaching the goals of the Foundation. It’s an organic moment for the Foundation to evolve to new niches and markets. There is much demand for high-performance scalable and affordable storage systems not limited by constraints of traditional tech and design approaches. YADRO proves the technical advantages and applicability of the OpenPOWER technology for storage with the market success of the storage platform YADRO TATLIN.

Speakers

Nikolay Semenov

Product Manager, Yadro

I am a product manager at YADRO responsible for server products. My product family is VESNIN and appliances made on this platform.

2019 OPS USA keynote Semenov v3 pdf

Monday August 19, 2019 09:05 - 09:15 PDT
Harbor ABC

Keynote Sessions

Session Slides Included YES

09:15 PDT

Keynote: Innovative OpenPOWER Servers Bring Impactful Alternatives to China Market - John Hu, GM of Inspur Power Systems, SVP of Inspur*

IPS is using POWER technology with leveraging X86 experience and asset, to create multiple modality and better cost preference servers which fulfill Enterprise and Internet company’s demand.

Speakers

John Hu

Vice President, Inspur Group Co., Ltd.

Mr. John Hu currently serves as Vice President of Inspur Group Co., Ltd, Vice Director of State Key Laboratory of High-end Server & Storage Technology, Vice President of Inspur Electronic Information Industry Co., Ltd, and General Manager of Inspur Power Commercial Systems Co., Ltd... Read More →

03 John Hu Inspur 20190816a FINAL pdf

Monday August 19, 2019 09:15 - 09:25 PDT
Harbor ABC

Keynote Sessions

Session Slides Included YES

09:25 PDT

Keynote: IBM OpenPOWER Roadmap - Steve Fields, Fellow & Chief Engineer of Power Systems, IBM*

Speakers

Steve Fields

IBM Fellow, Chief Engineer of Power Systems, IBM

04 Steve Fields IBM 20190818a pdf

Monday August 19, 2019 09:25 - 09:35 PDT
Harbor ABC

Keynote Sessions

Session Slides Included YES

09:35 PDT

Keynote: Pete Hazen, VP & Business Unit Manager, Data Center Solutions, Microsemi*

Speakers

Pete Hazen

VP & Business Unit Manager, Data Center Solutions, Microchip

Pete Hazen is Vice President and Business Unit Manager for Data Center Solutions at Microchip (Nasdaq: MCHP). In this role, Pete is responsible for storage and memory connectivity solutions as well as SSD controller solutions for next generation data centers. Pete is a 33-year industry... Read More →

05 Pete Hazen Microchip 20190816a FINAL pdf

Monday August 19, 2019 09:35 - 09:45 PDT
Harbor ABC

Keynote Sessions

Session Slides Included YES

09:45 PDT

Keynote: Brian Pan, General Manager, H3 Platform Inc.*

Speakers

Brian Pan

General Manager, H3 Platform Inc.

Brian Pan is the founder of H3 Platform, a leading cloud solution provider that provide the PCIe composable technology to cloud operators. He received a BA degree in ME of NCKU and MBA master degree of NCCU. His is engaging with Tier 1 AI research lab in North American The biggest... Read More →

06 Brian Pan H3 Platform And Wistron 20190818b FINAL WITH NOTES pdf

Monday August 19, 2019 09:45 - 09:55 PDT
Harbor ABC

Keynote Sessions

Session Slides Included YES

09:55 PDT

Keynote: Timothy Pearson, Chief Technology Officer, Raptor Computing Systems*

Speakers

Timothy Pearson

Chief Technology Officer, Raptor Computing Systems

Timothy has worked in multiple areas of high performance systems designover the past 15 years, from hardware / firmware to kernel andapplication development, with a strong emphasis on open solutions andsystem security. In his time at Raptor he has been heavily involved inthe inception... Read More →

07 Tim Pearson RaptorCS 20190818b FINAL pdf

Monday August 19, 2019 09:55 - 10:05 PDT
Harbor ABC

Keynote Sessions

Session Slides Included YES

10:05 PDT

Keynote: Closing Remarks - Hugh Blemings, Executive Director, OpenPOWER Foundation

Speakers

Hugh Blemings

Executive Director, OpenPOWER Foundation

Monday August 19, 2019 10:05 - 10:10 PDT
Harbor ABC

Keynote Sessions

10:10 PDT

Coffee Break

Monday August 19, 2019 10:10 - 10:40 PDT
Harbor Foyer

Event Experiences / Breaks / Meals

10:40 PDT

Introduction & Welcome - Peter Hofstee, IBM

Monday August 19, 2019 10:40 - 10:50 PDT
Harbor Ballroom D (Track 1)

OpenCAPI

10:40 PDT

Pack-em or Stack-em: Advantages of Running OpenShift on Power - Krishna Harsha Voora, IBM*

In this session, you will hear about the (container density, pod scaling) tests we performed with OpenShift Container Platform (OCP) and compared different architectures. You will ascertain the effectiveness of using POWER Systems with OpenShift Container Platform which can host most containers or pods without performance degradation compared to other platforms. A “DevOps” scenario with a workload similar to a MEAN stack (Mongo, Express, Angular, and Node) is selected. The workload was purposeful to further demonstrate the price-performance advantage of using OpenShift with POWER Platform.

Speakers

Krishna Harsha Voora

Software Developer - Cognitive Systems, IBM India Pvt. Ltd.

Krishna Harsha Voora is a consultant in UNIX® and Linux operating systems with IBM Systems and Development Lab India. He is actively contributing to Hyperledger Fabric on POWER and he evangelizes Blockchain technology in various forums, conferences, and meet-ups. He also actively... Read More →

Advantages Of Running OpenShift On Power pdf

Monday August 19, 2019 10:40 - 11:20 PDT
Harbor Ballroom E (Track 2)

Cloud

Session Slides Included YES

10:40 PDT

VT OpenPOWER Hackathon: A Story of New Users, Open OnDemand and PowerAI Acceleration - Robert Settlage, Virginia Tech

In Spring of 2019, Virginia Tech hosted an OpenPOWER Hackathon with the aim of showing users how the PowerAI ecosystem can accelerate AI/DL frameworks with little to no modifications. Many of the students that signed up for the Hackathon were new to Linux and HPC. As a limited duration Hackathon with a goal of using the PowerAI toolset, it was important to get new users into the Power system quickly. Here we will discuss the Open OnDemand platform that allows for creating custom GUIs and portals of entry into HPC systems. We will show how this system allowed users to focus more on the Hackathon topic and give summaries of the winning projects and acceleration results achieved.

Speakers

Robert E. Settlage

Computational Scientist, United States

Dr. Settlage is a Computational Scientist and Data Analyst within the Advanced Research Computing(ARC) group at Virginia Tech. The ARC is a VT community resource to assist faculty in handling andanalyzing large and complex data sets. The ARC is building on Dr. Settlage’s prior experiences... Read More →

Monday August 19, 2019 10:40 - 11:20 PDT
Harbor Ballroom F (Track 3)

HPC OpenPOWER Ecosystem

10:50 PDT

Implementation of an FPGA Accelerator with Coherent Memory Access in Apache Arrow Format for Fully Homomorphic Encryption using IBM's CAPI - Noushin Azami, Texas State University*

We aim to implement a fully homomorphic encryption algorithm using an FPGA accelerator in a way that eliminates the overhead produced by the device driver and memory format transitions. Fletcher is a framework that integrates FPGA accelerators with Apache Arrow, an in-memory format that reduces overhead when transferring data between platforms by creating tabular memory blocks.
Fletcher is programmed to function on top of IBM’s CAPI SNAP which eliminates the need for a device driver in an FPGA accelerator, reducing the overhead of exchanging data between FPGA and CPU and giving the FPGA coherent access to data in the main memory.

