Hyperion Research forecasts that the worldwide HPC broader market (servers, storage, software, and services) will expand at a 6.4% CAGR to exceed $40B in 2026. HPC servers compose about half of the broader market revenues and are expected to grow at a similar rate (6.9% CAGR) to exceed $20B in 2026. Storage is forecasted to show the highest growth (8.7% CAGR).

The application software landscape is quickly evolving along with HPC workloads. Independent software vendor (ISV) applications, as opposed to open-source or home-grown, have traditionally been considered the gold standard as the source for many HPC applications and were frequently cited as HPC users' top applications. While ISV revenues continue to rise, open-source application use is growing as well. Although HPC users can be reluctant to change the applications used at their site, developments such as the onset of AI, new types of hardware, and the cloud have spurred many users to explore new applications they may not have considered otherwise. Overall, the HPC application software landscape is rapidly developing in tandem with new HPC infrastructure and use cases.

The Israel Innovation Authority and the Israel Ministry of Defense (IMOD) recently announced a $62 million USD collaborative effort to establish a domestic quantum computing (QC) infrastructure. The two-pronged approach calls for the Israel Innovation Authority, part of Israel's Ministry of Economy charged with fostering domestic industrial R&D, to focus on developing QC algorithms, applications, and a related software stack to support both on-premises or cloud QC access models. For its part, the IMOD will stand up a national center to develop a complete quantum computer including a quantum processor expected to consist of 30-40 qubits, quantum control capabilities, and I/O interfacing hardware.

Researchers at the US-based Lawrence Berkeley National Lab (LBNL) recently reported a new approach to error mitigation in a quantum computer (QC) that targets error-producing noise, a ubiquitous problem that can severely limit the performance and utility of existing and near-future quantum computers. The method developed at LBNL consists of taking an initial noisy target circuit and constructing an analogous estimation circuit that is configured specifically for accurate noise characterization. The information gathered from running the estimation circuit is then applied to correct the noise in the original target circuit.

Seeking to establish a die-to-die interconnect standard and foster an open chiplet ecosystem, a strong collection of major chip makers and users recently announced the formation of the UCIe (Universal Chiplet Interconnect Express) industry consortium. The consortium has published version 1.0 of the UCIe specification, covering the die-to-die I/O physical layer, die-to-die protocols, and software stack. Promoter members of the consortium are Advanced Semiconductor Engineering, Inc. (ASE), AMD, Arm, Google Cloud, Intel Corporation, Meta, Microsoft Corporation, Qualcomm Incorporated, Samsung, and Taiwan Semiconductor Manufacturing Company (TSMC)

Hyperion Research has been following the rise of Arm processors in HPC for the past few years, tracking the development of processors as well as the evolution of the HPC software ecosystem to take advantage of Arm. Arm recently came into the spotlight with the acceptance of the Fugaku machine at Riken, which claimed the top spot on the Top500. Based on Fujitsu's Arm-based A64fx processor, Fugaku is a powerful and effective HPC system, rising to the top of many major HPC and AI benchmarks. Hyperion Research believes that the Fugaku machine is part of a wider growth of Arm adoption in HPC.

On February 8, 2022, the European Commission formally proposed what's commonly referred to as the European Chips Act. The legislation plans to build on Europe's strengths and address weaknesses to develop a thriving domestic semiconductor ecosystem and resilient supply chain, while setting measures to anticipate and respond to future supply chain disruptions. In the short term, the Act seeks to bolster EU capabilities to anticipate future chips crises, strengthen manufacturing activities in the EU, and support scale-up and innovation across the whole value chain. In the mid- to long-term, it seeks to reinforce Europe's technological leadership while developing mechanisms to support transfer of knowledge from the lab to the fab and position Europe as a technology leader in innovative downstream markets.

At the conclusion of 2021, researchers at UNC Charlotte and Duke University Medical Center published results of work done to use transfer learning methods to generate fluence maps for radiation therapy, aimed at providing medical professionals with more capability and information in fighting adrenal cancers. The technique uses a deep transfer learning model trained on a much larger dataset that can be applied to a smaller data set for a specific application.

• The initial model was trained on pancreas treatment plans, then retuned and applied to a smaller set of data points on adrenal cancers. The output of the model generates a fluence map for specific IMRT beam-based treatments for adrenal cancers.
• According to the researchers, this approach is meant to supplement but not replace human expertise in the field and is reliant on human expertise to finetune and improve the AI model.

Storage and interconnects continue to be important elements of HPC system architecture and are expected to take on even greater significance with increasing demanding and diverse requirements driven by both traditional compute-intensive HPC mod/sim workloads and data-intensive AI workloads. The significance is reflected in recent market data and near-term forecasts, technology adoption and utilization trends, industry announcements in the second half of 2021, and future storage and related technology research direction of Hyperion Research.

In an announcement made in December, 2021, Verusen, a startup specializing in leveraging AI resources to support global supply chains, detailed a recently formed partnership with Machine Compare, a supplier of one of the world's largest databases for machinery and leading B2B marketplace for buyers and sellers of industrial spare parts. The partnership is aimed at enhancing the customer experience, limiting risk, reducing waste, and helping companies conduct materials management and commerce in a new and efficient way. Verusen founder and CEO Paul Noble explains the partnership is targeted to resolve a painful and wasteful process and will ultimately allow manufacturers to realize a whole new level of sustainability. For his part, Machine Compare CEO Ben Findlay is looking for a reduction in downtime, stockouts, and costs. Furthermore, the burdens lifted by the Verusen AI capabilities are targeted to reduce the amount of manpower committed to time-consuming, reactive tasks, allowing for a more proactive and long-term management of goals.

Open source and standards have long been a hallmark of the HPC community. Each encourages broad industry participation, wide-ranging collaboration, and a thriving ecosystem with a common goal of robust, interoperable solutions. Key to the success of standards and open source is a stable, neutral governance and support structure. Two new unrelated industry efforts (Apptainer evolving under the auspices of the Open Linux Foundation and DMTCP establishing a public/private open-source partnership) have been embraced by the HPC community to strengthen the standards and open source activities of the HPC ecosystem.

Results from the most recent Hyperion Research Multi-Client Study (MCS) show continued evolution within HPC file system adoption. The 2021 HPC Multi-Client Study: Trends and Forecasts in HPC Storage and Interconnects report details the storage and interconnect-related responses and analysis associated with the 252 HPC storage systems deployed at 141 global HPC respondent sites representing 2,006 HPC systems. The survey sample had a total of 338 file systems. This Special Study examines file system utilization.

