Showing papers in "Ibm Journal of Research and Development in 2018"

Potential of quantum computing for drug discovery

Abstract: Quantum computing has rapidly advanced in recent years due to substantial development in both hardware and algorithms. These advances are carrying quantum computers closer to their impending commercial utility. Drug discovery is a promising area of application that will find a number of uses for these new machines. As a prominent example, quantum simulation will enable faster and more accurate characterizations of molecular systems than existing quantum chemistry methods. Furthermore, algorithmic developments in quantum machine learning offer interesting alternatives to classical machine learning techniques, which may also be useful for the biochemical efforts involved in early phases of drug discovery. Meanwhile, quantum hardware is scaling up rapidly into a regime where an exact simulation is difficult even using the world’s largest supercomputers. We review how these recent advances can shift the paradigm with which one thinks about drug discovery, focusing on both the promises and caveats associated with each development. In particular, we highlight how hybrid quantum-classical approaches to quantum simulation and quantum machine learning could yield substantial progress using noisy-intermediate scale quantum devices, whereas fault-tolerant, error-corrected quantum computers are still in their development phase.

Bayesian optimization for accelerated drug discovery

Abstract: The space of potential drug-like molecules is vast, precluding “random-walk”-like searches from achieving any reasonable effectiveness. Active search techniques have been increasing in popularity in recent years as a method for accelerating the discovery of novel pharmaceutical molecules. By providing an effective method for prioritizing molecules within the discovery process, the efficiency of the discovery process can be dramatically improved. In this paper, we describe the use of Bayesian optimization, a method for iterative optimization of black-box functions for achieving this end, balancing the exploitation of current knowledge acquired from data, with the acquisition of new knowledge about which little is known.

Scalable Molecular Dynamics with NAMD on the Summit System.

Abstract: NAMD (NAnoscale Molecular Dynamics) is a parallel molecular dynamics application that has been used to make breakthroughs in understanding the structure and dynamics of large biomolecular complexes, such as viruses like HIV and various types of influenza. State-of-the-art biomolecular simulations often require integration of billions of timesteps, computing all interatomic forces for each femtosecond timestep. Molecular dynamics simulation of large biomolecular systems and long-timescale biological phenomena requires tremendous computing power. NAMD harnesses the power of thousands of heterogeneous processors to meet this demand. In this paper, we present algorithm improvements and performance optimizations that enable NAMD to achieve high performance on the IBM Newell platform (with POWER9 processors and NVIDIA Volta V100 GPUs) which underpins the Oak Ridge National Laboratory's Summit and Lawrence Livermore National Laboratory's Sierra supercomputers. The Top-500 supercomputers June 2018 list shows Summit at the number one spot with 187 Petaflop/s peak performance and Sierra third with 119 Petaflop/s. Optimizations for NAMD on Summit include: data layout changes for GPU acceleration and CPU vectorization, improving GPU offload efficiency, increasing performance with PAMI support in Charm++, improving efficiency of FFT calculations, improving load balancing, enabling better CPU vectorization and cache performance, and providing an alternative thermostat through stochastic velocity rescaling. We also present performance scaling results on early Newell systems.

IBM POWER9 opens up a new era of acceleration enablement: OpenCAPI

Abstract: Open Coherent Accelerator Processor Interface (OpenCAPI) is a new industry-standard device interface that enables the development of host-agnostic devices that can coherently connect to any host platform that supports the OpenCAPI standard. This in turn allows such devices to coherently cache host memory to facilitate accelerator execution, perform direct memory access and atomics to host memory, send messages and interrupts to the host, and act as a host memory home agent. OpenCAPI utilizes high-frequency differential signaling technology while providing the high bandwidth and low latency needed by advanced accelerators. OpenCAPI encapsulates the serializing cache access and address translation constructs in high-speed host silicon technology to minimize overhead and design complexity in attached silicon such as field-programmable gate arrays and application-specific integrated circuits. Finally, OpenCAPI architecturally ties together transaction layer, link layer, and physical layer attributes to optimally align to high serializer/deserializer (SerDes) ratios and enable high-bandwidth, highly parallel exploitation of attached silicon.