Speakers

Noushin Azami

graduate student, Texas State University

Noushin is an Iranian masters student in electrical engineering at Texas State University, she is currently working on her thesis and is graduating in August. she is continuing her studies toward PhD in CS at the same school.

OpenPowerReady1 Copy pdf

Monday August 19, 2019 10:50 - 11:20 PDT
Harbor Ballroom D (Track 1)

FPGA OpenCAPI

Session Slides Included YES

11:25 PDT

D4P: The Power Platform for Docker Online Container Authoring - Abdul Dakkak, UIUC

The current POWER container community lags behind that of some other architecture. Part of this is because software developers are unable to publish docker images without access to a Power system. Docker for Power (D4P) is a platform to develop and deploy Power containers within a web browser without access to a local Power system.

The aim of D4P is to enrich the Power container ecosystem by providing both a platform for developers to create docker containers and for Power community to find docker images. Already, we have built and published over 200 docker images that are available in the D4P image catalog. User
contribution is key to extend D4P's catalog.

D4P is available online https://dockerfile-builder.mybluemix.net and slated to be the hub for the Power
community to create, discover, and use docker images.

Speakers

Abdul Dakkak

PhD Student, UIUC

Abdul Dakkak is a PhD candidate in Computer Science at the University of Illinois at Urbana-Champaign (UIUC). He is a research assistant in the IMPACT Research Group working with Professor. Wen-mei Hwu. Abdul’s research interest are in functional programming languages and how they... Read More →

Monday August 19, 2019 11:25 - 12:05 PDT
Harbor Ballroom E (Track 2)

Cloud

11:25 PDT

The Fletcher Framework: Bringing Apache Arrow to FPGA - Johan Peltenburg, Delft University of Technology*

Modern big data systems are highly heterogeneous. Components are implemented in a wide variety of programming languages and frameworks. Due to implementation differences, interfaces between components are burdened by serialization overhead. The Apache Arrow project helps to overcome this burden through a language-agnostic columnar in-memory format for big data applications. The open-source Fletcher framework is an implementation of Arrow for FPGA accelerators. Fletcher allows developers to communicate in the Arrow format through specialized, high-performance and easy-to-use hardware interfaces. Serialization overhead is prevented, and integration with over 11 high-level languages is made possible and efficient. The OpenPOWER CAPI SNAP framework is a supported platforms of Fletcher. We will go over the benefits of Apache Arrow and Fletcher and show a hands-on example.

Speakers

Johan Peltenburg

PhD Candidate, Delft University of Technology

Johan Peltenburg is a PhD Candidate from the Delft University of Technology in The Netherlands. His research focuses on heterogeneous computing for big data applications. Johan received his B. Eng in Electrical Engineering at the Rotterdam University of Applied Science, followed by... Read More →

Fletcher OpenPOWER NA 2019 pdf

Monday August 19, 2019 11:25 - 12:05 PDT
Harbor Ballroom D (Track 1)

FPGA

Session Slides Included YES

11:25 PDT

OpenPower Lab@Unicamp and Minicloud: Free Developer Cloud for Open Source and Academic Members. Updates and Call-for-input. - Klaus Heinrich Kiwi, IBM*

UNICAMP (State University of Campinas) is an important OpenPower academic member providing community services such as Minicloud (http://openpower.ic.unicamp.br/minicloud/) that is one of the most widely used "free" developer clouds in the OpenPower ecosystem, providing virtual images hosting to dozens of Open-Source projects, academic students and faculty, as well as local IBM partners working on porting their solutions to POWER. This session will bring an update about our OpenPOWER Lab@Unicamp activities, including some statistics of usage such as geography, distro flavor, technology themes, etc. This session will also update the community on the current work to use the minicloud as a CI validation for multiple projects and will collect feedback and input on what other projects might be good targets to bring into minicloud, including a possible partnership with other foundation members.

Speakers

Klaus Heinrich Kiwi

Engineering Manager, OpenPower System Software Workgroup co-chair, OpenPower@UNICAMP Lab leader, IBM

Klaus has been working with Linux on Power development on IBM since 2006 on various topics such as Security, Cryptography, Performance, I/O, Toolchain and Linux Distro enablement Library development. More recently, Klaus took the responsibility for the OPAL Open-Source firmware and... Read More →

OpenPower Lab pdf

Monday August 19, 2019 11:25 - 12:05 PDT
Harbor Ballroom F (Track 3)

HPC OpenPOWER Ecosystem

Session Slides Included YES

12:00 PDT

Sponsor Showcase

Monday August 19, 2019 12:00 - 12:15 PDT
Harbor Foyer

Event Experiences / Breaks / Meals

12:05 PDT

Lunch (Provided for OpenPower Attendees)

Monday August 19, 2019 12:05 - 13:05 PDT
Sponsor Showcase

Event Experiences / Breaks / Meals

13:05 PDT

CAPI/OpenCAPI Accelerated PostgreSQL - Peng Fei Gou, IBM*

Heterogeneous computing, one of the most crucial technology to OpenPOWER systems, is extended by cache-coherent accelerator interfaces, i.e., CAPI/OpenCAPI. Database, critical workload in data-analytics, AI, etc., is key to the success of OpenPOWER ecosystem. In this proposal, a CAPI/OpenCAPI and FPGA based solution is described to accelerate database queries. In particular, an architecture with multithreading software and multi-engine FPGA logic is presented, with detailed analysis on how CAPI/OpenCAPI’s advantages of latency, bandwidth and simplified software stack can be leveraged to accelerate database. The accelerated database is based on the popular open source PostgreSQL. Query operators including regular expression matching and other general queries are implemented in FPGA logic, with 10x to 2x overall throughput improvement compared with state-of-the-art CPU version.

Speakers

Peng Fei Gou

Hardware Architect, IBM

Technical lead of IBM China OpenPOWER chip development team. Years of experience on state-of-the-art processor architecture research/development, industrial level CPU and SoC verification and compiler development. Works for projects such as IBM POWER9, OpenPOWER, OpenCAPI/CAPI, Nvidia... Read More →

PostgreSQL on CAPI regex OPS 2019 pdf

Monday August 19, 2019 13:05 - 13:45 PDT
Harbor Ballroom D (Track 1)

FPGA OpenCAPI

Session Slides Included YES

13:05 PDT

The POWER of Supercomputers at your Fingertips - Rafael Sene, IBM

A healthy ecosystem is fundamental for any computer architecture to thrive nowadays and developers are the fuel that keeps it moving and evolving. They need access to resources that can empower them to get the most of their applications and amaze their customers. As the architecture that is the core of the #1 and #2 fastest supercomputers in the world, POWER is where they need to be, but how developers can leverage it as part of their daily activities? This presentation will demonstrate how everyone can get FREE access to the POWER architecture and start building, porting and tuning their applications right away. It will show how to build on POWER using GitHub, Travis CI, and GitLab and how to get FREE access to POWER resources.