HPC has long been recognized as indispensable for advancing scientific research, providing more timely and accurate weather forecasting, building better products, and enabling broader adoption of technical computing for AI-driven training and inference. At the same time, HPC has also been recognized as being extremely complex to set up, operate, and maintain. According to recent Hyperion Research studies, over half of the overall HPC market identified the lack of staffing and ease-of-use related issues as barriers for them to acquire additional HPC-based solutions. NVIDIA, a leading HPC hardware component supplier, has announced that it is addressing this issue by acquiring Bright Computing, a leader in software for managing high-performance computing systems used by more than 700 organizations worldwide.

At the recent CES 2022 tech event, Japanese corporation OMRON, the world’s leading manufacturer and distributor of personal heart health products and other medical devices, highlighted (and later announced on their website) the activities of their partnership with Kyoto University to develop an AI[1]powered platform that uses remotely gathered patient data to predict cardiovascular diseases at an earlier stage than current averages. Kyoto University is closely tied to the identity of Japan's government and considered Japan's leading research university. They operate a Top500 HPC system on-site in addition to conducting research on the cutting edge Fugaku supercomputer. This new program, part of an ongoing partnership between the two organizations, seeks to explore the use of AI to analyze blood pressure metrics for early detection of cardiovascular diseases faster and with greater accuracy allowing for treatment courses to be changed or taken more quickly.

Understanding the HPC market by application segment continues to be essential as many key behaviors, such as spending habits and emerging technology adoption, are closely related to specific application areas. Verticals, applications, software, and middleware are all inextricably linked and key to accurately characterizing HPC site workloads. Important developments in these areas are explored in the 2021 iteration of Hyperion Research's annual HPC Multi-Client Study: Vertical/Application Workload Areas and Technical Computing System Software and Middleware report. Key findings from the report are summarized in this document.

The HPC User Forum was established in 1999 to promote the health of the global HPC industry and address issues of common concern to users. In September 2021, the 77th HPC User Forum took place virtually. This update summarizes a presentation from that virtual conference given by Leonardo Flores Añover, Senior Expert at the HPC & Quantum Technology Unit at the European Commission. He provided insights and updates on the current and future developments of HPC in the European Union and the EuroHPC joint undertaking as well as insights into their new HPC policy environment, the state of play in the EU, and drivers of the Union's HPC strategy.

Hyperion Research recently released its updated annual 5-year forecast for the on-premises technical server market, also referred to as the HPC server market, Worldwide On-premises HPC Broader Market Forecast Updates, 2020-2025. While projecting a steady overall 8% CAGR over the 5-year period, the forecast includes a drop in projected on-premises technical server revenues between 2024 and 2025. Although not a common occurrence, there is precedence for such a pullback in the sector, and in this particular instance a confluence of three trends is contributing to this phenomenon.

The US Office of Science and Technology Policy recently released its second annual National Quantum Initiative (NQI) report, a supplement to the President's FY22 Budget Request that outlines the major US government quantum information science (QIS) research activities and related funding levels out to FY 2022. As seen in Figure 1, the proposed FY2022 budget, which is targeted for about $880 million, calls for an increase of nearly 11% from the previous year. Roughly half of the funding is to come from the NQI and the other half from base agency-specific QIS R&D budgets. The figure represents the sum of Federal budgets for U.S. QIS R&D efforts in over a dozen agencies including NIST, NSF, DOE, NASA, DOD, and DHS, and it also aggregates several QIS subtopics such as computing, networking, sensing, fundamental science, and end quantum-related use cases

As AI and Big Data/HPDA applications continue to grow in importance and focus for HPC data centers and users worldwide, the demand for dedicated HPDA and AI systems are increasing at a rate nearly double that of the overall HPC market. By 2025, Hyperion Research expects roughly one-third of all system revenue to be dedicated to HPDA and AI-centric systems, in a category defined as Data[1]Centric HPC Systems. The portion of the data-centric market focused on AI applications is experiencing even higher growth, nearly 23% CAGR over the five year period, driven by the influx of AI workloads not only for new application spaces but also as a supplemental tool for the traditional simulations that have dominated the HPC space for decades. Some important HPDA workloads require analytics alone, while many benefit from combining established simulation and newer analytics methods, especially machine and deep learning.

AI and HPDA applications continue to grow in utilization and importance in the HPC space, altering the direction of system architecture and design as well as widening the scope of HPC and AI. HPC system architects now have to factor in a variety of processor technologies as well as interconnect, storage, and memory configurations to handle a more diverse set of workloads. Not only are HPC sites adopting emergent AI and HPDA workloads, but some are applying AI techniques to traditional modelling and simulation workloads to uncover new capabilities and solutions. This frequent pairing of simulation and analytics requires that HPC system designs be both compute and data friendly. As a result, recent designs are starting to reverse the emphasis on compute-centrism of past decades and establishing a better balance. Insights into the critical factors driving these and other trends are detailed in the 2021 iteration of Hyperion Research's annual Multi-Client Study (MCS) end users' report, AI and HPDA Usage and Future Technology Trends. Key findings from the report are summarized in this document.

Although technical servers and their associated compute and memory elements receive much of HPC infrastructure focus and attention, storage and interconnects are becoming increasingly prominent elements within the HPC ecosystem. This is reflected in the 2021 iteration of Hyperion Research's annual HPC Multi-Client Study: Trends and Forecasts in HPC Storage and Interconnects report. Key findings from the report are summarized in this document.

The HPC User Forum was established in 1999 to promote the health of the global HPC industry and address issues of common concern to users. In September 2021, the 77th HPC User Forum took place virtually. This update summarizes a presentation from that virtual conference given by Dave Turek, the Chief Technology Officer at Catalog, a new company focused on biological computing. Turek provided insights on biological methods of storing information and how his team is seeking to help shift storage and compute paradigms and create tools to manage the recent and continuing explosion of useful data.

Hyperion Research forecasts that the worldwide HPC broader market (servers, storage, software, and services) will expand at a 7.9% CAGR to nearly $40B in 2025. HPC servers comprise about half of broader market revenues and are expected to grow at a similar rate (8.0% CAGR) to reach nearly $20B in 2025. Storage is forecasted to show the highest growth (9.3% CAGR).

Hyperion Research forecasts that the worldwide HPC server market will reach $14.5B in 2021. Nearly half of the worldwide spend on HPC servers currently come from North America, primarily from US customers. Hyperion Research anticipates that the worldwide HPC technical server market will grow at an 8.0% CAGR over the forecast period (2020-2025) to reach nearly $20B in 2025.

Hyperion Research forecasts that the worldwide HPC technical server market to grow at 8.0% CAGR to 2025. Within the 13 verticals tracked, five are expected to exceed $1B in revenue in 2021, and three others are poised to pass that mark in the forecast period.