Data quality challenges for person-generated health and wellness data

Abstract: Person-generated health data (PGHD) generated by wearable devices and smartphone applications are growing rapidly. There is increasing effort to employ advanced analytical methods to generate insights from these data in order to help people change their lifestyle and improve their health. PGHD—such as step counts, exercise logs, nutritional diaries, and sleep records—are often incomplete, inaccurate, and collected over too short a duration. Insufficient user engagement with wearable and mobile technologies, as well as lack of sensor validation, standardization of data collection, transparency of data processing assumptions, and accessibility to relevant data from consumer-grade sensors, also negatively affects data quality. The literature on data quality for PGHD is sparse and fragmented, providing little guidance to data analysts on how to assess and prioritize data quality concerns. In this paper, we summarize our experiences as data analysts working with PGHD, outline some of the challenges in using PGHD for insight generation, and discuss some established methods for addressing these challenges. We review the literature on PGHD data quality, identify the major stakeholders in the PGHD ecosystem, and apply an established data quality framework to present the most relevant data quality challenges for each stakeholder.

A homomorphic encryption-based system for securely managing personal health metrics data

Abstract: Hardware and software solutions for the collection of personal health information continue to evolve. The reliable gathering of personal health information, previously usually possible only in dedicated medical settings, has recently become possible through wearable specialized medical devices. Among other drawbacks, these devices usually do not store the data locally and offer, at best, limited basic data processing features and few advanced processing capabilities for the collected personal health data. In this paper, we describe an integrated personal health information system that allows secure storage and processing of medical data in the cloud by using a comprehensive homomorphic encryption model to preserve data privacy. The system collects the user data through a client application module, typically installed on the user's smartphone or smartwatch, and securely transports the data to the cloud backend powered by IBM Bluemix. The data are stored by the IBM Cloudant infrastructure, while the homomorphic processing of the encrypted data is performed using the Apache Spark service, which is also made available by the IBM Bluemix platform. The event-based handlers are triggered by the IBM OpenWhisk programming service. The initial prototype has been tested using a real-world use case, which is described.

Decomposition of complex movements into primitives for Parkinson's disease assessment

Abstract: Recent advances in technology present an important opportunity in medicine to augment episodic, expert-based observations of patients’ disease signs, obtained in the clinic, with continuous and sensitive measures using wearable and ambient sensors. In Parkinson's disease (PD), such technology-based objective measures have shown exciting potential for passively monitoring disease signs, their fluctuation, and their progression. We are developing a system to passively and continuously capture data from people with PD in their daily lives, and provide a real-time estimate of their motor functions, that is analogous to scores obtained during Part III of the human-administered Movement Disorder Society's Unified Parkinson's Disease assessment (MDS-UPDRS3). Our hypothesis is that complex human movements can be decomposed into movement primitives related to the performance of the MDS-UPDRS3 motor assessment. Toward this hypothesis, we developed a system for integrating and analyzing multiple streams of sensor data collected from volunteers executing the tasks based on the MDS-UPDRS3. In this paper, we show how we can leverage the data collected from MDS-UPDRS3 tasks to develop machine learning models that can identify movement primitives in activities of daily living.

An optimized blockchain solution for the IBM z14

Abstract: Blockchain is a technology that has been available for several years and has already had several generations. Multiple implementations exist. In this paper, we describe how one of them, the Hyperledger Fabric, is used today to demonstrate the wide usage of the technology for multiple markets beyond finance and cryptocurrency. The use cases highlight the relevance of security—due to the type of data and the distributed foundation of blockchain—and of system performance to enable the use cases in production. Security and performance are the base for the architecture of blockchain on the IBM LinuxONE systems. The Secure Service Container prevents tampering, and the cryptocard usage hardens the solution—and the performance optimization in the compiler, cryptoinstructions, and large caches provides the performance. These optimizations have been further extended with the IBM z14 beyond previous IBM Z generations. In this paper, we describe how a cluster topology is used to implement a high-availability service that scales dynamically from small networks to large ones using the hot plug capabilities of the IBM LinuxONE systems.

