Recent AI applications have revolutionized the modeling of structurally unresolved protein regions, thereby complementing traditional computational methods. These state-of-the-art techniques can generate numerous candidate structures, significantly expanding the scope of structural biology. However, to effectively prioritize these models, a physics-based approach is required to assess the energy landscape. Such integration can bridge the gap between rapid model generation and precise determination of functional conformations. To address this challenge, we propose an integrated approach that combines molecular modeling with AI and HPC. Metadynamics simulations in latent space are used to explore potential energy landscapes based on initial approximations of flexible region structures derived from modeling tools such as AlphaFold, RosettaFold, Modeller, SwissModel, etc. The approach was validated by modeling folding of Trp-cage protein and conformational plasticity of ubiquitin. The predominant conformations of previously unresolved mobile regions in the active center of flavin-dependent 2-hydroxybiphenyl-3-monooxygenase (EC 1.14.13.44) were identified, while estimating the energy associated with these conformational changes.

We present the algorithm of exoplanets transit search in MASTER global robotic telescope network database, that has been created with the involvement and use of artificial intelligence methods. MASTER’s 2003–2023 years archive images are stored at data storage and analysis center in Lomonosov supercomputer. For exoplanets transit search we used the images of SWIFT gamma-ray bursts alerts (GRB), received by MASTER telescopes during several hours at target nights. The algorithm includes cross-correlation of GRB error-boxes coordinates and TESS exoplanetary transits candidates inside 4 square degrees of each field. Our analyses is based on wide-field images obtained on MASTER-Amur, -Tunka, -Kislovodsk, -Tavrida, MASTER-IAC, MASTER-OAGH, MASTER-OAFA and MASTER-SAAO robotic telescopes. We used data storage and analysis center in Lomonosov supercomputer with MASTER images archive, calculated light curves for target stars and approximate them to find eclipse by exoplanets. The centralized storage of this data allows us to simplify and significantly speed up access to the huge archive of images formed during the entire operation of the robotic network for automatic processing and facilitate the manual search for the necessary information by the researcher. From this storage we compared the program with unique long homogeneous series of Swift GRBs alert and inspection observations, which we analyzed. The main result is reduction of unique light curves of exoplanetary transits of TIC 127115861.01 TESS candidate, obtained by MASTER long before the TESS space observatory.

Feynman integral reduction by means of integration-by-parts identities is a major power gadget in a theorist toolbox indispensable for calculation of multiloop quantum effects relevant for particle phenomenology and formal theory alike. An algorithmic approach consists of solving a large sparse non-square system of homogeneous linear equations with polynomial coefficients. While an analytical way of doing this is legitimate and was pursued for decades, it undoubtedly has its limitations when applied in complicated circumstances. Thus, a complementary framework based on modular arithmetic becomes critical on the way to conquer the current `what is possible' frontier. This calls for use of supercomputers to address the reduction problem. In order to properly utilize these computational resources, one has to efficiently optimize the technique for this purpose. Presently, we discuss and implement various methods which allow us to significantly improve performance of Feynman integral reduction within the FIRE environment.

Supercomputer co-design provides for coordination of the specifics of solved problems, methods, algorithms and their implementations with the architecture of computing systems. In this case, supercomputer co-design technologies usually provide for the solution of the following tasks: search for an optimal method or algorithm for solving a certain problem on a particular supercomputer, construction of an optimal implementation of the selected method and algorithm on the target computer system, as well as the selection or construction of the most appropriate hardware and software platform for solving the problem using the selected method or algorithm on the basis of this or that software implementation. All three variants of the scientific problem statement are meaningful, require different approaches for solution, but have in common—the necessity of joint analysis of properties of problems, methods, algorithms, software implementations and computer architectures for effective solution.

Performance analysis of supercomputing applications is an essential task that should be performed by any HPC user, since optimizing job performance reduces, sometimes by several times and even by orders of magnitude, the speed of carrying out computational experiments. There are many advanced analysis software tools that can be used for this purpose, but it is difficult for many users to figure out which tool to choose in their particular situation, how to use it and how to interpret the obtained results. In order to tackle this challenge, we are developing a guide, which describes a number of common steps that can be helpful for analyzing the performance of most applications running on modern CPUs. The paper also shows how the proposed guide was used for the analysis of real-life application for modeling the atmospheric boundary layer, which was performed on the Lomonosov-2 supercomputer.