Speakers

Rafael Sene

Advisory Software Engineer, IBM

Advisory Software Engineer who have been working at IBM from low level source code analysis and migration to the cloud development and deployment. Currently his focus on RedHat OpenShift Development on Power, workloads enablement and ecosystem expansion of IBM POWER by enabling teams... Read More →

Monday August 19, 2019 13:05 - 13:45 PDT
Harbor Ballroom F (Track 3)

OpenPOWER Ecosystem, Subtype1, Subtype2

13:05 PDT

OpenPOWER in Enterprise: Gaps & Challenges - Nikolay Semenov & Alexey Stepanov, Yadro*

Following the Foundation strategy to make OpenPOWER a default choice for modern workloads, at YADRO we are committed to deliver OpenPOWER solutions to Enterprises, allowing them to experience efficient hyperscale-designed hardware benefits. Enterprise world brings new requirements for system design, accompanying tools and services offered by vendor.
Making YADRO OpenPOWER products Enterprise ready, we revealed gaps and challenges that have to be addressed on a vendor side or community level. We will share YADRO field experience dealing with these gaps, so that we all could get more mature platform and products closer to Enterprise needs.

Speakers

Alexey Stepanov

R&D Engineer, YADRO

I am a member of R&D team at YADRO with focus on hardware and software troubleshooting, engineering support, as well as performance, benchmarking and proof of concept tests. My Enterprise IT career started in 1995, I was working as technical support and professional services engineer... Read More →

Nikolay Semenov

Product Manager, Yadro

I am a product manager at YADRO responsible for server products. My product family is VESNIN and appliances made on this platform.

2019 OPS USA Semenov Stepanov — Gaps and Challenges pdf

Monday August 19, 2019 13:05 - 13:45 PDT
Harbor Ballroom E (Track 2)

Systems OpenPOWER Ecosystem

Session Slides Included YES

13:50 PDT

Enabling SQL to FPGA Using HLS for Filtering and Stream Data Transformations - Erwin de Haan, Delft University of Technology*

This talk discusses automating the transpiling of SQL to High-Level Synthesis C++ to generate hardware for filtering and data stream transformations using CAPI on POWER systems. This work uses the Fletcher framework to automate the handling of data movement between memory and the FPGA. The use of HLS technologies can greatly reduce the development time of FPGAs compared to manual FPGA development workflows.
Deploying FPGAs in fast changing data processing pipelines, can be very complicated or limit the use of the FPGA hardware. This work investigates if HLS is good enough for these kinds of applications to meaningfully reduce total development time while still maintaining performance. Additionally, the use of the Fletcher framework further reduces required developer time.

Speakers

Erwin de Haan

Graduate Student, Delft University of Technology

Erwin de Haan is an MSc student of Computer Engineering and Embedded Systems at the Delft University of Technology, the Netherlands, where he received his BSc degree in Electrical Engineering in 2016. He was a research intern at the Tokyo Institute of Technology, Japan in 2018. Erwin... Read More →

Presentation Erwin de Haan pdf

Monday August 19, 2019 13:50 - 14:30 PDT
Harbor Ballroom D (Track 1)

FPGA OpenCAPI

Session Slides Included YES

13:50 PDT

AutoML for Efficient Neural Architecture Design - Ligeng Zhu, MIT*

Deep learning has drawn growing attention in recent years. The impressive performance comes at scale of computational power, which makes it hard to execute especially on mobile devices. Besides, designing a good and efficient neural-net requires a lot of engineering efforts, let alone the following platform/hardware specific deployment. To reduce the human-bandwidth, we propose ProxylessNAS -- an efficient framework that automatically specializes neural architecture for different hardwares. With >74.5% top-1 accuracy, the measured latency of ProxylessNAS is 1.8x faster than MobileNet-v2, a widely used human design for mobile vision.

Speakers

Ligeng Zhu

Research Assistant, MIT

Ligeng Zhu is a research assistant of Professor Song Han’s group at MIT. His research focuses on efficient machine learning, with a special interest in the design automation. His recent AutoML works aim to automatically search the optimal neural-net architecture for a specific task... Read More →

proxyless@summit pdf

Monday August 19, 2019 13:50 - 14:30 PDT
Harbor Ballroom E (Track 2)

Machine Learning

Session Slides Included YES

14:35 PDT

Protected Execution Facility On Power - Guerney Hunt, Ram Pai & Michael Anderson, IBM*

IBM POWER9 architecture introduces Protected Execution Facility (PEF) which enables secure memory, which can only be accessed when the processor is in a new higher privileged mode called Ultravisor mode. Some facilities that were previously hypervisor privileged are now ultravisor privileged. This talk presents architectural changes for PEF, the Protected Execution Ultravisor, and new services called ultra calls. Combined with enhancements to the Virtual Machine format a new class of VMs called Secure Virtual Machines (SVM) is enabled. SVMs utilize hypervisor services and run in secure memory under the control of the Ultravisor. SVMs are protected from attacks from the Hypervisor, other privileged software entities, and privileged users on the system. POWER9 systems can concurrently support both normal VMs and SVMs. All firmware and tools for PEF will be Open Sourced.

Speakers

Guerney Hunt

Research Staff Member, IBM

Dr. Guerney D. H. Hunt has been a Research Staff Member at IBM’s T. J. Watson Research Center since 1995. He is currently working on transferring security technology into IBM products, and developing additional security technology. He participated in a team funded by the Department... Read More →

Ram Pai

Senior Software Engineer, IBM

Ram Pai leads the development of the Secure Virtual Machines at IBM's Cognitive Systems Group.He has enabled large teams globally towards enablement of various contributions to the Linux ecosystem.Among many other features, VFS Shared-Subtree and POWER memory-key features are some... Read More →

Michael Anderson

Senior Software Engineer, IBM

Mike Anderson is a developer in IBM's Cognitive Systems Group.He is currently working on Ultravisor firmware. This work provides support for secure virtual machines (SVMs) on IBM Open Power Systems supporting Protected Execution Facility (PEF) capability.He is an open source developer... Read More →

OpenPOWER NA Final (2) pdf

Monday August 19, 2019 14:35 - 15:15 PDT
Harbor Ballroom E (Track 2)

Cloud Security

Session Slides Included YES

14:35 PDT

Thymesis-P: An Approach to Rack-scale Disaggregation Over OpenCAPI - Dimitris Syrivelis, IBM*

OpenCAPI has the potential to revolutionize the way computers are organized. Custom off-chip components may coherently master memory transactions like any other on-chip component towards system memory, while slave components get integrated at the lowest point of coherency.

Taking advantage of the exciting new capabilities, the ultra-low latency and high bandwidth we took things a step further: We developed ThymesisFlow-P, a software defined interconnect that mixes OpenCAPI M1 and C1 modes to enable the integration of off-tray components in the same spirit as off-chip counterparts and thus enable Rack-scale disaggregation. ThymesisFlow-P materializes a resource “stealing” concept, where each server can request and use resources from a neighbor. This talk presents ThymesisFlow-P approach potential, the design, a live disaggregated memory demo and shares key experiences of the development journey.