IBM POWER9 processor core

Abstract: The IBM POWER9 processor is the latest Reduced Instruction Set Computer microprocessor from IBM. POWER9 employs a new modular core microarchitecture to counter the technology trend of decreasing frequency and increasing power density from generation to generation. The new POWER9 design enables a family of processors optimized for a broad range of server applications. The new microarchitecture is closely coupled with a rich set of new instructions geared toward data-centric applications. In this paper, we describe the POWER9 core microarchitecture innovations, its new instructions and features, and the exploitation of this new design for computing in the cognitive era.

IBM POWER9 package technology and design

Abstract: The first-level package that contains the IBM POWER9 processor chip is designed to achieve the high computational performance needed for cognitive systems in a cost-effective design. The throughput data bandwidth of the POWER9 package for high-end scale-up systems is more than 1 TB/s, which is double the data bandwidth of the previous generation. This increase in bandwidth is achieved by introducing a dielectric with a loss tangent of 40% of the predecessor material, a C4 density increase of 15%, higher number of stacked vias to reduce jogging, and improved via pattern and placement to increase the frequency and density of signals. The cloud platform scale-out POWER9 package leverages the high-end and cognitive platform package attributes to maintain signal frequency while introducing novel chip-package-system co-design techniques. These design techniques were used to produce a well-balanced two-socket entry-level package with four build-up layers above and below the core, instead of six, resulting in a significant cost reduction from the previous generation while supporting the signal frequencies of POWER9. POWER9 systems are the first to offer 16-Gb/s PCIe Gen4 and 25.8-Gb/s open coherent accelerator processor interface that interconnect the processor to the I/O, networking, and accelerators required for systems in the cognitive computing era. In this paper, we present the material and wiring technology needed to achieve the signal performance up to 25.8 Gb/s per channel, the package physical attributes, and the chip-package-system co-design methodology to achieve the increased signal density, minimize the crosstalk, and maximize the frequency while reusing the package form factors of the previous generation, IBM POWER8.

Process technology for IBM 14-nm processor designs featuring silicon-on-insulator FinFETs

Abstract: A highly optimized silicon-on-insulator FinFET technology is utilized for the IBM processor designs in the 14-nm node. This process technology has a number of unique elements that enable these prod...

Functionality and performance of NVLink with IBM POWER9 processors

Abstract: Heterogeneous computer systems with multiple types of processing elements (PEs) are becoming a popular design to optimize performance and efficiency for a wide variety of applications. Each part of an application can be executed on the PE for which it is best suited. In heterogeneous systems, communication, efficient data movement, and memory sharing across PEs are critical to execute an application across the different PEs while incurring minimal overhead for communication and synchronization. The IBM POWER9 processor supports the NVIDIA NVLink interface, a high-performance interconnect with many such capabilities. In the IBM Power System AC922, IBM POWER9 processors directly connect to multiple NVIDIA GPUs using NVLink. In this paper, we highlight the important functional and performance capabilities of NVLink with the POWER9 processor. These include high bandwidth, hardware cache coherence, fine-grained data movement, and hardware support for atomic operations across all PEs of a compute node. We also present an analysis of how these performance and functional capabilities of POWER9 processors and NVLink are expected to have significant impacts on performance and programmability across a variety of important applications, such as machine learning and domains within high-performance computing.