Users continue to run an increasing amount of their HPC workloads in the cloud. Recent Hyperion Research studies show that while heretofore the HPC cloud spending was largely complementary and incremental to on-premises HPC spending, there is growing evidence that on-premises spending is being either delayed or foregone altogether in lieu of cloud spending. Insights into the critical factors driving the trend are detailed in the 2021 iteration of Hyperion Research's annual MCS end user study, Use of Public/External Clouds for HPC Workloads, Trends, and Drivers report. Key Findings from the report are summarized in this document.

Hyperion Research is projecting the HPC cloud market to exhibit a five-year CAGR of nearly 17% and reach $9.3 billion in end user spending in 2025. The key verticals driving much of the growth include: the bio-sciences and manufacturing sectors, as well as the economics/finance sector and the EDA sector, to name a few.

A recent survey revealed that HPC end users are looking to change the way they outfit their next HPCs, with many interested in exploring new types of processors, relying increasingly on GPUs and other accelerators to support computational gains, and preparing to increase their budgets to afford continued acquisition of more powerful HPCs. Insights into the critical factors driving these and other trends are detailed in the 2021 iteration of Hyperion Research's annual MCS end users' study, Processors, Coprocessors/Accelerators, and HPC budgets. Key Findings from the report are summarized in this document.

The August 2021 issue of the International Research Journal of Engineering and Technology (IRJET), a peer-reviewed research journal, included a paper based on the work of three researchers from India's St. Francis Institute of Technology (SFIT) summarizing their use of artificial intelligence (AI) and machine learning (ML) methods to help alleviate water shortages in India caused by population growth, urbanization, and climate change. Verlekar, Shah, and Kulkarni used a machine learning model to create a proactive scheme for managing local water resources, work that was prompted by a 2019 drought that impacted the Chennai area of India.

In September 2021, the 77th HPC User Forum took place virtually. This update summarizes a presentation from that virtual conference given by Dr. Sean Peisert, computer security research lead and staff scientist and Lawrence Berkeley National Laboratory. He provided insights on security in high-performance computing environments, specifically about how emerging technologies and research can advance cybersecurity as an enabling capability as opposed to an impedance to optimal performance or other barrier for users.

HPC datacenters have been improving their energy consumption and environmental sustainability profiles. Verne Global, a high-intensity-compute cloud colocation services provider, has emerged as a leader in delivering 100 percent renewable, global HPC cloud resources, powered by Iceland's natural hydroelectric and geothermal energy. Digital 9 Infrastructure plc (D9), a newly formed UK- based investment trust company that invests in a range of digital infrastructure assets, has acquired Verne Global in a deal valued at approximately £231 million.

During the recent conference held by the Special Interest Group on Knowledge Discovery and Data Mining held in Singapore, three international public science policy advocacy groups presented a guide, Using Artificial Intelligence to Support Healthcare Decisions, aimed at empowering and educating the public on the growing use of AI platforms in the healthcare decision-making process. The guide covers explanations of common applications of artificial intelligence platforms in healthcare and, more importantly, outlines specific questions one can pose to cut to the core of the efficacy and reliability of an AI platform in those applications.

CATALOG recently announced major advancements in developing both storage and computers based on synthetic DNA as the fundamental building block. Continuous innovations in high performance computing (HPC) architectures are required by the ever-increasing challenges that scientists, researchers, and data analysts are expected to overcome. Traditional architectures are evolving as best they can to address each of the above areas. At some point, however, investments in traditional architectures will reach diminishing returns for certain workloads and use cases, driven by some economic reality. One promising non-traditional approach being actively pursued to address these complex challenges is the use and application of synthetic DNA for both storage and computational applications.

Recently, the Barcelona Supercomputing Center (BSC) trained the first large artificial intelligence (AI) model designed to understand, speak, and write in the Spanish language. The system, named MarIA, was trained on the MareNostrum supercomputer at the BSC, leveraging 59 TBs of language data from the Biblioteca Nacional de España, one of the world's largest public libraries. The model is said to be an expert in both writing and understanding the Spanish language and is free to use by any developer, company, or entity. The system has a wide variety of potential applications including summarization applications, chatbots, smart searches, translation engines, and automatic subtitling chatbots.

The HPC market is a dynamic environment from just about every perspective. Whether it be technology, business engagements and partnerships, or datacenter upgrade plans, there is rarely a dull moment. The first half of 2021 was no exception. This report provides HPC storage-related highlights and analysis of notable activity that occurred within the global HPC community over the past six months across several fronts including:

Using biological building blocks in place of traditional materials to assemble computers has been a research topic for many years, but recently the first potential commercial use cases have begun to emerge, centered on storage for large data sets. The DNA Storage Alliance, created to promote a storage ecosystem based on synthesized DNA strands, recently shared their aspirations for the emerging technology that offers significant promise in durability, simplicity, cost, and density over traditional magnetic counterparts. The initial goals of the alliance are to educate the public and raise awareness about DNA-based storage. Further out, the alliance may pursue the creation of specifications and standards, such as encoding, physical interfaces, retention, and file systems, to ensure that DNA-based solutions complement existing storage hierarchies. The alliance notes that expectations for the growth rate of current storage mechanisms cannot keep pace with the rising demand for data storage, particularly where growing data retention and related data mining efforts are driving the need to save increasingly larger data sets for longer periods of time. Such requirements are well suited to DNA-based archive storage characteristics in applications including digital content creation, robotics, smart cities, autonomous vehicles, healthcare, astronomy, and climate science.

Late last month, US Department of Defense (DOD) leadership explored the potential to inculcate artificial intelligence (AI) processes into its overall military operations, signaling a fundamental change in how information and data are used to increase the decision space for leaders both in military and civilian domains. Delivered during the third and most recent iteration of the Global Information Dominance Experiment (GIDE 3), which included representatives from all 11 combatant commands, NORTHCOM Commander Gen. Glen D. VanHerck's remarks on AI were aimed at progressing the ability to maintain domain awareness, achieve information dominance, and provide decision superiority in both competition and crisis.

In May 2021, the 76th HPC User Forum took place virtually. This update summarizes a presentation from that virtual conference given by Richard Lawrence, IT Fellow for Supercomputing at the Met Office, the UK’s national meteorological service. He provided updates on their current activities, challenges, and goals, as well as gave insights into their procurement process and what the community can expect from the Met Office in the upcoming years.