Speakers

Dimitris Syrivelis

Research Engineer, IBM

Dimitris is a member of the High Performance Systems team at the IBM Research Lab in Dublin, Ireland. His work primarily targets vertical system support for next generation datacenters. Pushing forward the vision of Software Defined Infrastructures, he is involved in projects that... Read More →

ThymesisFlowP OPSummit2019 pdf

Monday August 19, 2019 14:35 - 15:15 PDT
Harbor Ballroom D (Track 1)

OpenCAPI

Session Slides Included YES

14:35 PDT

Developing Applications using Acceleration Hardware on Power - Gerrit Huizenga, IBM*

This session will share the goals of the newly chartered Accelerated Application Ecosystem Working Group, which will help identify resources needed when porting accelerated applications to Power specific hardware accelerators. This session will provide an overview of some of the available accelerators, some of the existing porting resources, and solicit information from ISVs and software developers on any gaps in the software ecosystem today.

Speakers

Gerrit Huizenga

Power Ecosystem Lead, IBM

Gerrit is currently the technical lead for the Power Open Source Ecosystem, which includes porting open source software to Power, helping ISVs and IBM Cloud and Cognitive products port to Power, and provides leadership for the Power Developer's Portal. Gerrit is also the acting lead... Read More →

Accel App Ecosystem Talk pdf

Monday August 19, 2019 14:35 - 15:15 PDT
Harbor Ballroom F (Track 3)

OpenPOWER Ecosystem Software Development

Session Slides Included YES

15:15 PDT

Coffee Break

Monday August 19, 2019 15:15 - 15:45 PDT
Harbor Foyer

Event Experiences / Breaks / Meals

15:45 PDT

OpenCAPI Acceleration Framework: Unleash the Power of Customized Accelerators - Yong Lu, IBM*

OpenCAPI Acceleration Framework, abbreviated as OC-Accel, is a platform to enable programmers and computer engineers to quickly create FPGA-based accelerations. The acceleration action's software part and hardware part share the server host memory data through OpenCAPI interface. With it, people can quickly design an accelerator and benefit from the bandwidth, latency, coherency and programmability advantages of OpenCAPI.

Speakers

Yong Lu

OpenCAPI Acceleration Architect, IBM

Yong Lu is a hardware engineer in IBM Systems group. She leads the development of OpenCAPI acceleration framework, which is a platform to enable programmers and computer engineers to quickly create FPGA-based accelerations. She also actively creates and incubates CAPI/OpenCAPI based... Read More →

OpenCAPI Acceleration Framework YongLu OPS SanDiego2019 upload pdf

Monday August 19, 2019 15:45 - 16:25 PDT
Harbor Ballroom D (Track 1)

OpenCAPI

Session Slides Included YES

15:45 PDT

OpenPOWER Foundation Hub BoF Session - Jeffrey Scheel, IBM

Speakers

Jeffrey Scheel

Chair, Technical Steering Committee, IBM

Monday August 19, 2019 15:45 - 16:25 PDT
Harbor Ballroom F (Track 3)

OpenPOWER Ecosystem

15:45 PDT

High-Performance MPI and Deep Learning on OpenPOWER Platform - Ching-Hsiang Chu, X-ScaleSolutions

This talk will focus on high-performance and scalable middleware for Message Passing Interface (MPI) and Deep Learning on OpenPOWER platform with NVIDIA GPGPUs and RDMA-enabled interconnects (InfiniBand and RoCE). The focus will be on two packages with commercial support being available from X-ScaleSolutions. The first package will focus on the OSU MVAPICH2 MPI libraries and their capabilities for high-performance computing with both CPUs (OpenPOWER) and GPUs (NVIDIA). The second package will focus on tight integration between the OSU MVAPICH2-GDR MPI library and the Horovod stack to provide high-performance and scalable Deep Learning with multiple frameworks like TensorFlow and PyTorch. Performance results from the SUMMIT system (#1 supercomputer in the World) with thousands of GPUs and POWER9 CPUs will be presented.

Speakers

Ching-Hsiang Chu

Software Engineer, X-ScaleSolutions

Mr. Ching-Hsiang Chu received B.S. and M.S. degrees in Computer Science and Information Engineering from National Changhua University of Education, Taiwan in 2010 and National Central University, Taiwan in 2012, respectively. He is currently working towards a Ph.D. degree in Computer... Read More →

Monday August 19, 2019 15:45 - 16:25 PDT
Harbor Ballroom E (Track 2)

Performance GPU

16:30 PDT

Update from SmartDV

Speakers

Bipul Talukdar

Director-Applications Engineering, SmartDV

Monday August 19, 2019 16:30 - 16:45 PDT
Harbor Ballroom D (Track 1)

OpenCAPI FPGA

16:30 PDT

OpenPOWER in Enterprise: Implementing OpenBMC/OPAL Features - Alexey Stepanov, Yadro

Enterprises engaged in the digital transformation are welcoming high performance, low latency and speedy data access of OpenPOWER-based solutions. Unlike their counterparts in research institutions and labs, they must be confident that these solutions will also provide familiar Enterprise level of reliability, manageability, and serviceability.
FromYADRO diverse field expertise and deep interaction with Enterprise customers, we’ve collected a selection of tech requirements for systems to fit in Enterprise infrastructure. In this talk, we’ll walk you through YADRO experience and lessons learned.

Speakers

Alexey Stepanov

R&D Engineer, YADRO

2019 OPS USA tech Stepanov — OpenBMC and OPAL PDF

Monday August 19, 2019 16:30 - 17:10 PDT
Harbor Ballroom E (Track 2)

Cloud

16:30 PDT

OpenPower Solution Builder Community - John Pace, Mark III & Christopher Sullivan, Oregon State University*

The OpenPower Solution Builder Community is for those who design, implement, operate and maintain Power and OpenPower solution stacks. The Group is self-governing and provides a means for builders to share insights and innovation. There are three primary elements 1) questions & answers and pointers to relevant information, 2) best practices on building hardware-software solution stacks and reference architectures, 3) innovations of applications, libraries, methods, and approaches. The initial focus is solution stacks around the IBM POWER9 AC922 integrated GPU Server but will expand to additional systems and workloads. You do not need to be an OpenPower member to participate. This talk will discuss the community formation, topics that are active, and will solicit participation.

Speakers

Chris Sullivan

Assistant Director for Biocomputing, Oregon State University

Christopher Sullivan has been working in the field of computational science for 19 years and currently has over 30 co-authored scientific publications. His work has focused on Biocomputing and life sciences that change the way we look at and interact with the environment. Over the... Read More →

John Pace

Sr. Data Scientist, Mark III Systems

John is Senior Data Scientist for Mark III Systems.

Solution Builder Group OPF Summit 8 19 2019 pdf

Monday August 19, 2019 16:30 - 17:10 PDT
Harbor Ballroom F (Track 3)

OpenPOWER Ecosystem

Session Slides Included YES

16:45 PDT

Hybrid Memory Project and OpenCAPI - Michael Miller, Rambus*

Rambus Labs will present its Proteus Hardware Platform -- a flexible, high performance platform that is well-suited for hybrid memory research and beyond. Combining DRAM with emerging memories has shown promise as a cost-performant way of increasing system capacity.

Speakers

Michael Miller

Principle Engineer, Rambus, Inc.