Curating and integrating user-generated health data from multiple sources to support healthcare analytics

Abstract: As the volume and variety of healthcare-related data continue to grow, the analysis and use of this data will increasingly depend on the ability to appropriately collect, curate, and integrate disparate data from many different sources including user-generated health data. We describe our approach to, and highlight our experiences with, the development of a robust data curation process that supports healthcare analytics. The process consists of the following steps: collection, understanding, validation, cleaning, integration, enrichment, and storage. It has been successfully applied to the processing of a variety of data types including clinical data from electronic health records and observational studies, genomic data, microbiome data, self-reported data from surveys, and self-tracked data from wearables from more than 600 subjects. The curated data have been used to support a number of healthcare analytic applications, including descriptive analytics, data visualization, patient stratification, and predictive modeling.

The Pharmit Backend: A Computer Systems Approach to Enabling Interactive Online Drug Discovery.

Abstract: Pharmit (http://pharmit.csb.pitt.edu) is an open-source online resource that allows users to interactively search libraries of millions compounds as part of a structure-based drug discovery workflow. Here we describe the systems-level implementation decisions made in designing Pharmit that, when combined with novel sub-linear time search algorithms, allow it to screen millions of molecules in seconds. The key concepts are to maximize parallelism while minimizing intra-thread communication, optimize data layout for sequential processing, and efficiently manage memory allocation. We describe how these concepts are applied to the cheminformatic data inherent to Pharmit and discuss limitations and possible future directions.

IBM z14 design methodology enhancements in the 14-nm technology node

Abstract: In designing the IBM z14 microprocessor chipset, we discarded many of our previous assumptions and processes in favor of newer, more radical approaches. These new approaches were the result of learning from previous designs as well as this design's performance and schedule requirements. In this paper, we discuss some of the more significant changes to our methodology, including a dramatic departure from our very hierarchical integration design style to a flatter, more nimble approach. We also discuss improvements to our design's power management architecture for managing power-supply noise. Finally, we discuss changes to our simulation environment, targeted at increasing both the number of simulation cycles and simulation logic coverage.

Large-scale application of IBM Design Thinking and Agile development for IBM z14

Abstract: In this paper, we describe the application of IBM Design Thinking and Agile development for the IBM z14 project. We discuss the adoption of IBM Design Thinking practices and its impact on the development and client reception of IBM z14. Furthermore, we show how the design direction of the IBM z14 system was iteratively shaped by involving key members of client organizations (“sponsor users”) early in the development process. Next, we discuss how applying Agile practices in a hardware/firmware environment—with iterative development and test, playbacks, and continuous integration—allowed for faster course correction and delivery on a challenging schedule. This paper provides a case study for the adoption of IBM Design Thinking and Agile practices, which were originally designed for small team operations, and the challenge of applying them to a large-scale development organization such as the IBM team developing the z14 with more than 1,000 engineers.

IBM POWER9 circuit design and energy optimization for 14-nm technology

Abstract: Modern multicore microprocessors require attentive development to energy requirements when maximizing power-performance efficiency and ensuring reliable plus scalable functionality. IBM POWER9 relies on extensive modeling to identify representative workloads used when analyzing thermal design power and regulator design power against product requirements. Compounding benefits of circuit optimizations applied to the diverse subcomponents of the chip results in lower power cores, caches, and memory/IO interconnect. Specific dc- and ac-current analyses ensure proper definition of chip specifications for system voltage and current delivery. Finally, a systematic exploration of microbenchmarks on intermediate and final POWER9 hardware provides insight into processor core requirements while validating model accuracy.

Evaluating speech-based question–answer interactions for elder-care services

Abstract: Japan has the highest percentage of population deemed elderly in the world (i.e., people over the age of 65). The portion is predicted to reach 30% by 2025. An aging society poses various societal challenges including societal isolation. Mobile devices, such as smartphones and tablets, can play an important role in connecting elderly persons with family members and can provide cost-effective daily services and opportunities to join community activities. However, lack of experience with mobile devices often prevents elderly persons from adopting such services. It is thus necessary to assist the elderly by giving them answers to questions they may have about their devices and services, to enable large-scale deployment of mobile devices to the aging population. In this study, we evaluated a speech-based question-and-answer system that we designed for elderly novice users of mobile devices through a pilot study on an eldercare service platform with 1,011 elderly participants who were encouraged to use health-check services, video telephony, and other services on a tablet device. The participants could ask questions about the services, the application, and device by using speech whenever they wanted. The results suggest the feasibility of using speech-based interfaces as the main interaction medium for the elderly.