В настоящее время графовые алгоритмы очень часто применяются для решения различных задач моделирования, поскольку многие реальные объекты хорошо моделируются графами (например, дорожная сеть или социальные связи). При этом эффективная реализация таких алгоритмов зачастую очень сложна, что связано, в частности, с нерегулярным доступом к памяти при работе с графами и огромным размером входных графов. Помочь с решением этой проблемы могут графовые фреймворки — программные среды для решения графовых задач. Ранее был разработан архитектурно-независимый фреймворк VGL (Vector Graph Library), позволяющий эффективно реализовывать графовые алгоритмы на различных аппаратных платформах (на многоядерных процессорах с векторными расширениями, графических ускорителях и векторных процессорах NEC). В данной работе было проведено изучение производительности VGL на разных платформах, выполнено сравнение производительности с существующими аналогами, а также предложен и апробирован подход для автоматического выбора формата входного графа на основе методов машинного обучения. Graph algorithms are currently often used to solve various modeling tasks, since many real-life objects are well modeled by graphs (for example, a road network or social connections). At the same time, the efficient implementation of such algorithms is often very complex, which is due, in particular, to irregular memory access when working with graphs and the huge size of the input graphs. Graph frameworks — software environments for implementing graph algorithms — can help solve this problem. Previously, an architecture-independent VGL (Vector Graph Library) framework was developed that allows for efficient implementation of graph algorithms on various hardware platforms (multi-core processors with vector extensions, graphics accelerators and NEC vector processors). In this work, the performance of VGL was studied on different platforms, its performance was also compared with existing analogues, and an approach for automatic selection of input graph format based on machine learning methods was proposed and evaluated.

Motivation: With the increasing availability of 3D-data, the focus of comparative bioinformatic analysis is shifting from protein sequence alignments toward more content-rich 3D-alignments. This raises the need for new ways to improve the accuracy of 3D-superimposition. Results: We proposed guide tree optimization with genetic algorithm (GA) as a universal tool to improve the alignment quality of multiple protein 3D-structures systematically. As a proof of concept, we implemented the suggested GA-based approach in popular Matt and Caretta multiple protein 3D-structure alignment (M3DSA) algorithms, leading to a statistically significant improvement of the TM-score quality indicator by up to 220-1523% on 'SABmark Superfamilies' (in 49-77% of cases) and 'SABmark Twilight' (in 59-80% of cases) datasets. The observed improvement in collections of distant homologies highlights the potentials of GA to optimize 3D-alignments of diverse protein superfamilies as one plausible tool to study the structure-function relationship.

With the regular appearance of new multi-core processors, it is extremely important to detect strengths and weaknesses of developed hardware solutions. Every new prominent computational platform has a potential to outperform existing ones on a subset of application classes. This is also a crucial challenge for ARM processors, because of their relatively novelty in High-Performance Computing (HPC) area and lack of detailed performance evaluation. This paper covers the methodology of creating a full–fledged hierarchical benchmarking set. Such completeness is achieved by using a hierarchical set of benchmarking kernels and algorithms, which are classified according to the roofline model. This proposed approach allows us to estimate the performance of modern multi-core processors through relatively simple but frequently used fragments of code. All these fragments are aimed to stress a particular memory subsystem component or computing units of target processor, which allows to easily map the on the roofline classification. Among other ARM solutions for HPC, Huawei company developed its own family of 64-bit ARM multi-core server SoCs called Kunpeng. To evaluate Huawei Kunpeng 920 processor performance, we launched our developed benchmarking set on other HPC platforms: Intel Xeon 6140 and Fujitsu A64FX. The evaluation results allowed us to identify multiple strengths and weaknesses of Kunpeng 920 processors, as well as to demonstrate their relevance to some of the real-world algorithms and applications.

High Performance Computing is now one of the emerging fields in computer science and its applications. Top HPC facilities, supercomputers, offer great opportunities in modeling diverse processes thus allowing to create more and greater products without full-scale experiments. Current supercomputers and applications for them are very complex and thus are hard to use efficiently. Performance monitoring systems are the tools that help to understand the efficiency of supercomputing applications and overall supercomputer functioning. These systems collect data on what happens on a supercomputer (performance data, performance metrics) and present them in a way allowing to make conclusions about performance issues in programs running on the supercomputer. In this paper we give an overview of existing performance monitoring systems designed for or used on supercomputers. We give a comparison of performance monitoring systems found in literature, describe problems emerging in monitoring large scale HPC systems, and outline our vision on future direction of HPC monitoring systems development.