Building on its Budget 2021 commitment to launch a seven year $360 million National Quantum Strategy, the Government of Canada recently stood up a public survey website to gather insights on how to implement the plan. The survey covers a comprehensive set of issues including best practices on academic, government and private sector partnerships, talent and workforce challenges and opportunities, mechanisms to ensure equity, diversity, and inclusion, adoption of quantum technologies by Canadian entities, and identification of any significant gaps, barriers, or challenges that could hinder Canadian efforts to become a global leader in quantum technologies.

Ten leading German corporations recently stood up the Quantum Technology and Applications Consortium (QUTAC) to explore and promote the commercial application of quantum computing (QC) targeted for the German industrial base as a way to ensure German competitive advantage across a broad array of industries. The effort spans industrial sectors and founding members that include automotive manufacturing (Bosch, BMW, and Volkswagen), chemical and pharmaceutical (BASF, Boehringer Ingelheim, and Merck), insurance (Munich Re) and technology (Infineon, SAP, and Siemens). AIRBUS is participating as an external contributor.

The HPC User Forum was established in 1999 to promote the health of the global HPC industry and address issues of common concern to users. In May 2021, the 76th HPC User Forum took place virtually. This update summarizes a presentation from that virtual conference given by Tina Declerck, System Department Head at the National Energy Research Scientific Computing Center (NERSC), titled Perlmutter: A System for Simulation, AI, and Data. In addition to updates on NERSC and its role as the Mission HPC facility for the DOE Office of Science, Declerck provided information on the developments of the new Perlmutter system, its capabilities, structure, and future.

Two weeks ago, the US Department of Defense officially canceled its Joint Enterprise Defense Infrastructure (JEDI) cloud solicitation and contract, ending a long period of uncertainty and controversy. Originally, the contract, which designated $10 billion to support cloud computing capabilities for a variety of workloads and departments across the DoD, had been awarded to a single vendor, Microsoft Azure, in 2019. However, after appeals from other vendors, the process was reevaluated. Ultimately Microsoft was awarded the contract a second time. After nearly two years into the JEDI solicitation and award process, the DoD stated that their needs had evolved, and the original contract no longer aligned with the requirements of the department. A new solicitation was issued, the Joint Warfighter Cloud Capability (JWCC) contract, which indicated a plan to use multiple vendors to fulfill the needs of the contract. Currently, the DoD is seeking proposals from Microsoft and Amazon Web Services but will likely evaluate other qualified U.S. based CSPs.

The HPC User Forum was established in 1999 to promote the health of the global HPC industry and address issues of common concern to users. In May 2021, the 76th HPC User Forum took place virtually. This update summarizes a presentation from that virtual conference given by Sanzio Bassini, Director of SuperCalculation, Applications & Innovation Dept. at the CINECA consortium. He is currently a member of the EuroHPC Infrastructure Advisory Board and Leader of the EuroHPC Italian Pre-Exascale Infrastructure Leonardo Project. Bassini provided important updates on CINECA’s family of supercomputers, their evolving use cases and applications, as well as their plans to expand capabilities and continue empowering research scientists in Europe and around the world.

Several years ago, anecdotal evidence led Hyperion Research to compile a list of applications that promised to be the most economically important HPC-enabled AI use cases. Rather than simply drawing attention as interesting one-off examples, these applications had emerged as repetitive AI workloads that vendors could begin to pursue as emerging market segments. Hyperion Research's recently completed multi-client study of the worldwide HPC market presented a direct opportunity to ask HPC user organizations whether they use or plan to use any of the economically important HPC-enabled AI applications.

The HPC User Forum was established in 1999 to promote the health of the global HPC industry and address issues of common concern to users. In May 2021, the 76th HPC User Forum took place virtually. This update summarizes a presentation from that virtual conference given by Michael Brown, Principal Engineer, Intel. Brown provided updates on Intel's progress on optimizing NAMD for Intel's new instruction set architecture, how it interacts with their most recent processor offerings, and how Intel is looking to the future of molecular dynamics.

The European Processor Initiative (EPI) recently announced that it had released its European Processors Accelerator Chip (EPAC1.0) test chip design to a semiconductor maker for fabrication. The chip consists of multiple accelerator options including a vector processing unit, a stencil and tensor accelerator, and a variable precision processor, along with supporting RISC-V cores and memory access hardware. The 25 mm2 chip will be manufactured by US-based semiconductor maker Global Foundries using their low-voltage, low-power, 22 nanometer, 22FDX process platform.

Japan's NTT, the fourth largest telecommunications company in the world, recently announced the formation of a cooperative organization of Japanese firms designed to promote Japan's technical position in quantum technologies and to help Japan complete globally with US and Chinese rivals in both quantum computing and quantum communications. The inaugural meeting of the group, held in late May 2021, was attended by 11 Japanese companies, including leading IT suppliers Fujitsu, Hitachi, NEC, and Toshiba as well as industrial partners including Toyota Motor, Mitsubishi Chemical, and the Mizuho Financial Group. More than 50 companies are ultimately expected to join the group.

The U.S. Senate passed the United States Innovation and Competition Act (USICA) on June 8th, a major step forward in providing support and financial investment in furthering the United States' competitive capabilities in technology, including semiconductor capabilities. The USICA is bipartisan legislation intended to give federal funding to key science and technology areas, including STEM research, technology transfer, semiconductor research and manufacturing, as well as NASA research activities. The bill also seeks to establish a framework for agencies including the NSF and DOE to collaborate in these areas, to help ensure U.S. leadership in science and technology.

MLCommons, an international artificial intelligence (AI) standards body formed in 2018, launched MLPerf Tiny, their first benchmark targeted at the inference capabilities of edge and embedded devices, or what they call "intelligence in everyday devices". The new benchmark is now part of the overall MLPerf benchmark suite, which measures AI training and inference performance on a wide variety of workloads, including natural language processing and image recognition. The benchmark covers four machine learning (ML) tasks focused on camera and microphone sensors as inputs: keyword spotting, visual wake words, tiny image classification, and anomaly detection. Some important use cases include smart home security, virtual assistants, and predictive maintenance.

The HPC User Forum was established in 1999 to promote the health of the global HPC industry and address issues of common concern to users. In May 2021, the 76th HPC User Forum took place virtually. This update summarizes a presentation from that virtual conference given by Bill Mannel, Vice President and General Manager of HPC, HPE, and Justin Hotard, Senior VP and General Manager of HPC & Mission Critical Solutions (MCS), HPE, as an update on some of the important projects going on at Hewlett Packard Enterprise (HPE). Mannel and Hotard had three major topics for this update: ongoing integration with Cray in the wake of the 2019 acquisition, the recent leadership changes at HPE, and the role HPE plays in the emerging exascale environment.