An interpretable health behavioral intervention policy for mobile device users.

Abstract: An increasing number of people use mobile devices to monitor their behavior, such as exercise, and record their health status, such as psychological stress. However, these devices rarely provide ongoing support to help users understand how their behavior contributes to changes in their health status. To address this challenge, we aim to develop an interpretable policy for physical activity recommendations that reduce a user's perceived psychological stress, over a given time horizon. We formulate this problem as a sequential decision-making problem and solve it using a new method that we refer to as threshold Q-learning (TQL). The advantage of the TQL method over traditional Q-learning is that it is "doubly robust" and interpretable. This interpretability is achieved by making model assumptions and incorporating threshold selection into the learning process. Our simulation results indicate that the TQL method performs better than the Q-learning method given model misspecification. Our analyses are performed on data collected from 79 healthy adults over a 7 week period, where the data comprise physical activity patterns collected from mobile devices and self-assessed stress levels of the users. This work serves as a first step toward a computational health coaching solution for mobile device users.

Design of the IBM z14 microprocessor

Abstract: The latest-generation IBM Z processor provides enhanced performance and compute capacity compared to its IBM z13 predecessor. This paper describes some of the major improvements that include an additional perceptron branch predictor, a completely redesigned translation engine that is tightly integrated into the core pipeline, and an integrated level-1 cache directory and translation lookaside buffer design. Outside of the central processing unit (CPU), the cache sizes have increased on each cache level, and each processor chip now contains 10 CPUs. The system topology has been optimized to improve cache transfer latencies for workloads spanning multiple processor chips. The bus interfaces between the chips have been redesigned to improve peak bus traffic handling. In combination, these enhancements provide significant performance improvements in traditional data-serving workloads, as well as in virtualized Linux environments running database, analytic, and cognitive workloads.

IBM POWER9 memory architectures for optimized systems

Abstract: The IBM POWER9 processor chipset provides a variety of system memory architecture interfaces to enable highly differentiated system offerings: a high bandwidth, high capacity, highly reliable, buffered architecture; a compute-density-optimized direct DDR attach architecture; heterogeneous integration of graphics processing unit memory into the host system memory; and an agnostic, flexibly attached SCM architecture. In this paper, we explore these architectures and the targeted optimizations they provide for various classes of workloads. We also explore the development synergies and semiconductor physical design tradeoffs associated with the varying implementations, and finally, we describe several hypothetical systems that could be constructed by utilizing these memory architectures.

IBM POWER9 system software

Abstract: The IBM POWER9 architecture offers a substantial set of novel and performance-improvement features that are made available to both scale-up and scale-out applications via system software. These features provide significant performance improvements for cognitive, cloud, and virtualization workloads, many of which use dynamic scripting languages. In this paper, we describe some of the key features.

Hybrid natural language processing for high-performance patent and literature mining in IBM Watson for Drug Discovery

Abstract: IBM Watson for Drug Discovery (WDD) is a cognitive computing software platform for early stage pharmaceutical research. WDD extracts and cross-references life sciences information from very large-scale structured and unstructured data, identifying connections and correlations in an unbiased manner, and enabling more informed decision making through explainable analytics and scientific visualizations. This paper describes in detail the high-throughput natural language processing system implemented in WDD. This system enables a new WDD release every three weeks, comprising the latest publications as part of a continually growing corpus of over 30 million scientific and intellectual property documents, each reprocessed using the latest annotators and structured reference data to extract a set of domain-relevant entity and relationship concepts. The hybrid approach to natural language processing in WDD incorporates model- and rule-based techniques utilized in concert for high-performance named entity recognition, and a similar ensemble approach to named entity resolution tasks, culminating in semantic relationship extraction. Statistics on full-scale annotation results and example use cases are also provided.