Michael Miller is a Principle Engineer in Rambus Labs, the research division of Rambus, Inc. He is presently the hardware lead on the Hybrid Memory project, which is prototyping memory subsystems with multiple types of media. For most of his career, he has worked in embedded processor... Read More →

Rambus OpenPower 2019 San Diego pptx

Monday August 19, 2019 16:45 - 17:10 PDT
Harbor Ballroom D (Track 1)

OpenCAPI FPGA

Session Slides Included YES

17:15 PDT

FPGAs in the Datacentre - Steve Hebert, Nimbix

Speakers

Steve Hebert

CEO, Nimbix

Steve began his career in semiconductors with Texas Instruments after earning his degree in Electrical Engineering from Santa Clara University, having studied device physics during the peak of Moore’s law. Today, he is a technology industry veteran of over 25 years and served the... Read More →

Monday August 19, 2019 17:15 - 17:55 PDT
Harbor Ballroom D (Track 1)

OpenCAPI FPGA

17:15 PDT

OpenPower Ready Certification Process and Requirements - Ned Gamble, IBM

The Chairman of the Power Power Ready work group will present an overview of the requirements to achieve the OpenPower Ready certification for computer systems, IO adapters, operating systems and applications.

Speakers

Ned Gamble

IBM Senior Engineer, Open Power Chief Engineer and Chairman, Open Power Ready Workgroup, IBM

Ned is a Senior Engineer at IBM where he began there in the late 80s. He is currently the Open Power Chief Engineer and Chairman of the Open Power Ready Workgroup.

Monday August 19, 2019 17:15 - 17:55 PDT
Harbor Ballroom F (Track 3)

OpenPOWER Ecosystem

17:55 PDT

All-Attendee Reception & Sponsor Showcase

Monday August 19, 2019 17:55 - 18:45 PDT
Harbor Foyer

Event Experiences / Breaks / Meals

08:00 PDT

Breakfast

Tuesday August 20, 2019 08:00 - 09:00 PDT
Harbor Foyer

Event Experiences / Breaks / Meals

08:00 PDT

Registration

Tuesday August 20, 2019 08:00 - 17:00 PDT
Harbor Foyer

Event Experiences / Breaks / Meals

09:00 PDT

Keynote: Day Two Welcome - Hugh Blemings, Executive Director, OpenPOWER Foundation

Speakers

Hugh Blemings

Executive Director, OpenPOWER Foundation

Tuesday August 20, 2019 09:00 - 09:05 PDT
Harbor ABC

Keynote Sessions

09:05 PDT

Keynote: Ken King, General Manager, OpenPOWER Alliance, IBM*

Speakers

Ken King

General Manager, OpenPOWER Alliances, IBM

Ken King is General Manager, OpenPOWER Alliances, for IBM Systems & Technology Group. Ken is responsible for building the ecosystem of partners on top of the POWER architecture. This includes chip, system, software and go-to-market channel development; along with governments worldwide... Read More →

OpenPOWER Keynote Ken King 8.20 pdf

Tuesday August 20, 2019 09:05 - 09:20 PDT
Harbor ABC

Keynote Sessions

Session Slides Included YES

09:20 PDT

Keynote: Aaron Sullivan, Director of Hardware Engineering, Facebook*

Updated OpenPower 2019 AaronSullivan v3 pdf

Tuesday August 20, 2019 09:20 - 09:35 PDT
Harbor ABC

Keynote Sessions

Session Slides Included YES

09:35 PDT

Keynote & Technology Demonstration: Anton Blanchard, Distinguished Engineer, IBM & Joe DeLaere, Xilinx*

Speakers

Joe DeLaere

Alveo Marketing, Data Center Group, Xilinx

Joe is a 20+ year veteran of the technology industry covering product marketing/management and technical services roles with leading FPGA semiconductors. Today, he manages the ground-breaking Alveo Data Center accelerator card product line for Xilinx’s Data Center Group, leading... Read More →

Anton Blanchard

Distinguished Engineer, IBM

Anton has been involved with Linux and Open Source Software for over 20 years, much of that time with IBM. He leads a worldwide team dedicated to using Open Source technologies to build better products. In mid 2019he got the Open Hardware bug as a result of IBM's opening up of the... Read More →

Blanchard v4.2 pdf

Tuesday August 20, 2019 09:35 - 09:50 PDT
Harbor ABC

Keynote Sessions

Session Slides Included YES

09:50 PDT

Keynote: Michael Cunningham, Executive Vice President, Red Hat*

Speakers

Michael Cunningham

Executive Vice President, Red Hat

Michael Cunningham, is the Executive Vice President of Red Hat, Inc. Red Hat is the leading provider of open source solutions to enterprises, and is headquartered in Raleigh, NC. Cunningham has been with Red Hat since 2004.He serves on the board of a number of companies and organizations... Read More →

Michael Cunningham OpenPower Summit August 20, 2019 20190819 final pdf

Tuesday August 20, 2019 09:50 - 10:00 PDT
Harbor ABC

Keynote Sessions

Session Slides Included YES

10:00 PDT

Keynote Panel: Timothy Pearson, Raptor Computing Systems; Derek Chiou, Microsoft Azure/UT; Steve Hebert, Nimbix; Alan Clark, SUSE & Bapi Vinnakota, Netronome; Moderated by Peter Rutten, IDC

Speakers

Derek Chiou

Partner System Architect & Researcher, Microsoft & University of Texas at Austin

Alan Clark

Office of the CTO, Emerging Technologies, SUSE

Alan Clark, a member of the SUSE CTO Office, is an experienced industry, corporate leader, open source advocate and strategy adviser for new industry initiatives and open source. He helps foster the creation, growth, awareness and adoption of open source and open standards as an industry... Read More →

Timothy Pearson

Chief Technology Officer, Raptor Computing Systems

Bapi Vinnakota

Engineer, Netronome

Bapi is the Director of Silicon Architecture Program Management and OCP ODSA Workgroup Lead at Netronome. Bapi taught at the University of Minnesota, Twin Cities, immediately after a Ph.D. at Princeton. He joined Intel through an acquisition and was the architect of a flow processor... Read More →

Steve Hebert

CEO, Nimbix

Peter Rutten

Research Vice President, IDC

Peter Rutten is a Research Director within IDC's Enterprise Infrastructure Practice, covering research on computing platforms. In this role, Mr. Rutten focuses on high-end, accelerated, and heterogeneous infrastructure and their use cases, which include supercomputing, massively parallel... Read More →

Tuesday August 20, 2019 10:00 - 10:45 PDT
Harbor ABC

Keynote Sessions

10:45 PDT

Closing Remarks & Special Announcement - Mendy Furmanek, IBM; Artem Ikoev, Yadro & Hugh Blemings, OpenPOWER Foundation

Tuesday August 20, 2019 10:45 - 10:50 PDT
Harbor ABC

Keynote Sessions

10:50 PDT

Coffee Break

Tuesday August 20, 2019 10:50 - 11:20 PDT
Harbor Foyer

Event Experiences / Breaks / Meals

10:50 PDT

Sponsor Showcase

Tuesday August 20, 2019 10:50 - 17:50 PDT
Harbor Foyer

Event Experiences / Breaks / Meals

11:20 PDT

Open Announcements BoF - Anton Blanchard, IBM & Mendy Furmanek, OpenPOWER Foundation

Speakers

Anton Blanchard

Distinguished Engineer, IBM

Tuesday August 20, 2019 11:20 - 12:00 PDT
Harbor Ballroom E (Track 2)

CPUs Open Hardware

11:20 PDT

Super-Connecting the World’s Number One Supercomputers - Scot Schultz, Mellanox Technologies*

Mellanox continues leadership as the best interconnect worldwide for both HPC and AI. With the latest intelligent interconnect devices, such as Mellanox Quantum HDR 200Gb/s and ConnectX-6 are enabling the next generation of in-network co-processing. With a more effective mapping of communication between devices in the system dramatically increases system performance.