Any modern supercomputer has an extremely complex architecture, and efficient usage of its resources is often a very difficult task, even for experienced users. At the same time, the field of high-performance computing is becoming more and more in demand, so the issue of efficient utilization of supercomputers is very urgent. Therefore, users should know everything important about performance of their jobs running on a supercomputer in order to be able to optimize them, and administrators should be able to monitor and analyze all the nuances of the efficient functioning of such systems. However, there is currently no complete understanding of what data are best to be studied (and how it should be analyzed) in order to have a whole picture of the state of the supercomputer and the processes taking place there. In this paper, we make our first attempt to answer this question. To do this, we are developing a model that describes all the potential factors that may be important when analyzing the performance of supercomputer applications and the HPC system as a whole. The paper provides both a detailed description of this model for users and administrators and some interesting real-life examples discovered on the Lomonosov-2 supercomputer using a software implementation based on the proposed model. Любой современный суперкомпьютер имеет крайне сложную архитектуру, и эффективное использование его ресурсов зачастую является очень сложной задачей даже для опытных пользователей. В то же время высокопроизводительные вычисления становятся все более востребованными и вопрос эффективного использования суперкомпьютеров очень актуален. Поэтому пользователи должны знать все самое важное о производительности их работы, выполняемой на суперкомпьютере, чтобы иметь возможность ее оптимизировать, а администраторы должны уметь отслеживать и анализировать все нюансы эффективного функционирования таких систем. Однако в настоящее время нет полного понимания того, какие данные лучше всего для этого изучать (и как их следует анализировать), чтобы иметь полную картину состояния суперкомпьютера и происходящих на нем процессов. В этой статье мы делаем нашу первую попытку ответить на этот вопрос. Для этого мы разрабатываем модель, которая описывает все потенциальные факторы, которые могут быть важными при анализе производительности суперкомпьютерных приложений и системы HPC в целом. В документе представлено как подробное описание этой модели для пользователей и администраторов, так и несколько интересных реальных примеров, обнаруженных на суперкомпьютере Ломоносов-2 с помощью программного обеспечения, реализованного на основе предложенной модели.

Developing efficient graph algorithms implementations is an extremely important problem of modern computer science, since graphs are frequently used in various real-world applications. Graph algorithms typically belong to the data-intensive class, and thus using architectures with high-bandwidth memory potentially allows to solve many graph problems significantly faster compared to modern multicore CPUs. Among other supercomputer architectures, vector systems, such as the SX family of NEC vector supercomputers, are equipped with high-bandwidth memory. However, the highly irregular structure of many real-world graphs makes it extremely challenging to implement graph algorithms on vector systems, since these implementations are usually bulky and complicated, and a deep understanding of vector architectures hardware features is required. This paper presents the world first attempt to develop an efficient and simultaneously simple graph processing framework for modern vector systems. Our vector graph library (VGL) framework targets NEC SX-Aurora TSUBASA as a primary vector architecture and provides relatively simple computational and data abstractions. These abstractions incorporate many vector-oriented optimization strategies into a high-level programming model, allowing quick implementation of new graph algorithms with a small amount of code and minimal knowledge about features of vector systems. In this paper, we evaluate the VGL performance on four widely used graph processing problems: breadth-first search, single source shortest paths, connected components, and page rank. The provided comparative performance analysis demonstrates that the VGL-based implementations achieve significant acceleration over the existing high-performance frameworks and libraries: up to 14 times speedup over multicore CPUs (Ligra, Galois, GAPBS) and up to 3 times speedup compared to NVIDIA GPU (Gunrock, NVGRAPH) implementations.

This paper provides an overview of the current state of supercomputer technology. The review is done from different points of view — from the construction features of modern computing devices to the features of the architecture of large supercomputer complexes. This review includes descriptions of the most powerful supercomputers in the world and Russia since the early of 2021 as well as some less powerful systems that are interesting from other points of view. It also focuses on the development trends of the supercomputer industry and describes the most famous projects for building future exascale supercomputers.

The field of biology is producing data faster than ever before. To approach its analysis efficiently, one can take advantage of a systematic symbiosis between the power of huge distributed-memory systems and the flexibility of locally managed shared-memory servers. Here, we introduce the concept of a "small and smart" High Performance Computing resource as an intermediary between a personal computer and a supercomputer to accommodate complex tasks in biology. Such a supplementary server, specifically designed and configured for a particular purpose, can facilitate efficient and convenient data processing by maximizing shared-memory multi-tasking/threading on CPU/GPU. It seems most reasonable to consider it as a beacon for further co-development of the increasingly common "medium-hard" algorithms/pipelines in life sciences and hardware for shared-memory parallelism. This leaves supercomputers with specifically developed highly parallel implementations of the most resource-demanding operations, to ensure optimal performance amid efficient utilization of expensive infrastructure.