The HPC User Forum was established in 1999 to promote the health of the global HPC industry and address issues of common concern to users. In May 2021, the 76th HPC User Forum took place virtually. This update summarizes a presentation from that virtual conference given by Chuck Gilbert, the Technical Director of Solutions Architecture at AMD. He gave updates on the most recent AMD CPU offerings, developments in heterogeneous architecture, and AMD's role in the future of high-performance computing. To Gilbert and the leadership at AMD, high performance computing is at the center of today's world and has improved every aspect of our lives. He explained that the market is now in a high-performance computing megacycle driven by the vast demands of cloud computing, digital transformation, 5G, and AI.

The HPC cloud market is continuing to rise as new users begin to adopt HPC cloud resources and existing users increase their spending and utilization of the cloud for HPC applications. Adoption and utilization of the cloud for HPC has varied based not only on the type of user but also on geography and the sizes of systems HPC users deploy on-premises.

Japan's RIKEN and Fujitsu recently announced the opening of the RIKEN RQC-Fujitsu Collaboration Center to support joint research and development of foundational technologies necessary to the development of quantum computers (QC), with a near-term goal of producing a 1000 qubit superconducting quantum computer. RIKEN has world-class skills in leading-edge scientific research including a range of quantum technologies while Fujitsu has demonstrated long-standing expertise in developing and fielding some of the highest performing advanced computing systems in the world. Such a combination could yield significant technology gains in the sector and open new areas of quantum computing applications and end uses. If able to successfully transition the results of this collaboration into commercial offerings, Fujitsu stands to become a strong competitor to leading commercial QC developers in the US and EU.

As workloads become more diverse and traditional enterprise datacenters are increasingly adopting HPC and HPC-enabled AI workloads, middleware is receiving more attention as part of the overall solution. HPC users desire a more intuitive experience and want to deal with fewer vendors. Middleware vendors want to expand their product portfolios and increase the markets they serve. Expanding product portfolios and increasing market reach can be achieved largely through either investment in current product lines or via strategic acquisitions. A recent example of the latter is Altair's acquisitions of Univa and Ellexus.

RIKEN's Fugaku, currently the fastest high performance computer (HPC) in the world, recently passed a critical development milestone by completing its initial testing phase. As a result, the system will become available for general shared use on March 9, 2021. The Fugaku supercomputer has been in development since 2014, and in addition to its world-leading computational performance capabilities, its overall design was targeted with an application-first philosophy to tackle some of the world's most pressing challenges including medicine, pharmacology, disaster prediction and prevention, environmental sustainability and energy. Fugaku is a joint development effort between RIKEN, Japan's largest HPC research organization, and Fujitsu, one of Japan’s largest commercial suppliers of high[1]end computing systems.

As HPC workloads are diversifying, processor needs are evolving and HPC users are increasingly demanding greater processor variety. The processor market is also growing overall as HPC users are incorporating more processors, and a greater range of processors, into their computer architectures. This combination is generating opportunities for companies able to provide new and better performing processor and coprocessor solutions. In response, new players are emerging, and many existing players are pivoting to provide innovative products. Meanwhile, some major HPC companies are leveraging acquisitions, mergers, and partnerships to assimilate new product offerings.

As the world's two largest economies and recognized world-class science and technology (S&T) developers, China and the United States are aware of how important technological prowess has become in projecting geopolitical leadership from an economic and national security perspective, and each place significant weight on national S&T polices to help ensure those ends. Recently, the leadership of both China and the United States had occasion to make public their long[1]term strategic S&T visions:

• China's President Xi Jinping highlighted that China is seeking to replace the US as the acknowledged global leader in S&T capability, and Beijing will likely continue its on-going efforts to transform its overall S&T infrastructure from that of a fast follow nation to one of global leadership while reducing its dependency on foreign S&T sources.
• US President Joe Biden emphasized that although the US has a long legacy of world-class technological prowess, emphasis should shift to better position S&T progress to help support key societal imperatives including public health, climate change, and economic inclusiveness. In additional, President Biden charged policymakers with fostering renewed support for R&D funding that not only ensures US S&T primacy going forward but that also fosters the wide base of existing and potential new US commercial entities to quickly turn research breakthroughs into innovative products.

Now more than ever, the key elements needed for high performance computing (HPC) performance leadership are a keen understanding of the system requirements to optimally support a growing array of heterogenous workloads and applications combined with the knowledge and creativity to realize the most appropriate hardware and software designs. The El Capitan machine being jointly developed by HPE and LLNL for the DoE NNSA ASC program provides two strong examples of exascale-driven architectural innovation: the Rabbit near-compute storage element and the Flux resource manager.

HPC in the cloud is augmenting many HPC users' access to resources today as utilization of cloud grows at an aggressive rate, causing HPC centers to create plans for how to effectively use public clouds as part of their approach to running HPC workloads. This includes future system planning as well as in-house personnel decisions for handling data center management. Many HPC users and datacenter managers are investigating the potential tradeoffs, benefits, and downsides of on-premise deployments and cloud utilization.

The French Government recently announced a national program worth over 1.8 billion euros (US $2.18 billion) over five years targeted to propel France into a global leadership position in a wide range of quantum technologies, including quantum-based computers, sensors, post-quantum cryptography, and communications. Specific quantum computing budgets are shared across near-term quantum simulator and NISQ system development as well as longer-term universal error corrected quantum systems. Albeit a strong comprehensive plan, France is one of the last major nations to come out with a wide-ranging national plan to spur innovation in quantum technologies and could be in catch-up mode for at least the next few years.

As more HPC users look to leverage the cloud to augment their on-prem resources or run cloud-natively, containers have emerged as a viable portable method for deploying their workloads more easily, consistently, and repeatedly anywhere (on-prem, in the cloud, or hybrid), decoupled from any one vendor's HPC system or single CSP's framework. Containers also aim to address requirements demanded by both HPC and enterprise IT datacenters (performance, automation, ease of application deployment, security, reliability, and scalability), an important attribute as enterprise IT datacenters increasingly adopt HPC infrastructure to implement their HPDA and AI workflows.

Covid-19 has impacted the worldwide HPC ecosystem in a variety of manners, leading to unpredictable changes in spending trends on systems, production, and installation of those systems, as well as changes in the sources of users' compute resources. Hyperion Research conducts surveys throughout the year to ask buyers for their expected short-term and long-term HPC spending changes in order to keep a tab on the ebbs and flows of the market. This Special Analysis explores changes in short-term budgets and expected spending. Due to the ever-changing situation of the world, these surveys have become critical in understanding, and predicting, the short-term spending trends of HPC buyers.