Mellanox solutions are being used today in 59% of the HPC systems listed on the June 2019 TOP500 list, including the top three supercomputers in the world, and four of the top five in the list. InfiniBand accelerates the fastest HPC and AI supercomputer in the world – the Oak Ridge National Laboratory ‘Summit’ system, delivering 148.6 Petaflops, and the second world’s fastest “Sierra” at Lawrence Livermore National Laboratory, demonstrating the robust ecosystem supported by the OpenPOWER Foundation and its membership.

Speakers

Scot Schultz

Sr. Director, HPC/AI and Technical Computing, Mellanox Technologies

Scot Shultz is Senior Director at Mellanox and specializes in High Performance Computer Systems with a focus on Artificial Intelligence and Machine Learning. His responsibilities include understanding computational, operational and high-speed interconnect requirements of High Performance... Read More →

190815 Mellanox HPC AI Advantages OpenPOWER pdf

Tuesday August 20, 2019 11:20 - 12:00 PDT
Harbor Ballroom F (Track 3)

HPC Interconnects

Session Slides Included YES

11:20 PDT

Architecture Effects When Interacting with GPU Technologies - Christopher M. Sullivan, Oregon State University*

Oregon State University’s Center for Genome Research and Biocomputing (CGRB) works closely with its users to determine the best computing platform for computational research. In the past several years we have seen many differences associated with architecture and GPU interactions. As we continued our testing we found tremendous benefits from using the OpenPOWER platform with Power9 processors and NVIDIA GPU. In an effort to truly determine the best platform for different types of GPU processing we created a set of test. We will be showing all the current data comparing some of the most bleeding-edge platforms and a demo of the machines racing each other. Our goal is to provide users with the ability to bring online the best platform for the work they are planning with GPU technology.

Speakers

Chris Sullivan

Assistant Director for Biocomputing, Oregon State University

OSU CMS OPF NASummit AITalk 08192019 widescreen pdf

Tuesday August 20, 2019 11:20 - 12:00 PDT
Harbor Ballroom D (Track 1)

HPC Performance

Session Slides Included YES

12:05 PDT

GPU Accelerated RFI Mitigation for GMRT High Resolution Southern Sky Survey - Aditya, Center for Development of Advanced Computing (C-DAC)

Fast Radio Bursts (FRBs) are momentary pulses of intense radio emissions which blast for a few milliseconds. The origin and source of FRBs is not yet identified, which make it a very exciting area of study. Due to their short-lived nature, real-time detection and analysis of FRBs become extremely difficult. One of the major and compute-intensive steps in this is radio frequency interference (RFI) calculation.

Our project under the OpenPOWER academia aims to optimize the performance of the RFI mitigation computations using OpenMP and CUDA on high-performance CPUs and GPUs. The performance and accuracy of the pipeline is currently being tested on OpenPOWER system with one Nvidia V100 and Power 9 CPU, the results and improvements in performances will be shared. Feather testing and optimization on multi GPU environments should increase the performance significantly.

Speakers

Aditya Mitkari

Machine Learning Engineer, Center for Development of Advanced Computing (C-DAC)

I am a machine learning engineer, graduated from University of Pune in 2018. I have been working on different projects under the OpenPOWER Academia Umbrella at research facilities like C-DAC and National Center for Radio Astrophysics.I have previously worked on computer vision for... Read More →

Tuesday August 20, 2019 12:05 - 12:45 PDT
Harbor Ballroom D (Track 1)

HPC GPU

12:05 PDT

Composable AI Solution For Power 9 - Brian Pan, H3 Platform Inc.

H3 Platform will talk about their Composable AI box – Falcon II. This is the 1 st PCIe Gen4 Fabric GPU box, and could connect with Power 9 server, the only model with PCIe 4.0 capability, as a GPU expansion box. Falcon II could load up to sixteen GPGPUs and share the GPGPU up to four Power 9 servers. With the dynamical allocation capability, Falcon II could compose systems with different GPGPU and host ratio to meet the different resource requirement of AI applications. For example, 16 GPGPUs to 1 Power 9 server, 2 groups of 8 GPGPUs to 1 Power 9 server, and 4 groups of 4 GPGPUs to 1 Power 9 server.

Speakers

Brian Pan

General Manager, H3 Platform Inc.

Tuesday August 20, 2019 12:05 - 12:45 PDT
Harbor Ballroom F (Track 3)

Interconnects Systems

12:45 PDT

Lunch (Provided for OpenPower Attendees)

Tuesday August 20, 2019 12:45 - 13:45 PDT
Sponsor Showcase

Event Experiences / Breaks / Meals

13:45 PDT

POWER Roadmap - Madeline Vega & Brian Allison, IBM

In this session you will learn about IBM's Next Generation POWER processor roadmap and our systems memory strategy.

Speakers

Brian Allison

Senior Technical Staff Engineer, IBM

Brian Allison is an IBM Senior Technical Staff Engineer and Chief Engineer for External Engagements including CAPI and OpenCAPI. Over his 25 year career at IBM, Mr. Allison was the Chief Engineer of various industry leading IBM chip sets that spanned cache coherent, memory and... Read More →

Madeline Vega

Senior Technical Staff Member, IBM

Madeline Vega is a Senior Technical Staff Member in the IBM Cognitive Systems Development organization. She is the Systems Platform Chief Engineer for the next generation of POWER Systems. In this role she leads cross-functional teams on the architecture, design and development... Read More →

OpenPowerSummit 2019 Roadmap session final pdf

Tuesday August 20, 2019 13:45 - 14:25 PDT
Harbor Ballroom E (Track 2)

CPUs

13:45 PDT

OpenPOWER and GPU Porting Platform for FOSS Using OpenStack at OSUOSL - Lance Albertson, OSU Open Source Lab*

The Oregon State University Open Source Lab (OSUOSL) works closely with Free and Open Source Software (FOSS) projects to provide them with a stable environment to improve their OpenPOWER porting needs. Since 2011, we’ve deployed and managed an OpenStack cluster powered by POWER9 and POWER8) used by over 100 projects. In this talk we will describe the technical challenges we have faced from setting up the initial cluster, to managing new users’ needs, tooling involved for managing the system, upgrading OpenStack and its components and other related topics. In addition, we’ll discuss a continuous integration platform we deployed which includes access to Nvidia GPU hardware along with a few other new services that are upcoming.