There is a variety of known HPC ratings nowadays which represent machine capability for solving a fixed problem, based on a certain algorithm, but these ratings represent a top of the iceberg, and as a rule, one can’t compare application tuning features even for the selected system, and the details of system architecture are not usually described precisely. At the same time lots of efforts are made to describe diverse algorithm features formally, AlgoWiki is one of the most notable recent projects. The idea of Algo500 is joining precise description of computer system with detailed formal descriptions of algorithms using implementation performance data, and building an engine over such joint base to allow various queries, thus giving means of building user-defined ratings regarding selected method, algorithms and/or computer platform features. This paper gives an overview of Algo500 design and some use cases.

Breadth-First Search (BFS) is an important computational kernel used as a building-block for many other graph algorithms. Different algorithms and implementation approaches aimed to solve the BFS problem have been proposed so far for various computational platforms, with the direction-optimizing algorithm being the fastest and the most computationally efficient for many real-world graph types. However, straightforward implementation of direction-optimizing BFS for vector computers can be extremely challenging and inefficient due to the high irregularity of graph data structure and the algorithm itself. This paper describes the world’s first attempt aimed to create an efficient vector-friendly BFS implementation of the direction-optimizing algorithm for NEC SX-Aurora TSUBASA architecture. SX-Aurora TSUBASA vector processors provide high-performance computational power together with a world-highest bandwidth memory, making it a very interesting platform for solving various graph-processing problems. The implementation proposed in this paper significantly outperforms the existing state-of-the-art implementations both for modern CPUs (Intel Skylake) and NVIDIA V100 GPUs. In addition, the proposed implementation achieves significantly higher energy efficiency compared to other platforms and implementations both in terms of average power consumption and achieved performance per watt.

Managing and administering an HPC center is a real challenge. The Supercomputing Center at Moscow State University is the largest in Russia, with petascale machines running in the interest of thousands of users engaged in hundreds of research projects. Since its early stages of development, the Octoshell system has been instrumental in mastering the equipment and tackling technical issues. The more problems are solved, the more new ideas arise and chances for improvement are discovered. In this paper, we discuss major enhancements recently made to Octoshell, including the integration of new modules (such as extended object description, license management, hardware accounting, and others), and outline plans for the near future.

Предложен метод проведения анализа эффективности и оптимизации суперкомпьютерных приложений, примененный на практике для изучения задач одного из пользователей суперкомпьютера Ломоносов2. Этот метод затрагивает различные этапы исследования задач, начиная от изучения общего поведения всех запусков пользователя на суперкомпьютере и заканчивая детальным изучением и оптимизацией исходного кода выбранной программы. Приведено описание общих этапов анализа, которые были выполнены на практике, показаны метрики производительности, на которые следует обратить внимание при выполнении подобного анализа, а также продемонстрированы конкретные примеры поведения задач и эффект от оптимизации, выполненной для задачи расчета жидкокристаллических капель. A method for the efficiency analysis and optimization of supercomputer applications applied earlier in practice to study jobs of a user on the Lomonosov2 supercomputer is proposed. This method involves various stages of the jobs research, starting from studying the general behavior of all user launches on a supercomputer and ending with a detailed study and optimization of the source code of a selected program. The paper describes the general stages of the analysis that were carried out in practice, shows performance metrics that should be paid attention to when performing such an analysis, and shows also some specific examples of the job behavior and the effect of optimization carried out for the task of calculating liquid crystal droplets.

Supercomputer technologies are used everywhere in the world today. They determine the competitiveness of science, industry, and states. The United States, Europe, China, and Japan are investing billions in their development, promoting national programs in this field, and forming the basis for the future. Russia underestimates the importance of supercomputer technologies. Not only has it not joined the global trend, but it also misses a unique chance to use a huge potential of Russian specialists in this field. Our country needs a comprehensive supercomputer program.

Deep neural networks (DNN) offer great opportunities for solving many problems associated with processing large-scale data. Building and using deep neural networks requires large computational resources. In this regard, the question naturally arises about the possibility of using HPC-systems for the implementation of DNN. In order to better understand the performance implications of DNN on High Performance clusters we analyze the performance of several DNN-models over 2 HPC systems: the Lomonosov-2 supercomputer (the section with processors equipped with P100 GPUs) and the Polus high-performance cluster based on IBM Power8 processors with P100 GPUs. Comparing these frameworks is interesting as they represent different types of processors (Intel for Lomonosov-2 and IBM for Polus). Apart from different processor architectures, these systems feature different internode communications, which may affect the performance of the analysed algorithms in case of parallel and distributed implementation of neural network algorithms. The studies were carried out on the basis of the PyTorch framework.