In late January of 2021, Origin Quantum (OriginQ), China's first domestic quantum computing (QC) start up, successfully raised US$15.4 million in a Series A round needed to continue development of its full-stack QC system capability. Its current QC system, the Wu Yuan, incorporates OriginQ's inhouse developed quantum processor, the KF C6-130, that has six superconducting qubits. Despite such examples of progress, continued advancement by the small number of commercial Chinese QC developers is not assured. China's significant and broad base of government funding for QC research and its resulting world-class capabilities primarily resides within Chinese government facilities. Improved mechanisms for transferring or independently fostering those same skills within the nascent Chinese QC commercial sector may be needed if Chinese commercial QC vendors are to close the technology and product development gap with leading-edge US and European commercial QC counterparts.

The European Commission (EC) has announced the Digital Europe Program, a new €7.5 billion effort designed to accelerate supercomputing, artificial intelligence, cybersecurity, and advanced digital skills, across the economy and society. Its overall goal is to help EU member states increase their capabilities in crucial areas of high-end computing and the relevant technologies to provide a foundation for driving Europe’s 21st century economy.

Although China represents more than one-third of the global market for semiconductors, the country relies heavily on foreign sources, especially U.S. chipmakers and foundries headquartered in other Asian countries, for a majority of devices critical to Chinese-made products ranging from smartphones to industrial machinery and supercomputers. For the past few decades, China sought to develop an indigenous capability in an effort to reduce its dependence on the foreign supply of components. Its most recent major effort, the “Made-in-China” semiconductor initiative launched in 2015, targeted supplying 40% of China's semiconductor demand with homegrown products by 2020 and 70% by 2025.

Since the dawn of the Supercomputer Era in the 1960s, a succession of base processors (CPUs) has attempted to satisfy HPC users' nearly insatiable demands for fast solution times on a wide spectrum of challenging problems, combined with affordability and energy efficiency. Chief among these entrants were vector processors, which ruled the HPC market through the 1990s, then RISC processors, and then x86-based processors, which began their steep ascendancy at that time and continue to dominate the market today, often in tandem with GPUs or other accelerators.

The recovery of the on-premise HPC server market continues to gain momentum, showing consistent revenue increases over the last two quarters, while easing some of the first quarter 2020 revenue losses that resulted from the economic fallout of the covid-19 virus. Specifically, the third quarter of 2020 exhibited a seasonal revenue growth of 3.4% to reach $3.4 billion in sales for the quarter, up 11.7% from the previous quarter and building on the 12.0 % quarter on quarter increase from 2Q2020.

HPC in the cloud has continued on a strong upward trajectory over the last few years, fueled by a concerted effort from CSPs to address the technical capabilities needed to better run HPC jobs in cloud environments. Many CSPs have bolstered their HPC strengths through hiring HPC experts to speak with HPC customers on application-specific issues, such as porting and running hard HPC jobs in the cloud. Further, CSPs are continuously increasing the platforms offered for HPC with the addition of offerings such as high performance processors, access to bare metal, a variety of accelerator options, high-performance interconnects, multiple storage offerings, as well as software packages valuable to HPC customers. These steps have resulted in dramatic advances in cloud adoption over the past two years, including a major growth year in 2019.

According to a recent Hyperion Research study of 135 US quantum computing (QC) technology developers across the academic, commercial, and government sectors, selected financial indicators in the QC commercial sector have yet to settle into any easily characterized profile. For example, Hyperion Research estimates that while more than one third of the reporting QC suppliers in the survey had revenues of five million dollars or more in 2019, almost half that supplied revenue figures had revenues of less than $500 thousand or no sales in 2019. Likewise, survey respondents indicated that the appropriateness of QC R&D funding differ greatly across the corporate, government, and VC spaces, citing a need for more government funding while expressing concern that current VC funding for the sector may be excessive. Finally, when asked about a potential quantum winter, QC suppliers were much more pessimistic than the base of current and potential QC users, with about half of QC suppliers believing a quantum winter is somewhat or highly likely in the next decade, twice that of counterpart QC end users.

The US Government's Office of Science and Technology Policy (OSTP) recently released a Federal strategic plan for a whole-of-nation approach to pioneering the future national advanced computing ecosystem that targets key application drivers and strategic objectives considered essential to US leadership, economic competitiveness, and national security. The plan specifically called out three critical application areas:

• • Extreme scale modeling and simulation
  • Data-intensive application workflows
  • End-to-end and real-time application workflows

The plan also arrays a number of key strategic objectives, including:

• • The development of an advanced ecosystem that spans government, academia, non-profits and industry
  • The support of foundational, applied and translational research
  • The expansion of a diverse, capable and flexible workforce

The strategic plan delineates US government agency roles and responsibilities and describes essential operational and coordination structures necessary to support and implement its objectives.

Hyperion Research has followed the growth of accelerators in HPC for more than 10 years. Accelerators had an initially slow HPC adoption curve but have recently experienced amplified growth. Within the HPC accelerator market, GPUs, particularly those from Nvidia, have dominated. Although originally designed as gaming processors for graphics rendering, GPUs were found to be well suited to a range of HPC applications, plus critical underlying AI functions, such as matrix multiplication, driving impressive performance gains. As a result, many HPC sites across a broad range of verticals are now using accelerators to speed up a larger portion of their increasingly AI-based workloads.

The HPC market is experiencing many challenges and opportunities from the covid-19 virus and its resulting economic impacts. In July 2020, the situation appeared bleak, with a potential for 2020 HPC server revenues being down by as much as 20% year on year. More recently, surveys show an improved outlook, although still with a market contraction of approximately 14% compared with 2019. The Workgroup segment is bearing the near-term brunt of the downside impact.

Covid-related HPC investment impacts are not equally distributed across the vertical markets; several segments are already experiencing increased HPC investments. These include bio/life sciences and government labs. At the same time other verticals are facing a greater negative impact, including oil & gas, manufacturing, and DCC/content distribution

Genomic modeling using high performance computing is beginning to prove its worth in the fight against the 2019 SARS-COV-2 (covid-19) pandemic. Using supercomputing resources at the Department of Energy’s Oak Ridge National Laboratory (ORNL), researchers have found a specific gene expression profile that makes patients particularly susceptible to some of the worst effects of the virus. This genomic pattern appears to be tied to biological pathways that explain many of the symptoms encountered by patients struggling to overcome the infection. As a result of this work, clinicians may be able to exploit a set of existing drugs to treat high-risk covid-19 patients.

Dating back to 1986, ISC is the world's longest running and Europe's foremost HPC conference. In June 2020, ISC held ISC2020 Digital, an online event for the HPC community. This report summarizes a presentation from that digital conference: an update on the EuroEXA project given by Peter Hopton, founder and innovator behind Iceotope and the VeryPC Group, as well as the chair of both the Data Centre Strategy and Future Computing Systems committees at the British Computer Society (the Chartered Institute for IT).