Speakers

Lance Albertson

Director, OSU Open Source Lab

Lance Albertson is the Director for the Oregon State University Open Source Lab (OSUOSL) and has been involved with the Gentoo Linux project as a developer and package maintainer since 2003. The OSUOSL provides hosting for more than 160 projects, including those of worldwide leaders... Read More →

OpenPOWER porting platform for FOSS using OpenStack at OSUOSL pdf

Tuesday August 20, 2019 13:45 - 14:25 PDT
Harbor Ballroom D (Track 1)

HPC GPU

Session Slides Included YES

13:45 PDT

ExpEther - Providing IO Distributed System - Gursimranjit Singh, NEC Corporation of America*

ExpEther technology virtualizes PCI Express over Ethernet. Using ExpEther, one can configure a computing system with scale-up headroom by plugging arbitrary PCI Express-based I/O devices distributed on an Ethernet LAN. ExpEther enhances the PCI Express bus by extending connection distances, supporting device numbers beyond the limitation of local PCI port counts as well as alleviating localized electrical power and heat dissipation restrictions that a conventional box-type computer would encounter. In addition, with full support of the PCI Hot-Plug protocol, dynamic reconfiguration of the computing platform is also possible.

Speakers

Gursimranjit Singh

Senior System Engineer, NEC Corporation of America

Gursimranjit Singh is currently a Senior Systems Engineer working on High Performance Compute Technologies at NEC. He specializes in NEC's PCIe protocol based interconnect solutions. Gursimranjit is also involved in developing the business strategy and use cases for NEC's next-gen... Read More →

NEC ExpEther082019 OPS pdf

Tuesday August 20, 2019 13:45 - 14:25 PDT
Harbor Ballroom F (Track 3)

Interconnects Systems

Session Slides Included YES

14:30 PDT

PowerAI Vision & Other Solutions on an AI Box - Arvind Bharani & Adith SV, Object Automation Software Solution Pvt. Ltd.

We, here at Object Automation Software Solutions Pvt Ltd are trying to combine these two momentous assets with various other solutions into a low-cost development system. This system is a power-packed AI box accelerated with NVIDIA GPUs and connected with XILINX Alveo U200 board for easy and instant deployment or inference of the build upon solutions so that enterprises or organizations can rapidly deploy their solutions generating a fully optimized and efficient performance.

This will enable students, faculties and industries to experience the ease of developing solutions leveraging Artificial Intelligence capabilities with tools like PowerAI Vision and many others on a low cost development system powered by IBM and supported by NVIDIA and XILINX, also, opening up plenty of opportunities for developers to build solutions on top of this system.

Speakers

Adith SV

Student, Object Automation Software Solution Pvt. Ltd.

I am Adith SV doing my B.Tech(ME) at Amity University. I have done a report and 3D model on the parts of a transformer. I have worked with aatral academy and made a drone model. I worked as an intern with Acetech precision engineering and Object Automation Software Solution Pvt... Read More →

Arvind Bharani

Research Intern, Object Automation Software Solution Pvt. Ltd.

This is Arvind Bharani from B.S.Abdur Rahman Crescent Institute Of Science and Technology, currently studying B.Tech Information Technology.The Projects I have worked on are CrescentSAT - Near Space Launch Vehicle (High Altitude Balloon) on February 24th,and the project currently... Read More →

Summit Presentation pptx

Tuesday August 20, 2019 14:30 - 15:10 PDT
Harbor Ballroom F (Track 3)

GPU FPGA Systems

14:30 PDT

Introduction to CHIPS - Zvonimir Z. Bandic, Western Digital Corporation*

We have recently launched a CHIPS Alliance project: CHIPS (Common Hardware for Interfaces, Processors and Systems) Alliance harnesses the energy of open source collaboration to accelerate hardware development. The organization was created to host and curate high-quality, open source hardware design relevant to the design of silicon devices. By creating a neutral and collaborative environment, CHIPS Alliance intends to share resources to lower the cost of development and accelerate the creation of more efficient and innovative chip designs – covering the span from small IoT devices to large datacenter silicon solutions.

As an independent entity, companies and individuals can work together and contribute resources to help make open source chips, complex IP blocks and system-on-a-chip (SoC) design more accessible to the market.

Speakers

Zvonimir Bandic

Senior Director, Western Digital

Zvonimir Z. Bandić is the Research Staff Member and Senior Director of Next Generation Platform Technologies Department in a Western Digital Corporation in San Jose, California. He received his BS in electrical engineering in 1994 from the University of Belgrade, Yugoslavia, and... Read More →

Introduction to CHIPS Alliance August 20 2019 v3 pdf

Tuesday August 20, 2019 14:30 - 15:10 PDT
Harbor Ballroom E (Track 2)

Open Hardware

Session Slides Included YES

14:30 PDT

POWERing up gem5 - Basavaraj Talawar, National Institute of Technology Karnataka, Surathkal

For the first time ever, full system POWER9 CPU model support has been added to the popular industry-academia standard gem5 architecture simulator. The implementation is done via a 3 year university project at the National Institute of Technology Karnataka, Surathkal, India and co-mentored by the Linux Technology Center at IBM, Bangalore. The functionality of the models have been fully verified by booting the OPAL firmware and the Linux kernel on gem5 on the POWER9 model.Development milestones include enabling 64-bit, little-endian support, privilege instructions and Radix MMU support, adding exception handling, and serial console support. This talk will detail the development journey, demo the Linux boot, and elaborate the future goals of this exciting project that the OpenPOWER partners from
industry and academia will benefit immensely from.

Speakers

Basavaraj Talawar

Assistant Professor, National Institute of Technology Karnataka, Surathkal

Basavaraj Talawar is an Assistant Professor at the CSE department at the National Institute of Technology Karnataka, India where he heads the the SPARK (Systems, Parallelization and Architecture Research at NITK) lab. He has a PhD from the Indian Institute of Science, Bangalore. His... Read More →

POWERing.up.gem5 OpenPOWERSummit 2019 pdf

Tuesday August 20, 2019 14:30 - 15:10 PDT
Harbor Ballroom D (Track 1)

Simulation CPUs

15:15 PDT

An Overview of the Self Boot Engine (SBE) in POWER9 base OpenPOWER Systems - Amit Tendolkar, IBM India Pvt. Ltd.*

This talk will provide an overview of the on-chip Self Boot Engine (SBE) hardware and firmware first shipped on POWER9 based systems. This talk will further give an understanding of how the SBE plays a key role in enabling several significant and futuristic architectural imperatives in OpenPOWER systems like hardware abstraction, secure boot, RAS and host offloads that enable key features. The talk will finally conclude with examples of system-level use-cases and flows that highlight the role and significance of the SBE as critical and futuristic architectural element in OpenPOWER systems, starting POWER9.

Speakers

Amit Tendolkar

Senior Engineer, IBM India Pvt. Ltd.

Amit completed his Master's of Science in Electronics and Instrumentation in May 2000. Amit has a total of 19 years of industry experience, 8 years in Real Time Embedded Multimedia Systems at Philips Electronics India Ltd. and 11 years in Enterprise Server Firmware at IBM India Pvt... Read More →

OpenPOWERNA2019 Tendolkar SBE pdf

Tuesday August 20, 2019 15:15 - 15:55 PDT
Harbor Ballroom E (Track 2)

CPUs

Session Slides Included YES

15:15 PDT

Counting the Stars - Jez Wain, Atos

There are 200 billion stars in the Milky Way. The European Space Agency's Gaia project has mapped about 1% of them. The stars now need to be classified in much the same way as we classify animals and plants. Atos working with the University of Geneva has investigated the use of machine learning to help with this classification problem. Jez's talk describes the Gaia project before presenting the approach taken to construct and optimise a neural network capable of classifying the different star types. The techniques used are applicable to many different supervised machine learning problems. The talk will conclude with the initial results and a look towards future work.