IBM POWER9 systems designed for commercial, cognitive, and cloud

Abstract: A new era of computing has emerged that focuses on actionable insights and predictive analytics with machine learning and deep learning algorithms. This is referred to as the cognitive era of computing. Servers designed for cognitive computing require a much different architecture than a traditional commercial server designed for database transactional processing and process automation. For example, graphics processing unit acceleration and high-bandwidth I/O for scalability are some of the key requirements for cognitive computing. Another different set of requirements is driven by servers designed for cloud infrastructure. The requirements for a cloud server place an emphasis on the total cost of ownership, total cost of acquisition, as well as compute density and server management. In this paper, we describe the family of IBM POWER9 servers that have been designed to meet the differing requirements for the cognitive, commercial, and cloud market spaces. We describe how each server in the family has been optimized for one (or more) of these workloads by implementing different combinations of POWER9 module packages, memory subsystems, internal storage subsystems, system management, and different levels of reliability, accessibility, and serviceability.

New database compression assists in the IBM z14 processor

Abstract: In this paper, we describe how the IBM z14 processor, together with Db2 for z/OS Version 12, can improve data compression rates and thus reduce data storage requirements and cost for large databases. The new processor improves on the compression hardware accelerator available in earlier IBM Z generations by adding new hardware algorithms that increase the compression ratio and extend the applicability to additional data structures in Db2 for z/OS databases. A new entropy coding step employed after Ziv–Lempel compression reduces the size of data compressed with the prior algorithms by 30% on average. Moreover, the new order-preserving compression algorithm enables database index compression, reducing index sizes by roughly 30%. This results in an overall improvement of 30% of the database size for many applications, with associated benefits in storage requirements, input/output bandwidth, and buffer pool efficiency.

IBM POWER9 processor and system features for computing in the cognitive era

Abstract: IBM POWER9 is a family of processor chips designed to serve a diverse set of workloads. New features have been added to POWER9 to address emerging workloads such as cognitive and artificial intelligence applications. POWER9 also further enhances features introduced in IBM POWER8 for big data and cloud applications. Distinct chips using common intellectual property building blocks are provided to enable enterprise applications requiring large symmetric multiprocessor servers with large memory footprints, as well as one to two socket industry form-factor servers. In this paper, we describe new POWER9 features for both system types. Several highly differentiated new features are described in other papers in this issue of the IBM Journal, and they provide a more in-depth description of their unique design aspects.

Performance innovations in the IBM z14 platform

Abstract: The IBM z14 platform brings together numerous innovations that collectively provide considerable performance gains over prior designs for the full spectrum of modern workloads being run on today's enterprise systems. These workloads range from traditional online transaction processing and batch workloads to Linux-based and cloud-based workloads, leveraging advances in analytics, mobile transaction processing, social data mining, cognitive applications, security (blockchain), and other areas. The performance gains in the z14 platform required advancements in the processor hardware microarchitecture, physical design, and instruction-set architecture, as well as in the firmware and hypervisor stacks. The result is a comprehensive platform solution that enables efficient multiprocessor scaling for large single-image partitions and mixed-workload multipartition environments within a single platform, while remaining responsive to the “bursty” demands of batch and transaction-based workloads. The performance gains present in the z14 are a direct result of the collaborative efforts between diverse teams within IBM working together to produce advancements in hardware, firmware, and software that outpace the gains of the individual contributions. In this paper, we discuss these results with a focus on the client value delivered through system performance and workload responsiveness.

Message from the General Manager, IBM Z

Abstract: Trust is the foundation of digital relationships. Consumer, client, and partner expectations have grown. They demand security, transparency, and greater value in every interaction and transaction as this new currency is exchanged. Data is the new source of business value and differentiation, and data protection is paramount for trust across the ecosystem. Because digital experiences are delivered via the cloud, businesses must be open and connected in order to drive innovation at speed. They must accelerate development and delivery of secure, scalable services with new economic models.