Hyperion Research's continuing efforts in Return on Investment (ROI) research examine individual HPC projects and measure the amounts spent on the HPC resources compared with the projects' financial and innovation returns. This report provides an update on this research, including new ROI and Return on Research (ROR) data and additional analysis. Key findings of this updated research include:

• Updated results continue to indicate substantial returns for investments in HPC: * The data now covers 763 successful HPC projects
• On average $507 dollars in revenue per dollar of HPC invested was generated * On average $47 dollars of profit (or cost savings) per dollar of HPC invested was generated * The average HPC investment per innovation was $2.6 million

The growing proliferation of GPUs and related HPDA and HPC-based AI workloads is challenging the HPC storage architectures of conventional CPU-based designs and traditional HPC modeling/simulation workloads. Conventional HPC storage systems manage the well-understood needs of largely independent and segregated home directories and scratch files, keeping the system fully utilized for optimal performance. However, data-intensive HPDA and AI workloads, with much greater variety of heterogenous I/O profiles, are stressing the performance capabilities of the conventional storage systems and related architectures.

According to a recent Hyperion Research study of 135 US quantum computing (QC) technology developers across the academic, commercial and government sectors, the global QC market was worth about $250 million (+/- $30 million) in 2019. Based on an anticipated CAGR of 27% between 2020 and 2024, the global QC market will grow from approximately $320 million in 2020 to $830 million in 2024. On-prem and cloud access QC hardware will comprise about 50% of the global QC market for the next three years, while optimization, physical simulation, and machine learning will near equally divide the algorithm space. This is Hyperion Research's baseline estimate for the rapidly evolving market, and Hyperion Research anticipates providing periodic updates as events, information, and decision-making requirements within the sector dictate.

HPC cloud adoption has grown aggressively over the last few years as cloud providers have recognized the HPC market as a high value target and have brought HPC-specific capabilities to their platforms to entice users. This has resulted in growing number of HPC users that are now using external clouds for parts of their HPC workload portfolio.   Hyperion Research has followed the growth vector of cloud adoption for many years and conducted a number of broad studies to better understand the underlying trends of cloud usage. This research has shown that users are contemplating different approaches of cloud adoption including what types of applications to send to the cloud; how much to send to the cloud; and ultimately whether to run a specific workload in the cloud or not.   As the performance capabilities of the cloud for HPC workloads improve, many of the barriers to adoption have been reduced, but some remained consistent barriers despite the technological advancements. These adoption inhibitors may be offset in many situations by the corresponding drivers and benefits. This will lead to increased cloud usage for many HPC applications and workloads in both traditional modeling/simulation as well as artificial intelligence.

Dating back to 1986, ISC is the world’s longest running and Europe’s foremost HPC conference. In June 2020, ISC held ISC2020 Digital, an online event for the HPC community. This update summarizes a presentation from that digital conference entitled, DEEP Framework for Deep Learning: Deep learning on a Cloud and HPC environment, given by Valentin Kozlov, a researcher from the Karlsruhe Institute of Technology.

The HPC User Forum was established in 1999 to promote the health of the global HPC industry and address issues of common concern to users. Recently, the 73rd HPC User Forum took place in Chicago, Illinois. This update summarizes a presentation from that meeting entitled, Using Graphs for Unstructured Data, given by Keshav Pingali, William Moncrief Chair of Grid and Distributed Computing, University of Texas at Austin.

Hyperion Research projects that between 2021 and 2026, worldwide installations of leading-edge exascale and near-exascale systems will total 28-38 new systems, worth an estimated $10-$15 billion, with China, the EU, and the United States each fielding 7 -10 systems in the six year interval. Being an exascale trailblazer can be expensive, with systems on the near horizon costing upwards of $500 million or more each, and only the most ambitious governments can commit to participating in these early rounds of exascale progress. However, as HPC price/performance moderates, a long-verified phenomenon in the HPC space, additional nations will also be looking to provide their domestic R&D base with access to exascale-class HPCs either developed domestically or procured from foreign sources.

An acquisition of Arm Technologies by Nvidia could transform the global microprocessor market, while significantly expanding the influence of the GPU-maker. If the deal comes to pass, it will merge two top-tier semiconductor firms under one roof, giving the new company access to a much larger combined market. That enables some important synergies, but also brings with it a few discordant elements. The most prominent advantage to Nvidia is the addition of Arm’s CPU IP and design expertise to the company’s own GPU assets. The addition of Arm would enable Nvidia to provide a more diverse set of offerings, expanding its market considerably, especially in the mobile and IoT markets. Furthermore, by taking advantage of the expertise within Arm, Nvidia would potentially be better able to develop integrated CPU-GPU hybrid products for its traditional PC and datacenter markets.

Dating back to 1986, ISC is the world’s oldest and Europe’s foremost HPC conference. In June 2020, ISC held ISC2020 Digital, an online event for the HPC community. This update summarizes a presentation from that digital conference entitled, Running a Pre-Exascale, Geographically Distributed, Multi-Cloud Scientific Simulation, given by Igor Sfiligoi, Lead Scientific Software Developer and Research at University of California San Diego Supercomputer Center.

According to a recent Hyperion Research study of 115 current and interested QC end-users from both HPC and enterprise IT organizations, the majority would consider a wide range of quantum computing (QC) technology and use-case options spanning new QC and QC-inspired applications, as well as speed-ups of existing applications delivered through a mix of QC or hybrid QC/classical systems. Likewise, these same QC buyers/users reported relatively modest expectations for realized performance gains from QC technology: 78% of respondents would see a performance boost of less than 250X as justification for using QC, and 42% would only need 50X or below. Such expectations bode well for the quantum computing sector writ large as QC developers and suppliers can explore a broad array of quantum technologies in both hardware and software with some assurance that end-users will be open to a span of QC use-case options with relatively modest near-term performance gains.

A digital simulation of the COVID-19 (SARS-CoV-2) spike protein that was run on the Frontera supercomputer at the Texas Advanced Computing Center (TACC) may provide needed insights for the design of potential vaccines and drugs being developed against the virus. Understanding the structure of the spike protein is crucial because the virus uses this component to infect human cells. Since this simulation provides a level of structural detail beyond that which is achievable by experiment, it promises to help researchers figure out how the virus is able to infect humans so effectively.