Speakers

Jez Wain

Atos

Jez started out working on real-time airborne fly-by-wire systems and then on automated manufacturing systems. He moved to France to join Group Bull (now Atos) to work on the AIX kernel in it's migration from a single to multi-processor support. He later became an infrastructure consultant... Read More →

Tuesday August 20, 2019 15:15 - 15:55 PDT
Harbor Ballroom D (Track 1)

HPC

15:15 PDT

State of IBM POWER Support on FreeBSD - Piotr Kubaj*

FreeBSD is currently the only open-source operating system other than Linux that can run on modern POWER platforms. Its advantages (ZFS file system, jails, and conservative approach) encourage some users to use it instead of Linux-based systems.
In my project, I build the whole package repository on a POWER9 server and fix software that fails to build. In this talk, I review the state of things on three currently supported POWER architectures - powerpc, powerpcspe and powerpc64. I also talk about the challenges encountered when porting software to FreeBSD on POWER and the current work in progress.

Speakers

Piotr Kubaj

sysadmin and FreeBSD developer

Piotr Kubaj is a sysadmin and FreeBSD developer from Poland with an interest *BSD operating systems, open-source firmware (coreboot and OpenPOWER) and IBM POWER architecture. He works on improving software support on FreeBSD on POWER architecture and maintains several ports.

powerpc64 pdf

Tuesday August 20, 2019 15:15 - 15:55 PDT
Harbor Ballroom F (Track 3)

Software Development Systems

Session Slides Included YES

15:55 PDT

Coffee Break

Tuesday August 20, 2019 15:55 - 16:25 PDT
Harbor Foyer

Event Experiences / Breaks / Meals

16:25 PDT

The Open Domain-Specific Architecture - Bapi Vinnakota, Netronome

Domain-specific accelerators (DSAs) are expected to be used in the data center and at the edge for a wide range of applications including machine learning, inferencing, video transcoding, database acceleration, blockchain and other applications. The Open Domain-Specific Architecture (ODSA) is a new incubation effort in the Open Compute Project.

The ODSA project aims to enable silicon vendors, systems vendors, data center operators and others to create high-performance DSAs at low development cost and time through the use of chiplet-based designs. In chiplet-based designs, an integrated product is implemented across multiple die, referred to as chiplets, but functions as though it is implemented on a monolithic IC. Chiplet-based designs require physical and logical interfaces between chiplets.

Specifically, the ODSA aims to do the following:
1. Define open physical and logical interfaces to allow chiplets from multiple vendors to interoperate. Unlike today’s closed systems, product designers can combine best-in-class chiplets from multiple vendors.
2. Create a marketplace of chiplets that support these interfaces. Create reference implementations, business practices and operational workflows to assemble DSAs with chiplets from the marketplace.

This talk will provide a technical introduction to the ODSA, recent technical results from the group and discuss the potential application of Open Power technologies, including the OMI.

Speakers

Bapi Vinnakota

Engineer, Netronome

Tuesday August 20, 2019 16:25 - 17:05 PDT
Harbor Ballroom E (Track 2)

Open Hardware

16:25 PDT

New Innovations in Power and Performance Management - Todd Rosedahl, IBM*

This talk will provide an overview of the existing power measurement, power management, and performance-boosting capabilities of the POWER architecture. Various levels of the hardware/firmware and software stack will be described. Then, it will focus on a new data collection method enabled by OpenBMC that provides a rich data set including performance metrics, temperatures, and power consumption. This new technology has been deployed at scale at ORNL with excellent results. Additionally, a newly emerging power/performance-optimizing application, called GEOPM, will be described.

Speakers

Todd Rosedahl

Chief Power/Thermal Engineer, IBM

Todd Rosedahl has worked for IBM for over 25 years in the area of power, thermal, and performance management. He has spoken at prior Open Power Summits, SC, ISC, and IBM Technical University. Currently Todd is a leader in the HPC PowerStack initiative where industry and academic experts... Read More →

Power Performance Overview Open Power Summit v20.0 pdf

Tuesday August 20, 2019 16:25 - 17:05 PDT
Harbor Ballroom D (Track 1)

Power Management HPC CPUs

Session Slides Included YES

16:25 PDT

Linux on POWER Driver Porting, Optimization, and Test - Brian King, IBM*

This session will provide an overview of what is involved in porting an I/O device driver to the POWER processor. It will cover both high-level architectural content as well as diving deep into the code. It will include an overview of various architectural differences between POWER and other architectures. Specific examples of issues discovered in the past will be reviewed. The scope of this session will encompass functional porting, optimization, as well as test tools and methodologies.

Speakers

Brian King

Linux on Power I/O Chief Architect, IBM

Brian King is the Linux on Power I/O Chief Architect at IBM. He is responsible for Linux I/O enablement and strategy on the Power platform at IBM. He is the author of two Linux device drivers, and has contributed to many more. Brian is the chair of the OpenPOWER I/O Workgroup.

OpenPowerNA2019 Linux Driver Porting pdf

Tuesday August 20, 2019 16:25 - 17:05 PDT
Harbor Ballroom F (Track 3)

Software Development, Subtype1 | Systems, Subtype1

Session Slides Included YES

17:10 PDT

Overview of the SMC 1000 8x25G: OMI to DDR4 Memory Controller - Sanjay Goyal, Microchip*

Speakers

Sanjay Goyal

Architect, Data Center Solutions, Microchip

Sanjay Goyal is a senior architect in Microchip’s Data Center Solutions business unit. He has been working in the Storage industry for 20+ years and has been with Microchip for 13 years. His experience includes chip and solution architecture for SAS expanders and the recently... Read More →

Sanjay Goyal OpenPower Summit Technical Presentation pdf

Tuesday August 20, 2019 17:10 - 17:50 PDT
Harbor Ballroom E (Track 2)

Open Hardware

Session Slides Included YES

17:10 PDT

Linux on OpenPower Validation - Keeping Linux Bug Free - Sachin Sant & Nageswara Sastry, IBM*

Testing/QA is the most important activity in Software Development Life Cycle to ensure the Software and/or Hardware works as expected. This presentation will provide insights about Linux testing(at IBM) on OpenPower boxes, introduce various test suites used by IBM Linux on POWER functional test teams to validate Linux(avocado), KVM(tp-qemu, tp-libvirt), I/O(avocado) and OPAL(op-test). These test cases/test suites are available in open source projects and can be used by various OpenPower partners for Linux validation. This presentation will also share how OpenPower partners can use these tests and also contribute to it.

Speakers

Nageswara Sastry

Advisory Software Engineer, IBM India

Nageswara Sastry (nasastry@in.ibm.com) is an Advisory Software Engineer at Open Systems Development, IBM. He is passionate about new technologies and learning. He is having 17+ years of Test development experience in Linux and related areas.He has spoken at different conferences namely... Read More →

Sachin Sant

Senior Software Engineer, IBM

Sachin Sant is Linux on POWER Test Architect at Open Systems Development, IBM. He has 20+ years experience in Operating System Test and Development with 12+ years in Linux. In his current role he leads Linux on IBM POWER Test team. Additionally he is also involved in testing upstream... Read More →

OpenPOWER Summit Linux Validation pdf

Tuesday August 20, 2019 17:10 - 17:50 PDT
Harbor Ballroom F (Track 3)

Software Development Systems