According to a recent Hyperion Research study of 115 current and interested QC end-users from both HPC and enterprise IT organizations, the average budget of surveyed QC buyers/users, worth about $3.2 million in 2019, will grow at a 27% CAGR between 2019 and 2024. This growth includes follow-on budget commitments from existing QC users as well as new budget commitments from a tranche of survey respondents who are currently only considering QC involvement. However, the number of organizations that are not yet certain of budget timetables over the next four years remains relatively constant at about 15% of those surveyed. A strong anticipated growth rate in QC end-user budgets implies high expectations for QC developments that can support adequate return on investment or sector-relevant competitive advantage. Current and potential QC users likely will be looking for QC suppliers to roll out a series of steady technological and use case advances that provide assurance that the QC sector is successfully moving to achieve stability, reducing the risk of being an early adopter.

Running HPC workloads in the cloud has undergone a fundamental shift over the past few years as CSPs have made a concerted effort to address the needs of HPC users, both for traditional HPC workloads as well as AI applications. Cloud adoption continues to rapidly increase for both current and new users, leading to a projection of end user HPC cloud spending to reach almost $9 billion by 2024. The spending for cloud resources by HPC end users primarily augments their spending for on-premise servers, as most users employ the cloud to complement, rather than replace, on-premise resources. There is much higher cloud adoption by workgroup users, suggesting that workgroup users are shifting some of their on-premise workloads to the cloud.

The recent decision by the Compute Express Link (CXL) and Gen-Z consortiums to collaborate on their respective technologies is a positive development that is likely to accelerate the acceptance of both memory fabric standards. Although not strictly competing against one another, prior to the agreement there was no assurance that Gen-Z and CXL supported devices would be compatible on future hardware. With the Memorandum of Understanding (MOU) recently agreed to by the two organizations, a path forward for interoperability and other collaborations now exists.

This Hyperion Research study presents our updated five-year forecast (2019-2024) for HPC on-premise server systems. This forecast includes the potential impacts to the HPC market from the covid-19 virus' effect on the world’s economy. Overall, the forecast has been reduced dramatically for 2020, with a partial recovery starting in 2021 and impacts lasting over the entire period.

While SSDs were initially integrated almost exclusively into the largest and highest performing HPC systems, widespread HPC adoption of flash-based storage is now occurring due to the perfect storm evolution and intersection of requirements, technologies and economics.

Nvidia’s recently announced Ampere GPU for datacenters (A100) comes to the market at a time of increased market competition for HPC and AI silicon. However, rather than offering specialized datacenter products aimed at these two application categories, the new Ampere offering carries forward the dual HPC/AI approach previously introduced in Nvidia's Volta architecture. To realize this, the company has introduced a number of innovations that significantly boost performance in each area.

This Special Study looks at the evolution and changes in file systems used for technical computing. Although the file system sector may appear to be mature, significant changes are occurring across many fronts spanning business, technology, workload requirements, and competitive forces.

Researchers at the Tokyo Institute of Technological (Tokyo Tech) with assistance from colleagues at Hitachi, Hokkaido College, and the University of Tokyo recently announced a promising new processor architecture that could advance the performance and capability of quantum annealing systems. Known as STATICA (Stochastic Cellular Automata Annealer Architecture), the novel architecture seeks to enable more parallelism in both connectivity and computation compared with counterpart quantum annealing devices. The processor is designed to take on complex optimization problems typically found in important applications including portfolio, logistic, and traffic flow analysis workloads that often cannot be effectively computed on counterpart classical systems.

Hybrid quantum-classical algorithms show significant promise in effectively harnessing the computational capabilities of near-term quantum systems when combined with classical HPC systems on key applications. Two recent developments reveal a growing interest in advancing such algorithms: the first, a trio of European-based organizations partnering to build a complete quantum computing R&D ecosystem: the second, a commercial partnership between quantum annealing vendor D-Wave and the Japanese supercomputer vendor, NEC. Successful growth of hybrid capabilities will require a merging of quantum and related HPC skills and can only contribute to the overall development pace of the quantum computing sector writ large. Additional partnerships with similar intentions will likely emerge soon.

The US White House recently released its proposed FY 2021 budget to Congress that seeks to impose a number of budget cuts on federal agencies, some of which are the largest purchasers and users of HPCs in the world. Notable reductions are proposed for NOAA (-31%), DOE's Office of Science (-17%), and NASA (-11%).

• On the plus side, two new advanced computing technologies, artificial intelligence (AI) and quantum information science (QIS) are earmarked for near doublings in budget and are spread across a number of US government agencies.
• As has happened in the recent past, Congress will almost certainly increase funding levels for some critical R&D, but competing financial priorities driven by the covid-19 economic fallout could limit the ability of Congress to adjust funding levels for activities not directly linked to near-term covid-19 policies and programs.

The recent announcement of a novel neural network processor by Silicon Valley startup Groq reflects the diversification taking place in hardware accelerators being developed for the artificial intelligence (AI) market. Groq’s Tensor Streaming Processor (TSP) uses a simplified architecture to enforce deterministic behavior, which offers a unique attribute for machine learning inference in both datacenter applications and for those at the edge.

By most objective measures, the United States is the leader in AI capability, both in its AI technology and its application. Whether looking at R&D spending, research capability, ecosystem support, or commercial deployment, the US is at the top or near the top of its global competition. That strength is derived largely from successes in the US private sector. In particular, computer companies in Silicon Valley and elsewhere have been at the forefront in developing new AI hardware and software technologies. A receptive user environment in areas such as web services, financial services, healthcare, and scientific research, among others, has created a virtuous dynamic between supply and demand for AI technologies and products. As a result, the depth and breadth of AI capability in the US is such that the country is well positioned to maintain its lead position in the near-term.

The HPC User Forum was established in 1999 to promote the health of the global HPC industry and address issues of common concern to users. In September 2019, the 73rd HPC User Forum took place at Argonne National Laboratory in Illinois. This update summarizes a presentation from that meeting entitled, Argonne Training Program on Extreme - Scale Computing, given by Ray Loy, ATPESC 2020 Program Director, Argonne National Laboratory.

The US Department of Energy recently initiated a program to establish as many as ten Quantum Information Science (QIS) research centers over the next five years and is currently accepting applications for the first round of two to five centers. Each center will have a specific focus area among highlighted quantum technologies including computing, communication, and sensing.

Over the last three years, AMD has become a force to be reckoned with in the high performance computing market. The recent announcements by AMD that its future AMD EPYC™ CPUs and AMD Radeon Instinct™ GPUs will be used to power the El Capitan supercomputer at Lawrence Livermore National Laboratory (LLNL) in 2023 and the Frontier supercomputer at Oak Ridge National Laboratory (ORNL) in 2021, demonstrate that the chipmaker may be able to effectively compete against other vendors for inclusion in some of the most powerful HPC systems in the world.