• A
  • A
  • A
  • АБВ
  • АБВ
  • АБВ
  • A
  • A
  • A
  • A
  • A
Обычная версия сайта
  • RU
  • EN
  • HSE University
  • Publications
  • Book chapter
  • Multi-GPU GEMM Algorithm Performance Analysis for Nvidia and AMD GPUs Connected by NVLink and PCIe
  • RU
  • EN
Расширенный поиск
Высшая школа экономики
Национальный исследовательский университет
Priority areas
  • business informatics
  • economics
  • engineering science
  • humanitarian
  • IT and mathematics
  • law
  • management
  • mathematics
  • sociology
  • state and public administration
by year
  • 2028
  • 2027
  • 2026
  • 2025
  • 2024
  • 2023
  • 2022
  • 2021
  • 2020
  • 2019
  • 2018
  • 2017
  • 2016
  • 2015
  • 2014
  • 2013
  • 2012
  • 2011
  • 2010
  • 2009
  • 2008
  • 2007
  • 2006
  • 2005
  • 2004
  • 2003
  • 2002
  • 2001
  • 2000
  • 1999
  • 1998
  • 1997
  • 1996
  • 1995
  • 1994
  • 1993
  • 1992
  • 1991
  • 1990
  • 1989
  • 1988
  • 1987
  • 1986
  • 1985
  • 1984
  • 1983
  • 1982
  • 1981
  • 1980
  • 1979
  • 1978
  • 1977
  • 1976
  • 1975
  • 1974
  • 1973
  • 1972
  • 1971
  • 1970
  • 1969
  • 1968
  • 1967
  • 1966
  • 1965
  • 1964
  • 1963
  • 1958
  • More
Subject
News
September 4, 2026
Time to Showcase Your Research: Applications Are Now Open for Student Research Paper Competition 2026
Taking part in the Student Research Paper Competition (SRPC) gives you an opportunity to present your research to experts, receive an independent assessment, and determine the future direction of your work. The competition is open to students graduating in 2026 not only from HSE University but from universities in Russia and abroad. Papers may be submitted in Russian and English, and in some fields also in French, German, and Spanish.
September 4, 2026
‘Hedgehog Versus ‘Relatives: Researchers Measure How the Brain Responds to Unexpected Words During Natural Speech
Russian neurophysiologists, including researchers from HSE University, have demonstrated the feasibility of using event-related fields (ERFs) to study brain activity during natural speech perception. The researchers showed that this approach can be applied not only to individual words but also to continuous speech. Their findings indicate that words whose meanings differ significantly from the preceding context require longer processing times. The study also reveals that the brain processes function words in two stages: first, it identifies their grammatical role and then uses this information to predict the next word. The study has been published in Frontiers in Human Neuroscience.
August 25, 2026
Scientists Develop Algorithm for More Reliable Processors in Data Centres
Researchers from HSE MIEM and Samara University have developed the LRF-3D algorithm to automatically bypass idle nodes in three-dimensional networks-on-chip. Thanks to its hierarchical architecture, the algorithm outperforms existing solutions in both speed and path accuracy, improving processor reliability for use in data centres, supercomputers, and AI computing. The source code and test results are publicly available.

 

Have you spotted a typo?
Highlight it, click Ctrl+Enter and send us a message. Thank you for your help!

Publications
  • Books
  • Articles
  • Chapters of books
  • Working papers
  • Report a publication
  • Research at HSE

?

Multi-GPU GEMM Algorithm Performance Analysis for Nvidia and AMD GPUs Connected by NVLink and PCIe

Ch. 23. P. 281–292.
Choi Y. R., Stegailov V.

Modern types of multi-GPU servers combine up to 8 A100 GPUs connected by NVLink 3.0 links through NVSwitch. This connectivity provides unprecedented capabilities for multi-GPU algorithms. In this work, we analyze the performance of matrix-matrix multiplication algorithm developed by us previously. Tuning principles and limits for maximum performance are discussed. Algorithm performance for much more affordable 4 AMD Radeon RX 6900 XT based server with PCI 4.0 working under ROCm HIP is described for comparison.

Language: English
DOI
Text on another site
Keywords: parallel computingCUDAGEMMhigh-speed GPU interconnectmulti-GPU programmingHIP
Publication based on the results of:
Supercomputer technologies, theory and atomistic multiscale modeling in materials physics and life sciences (2022)

In book

22nd International Conference, MMST 2022, Nizhny Novgorod, Russia, November 14–17, 2022, Revised Selected Papers
Springer, 2022.
Similar publications
Performance Analysis of Computational Devices in Quantum Chemistry Tasks
Nedomolkin I., Konikov M., Fedorov I. et al., , in: Parallel Computational Technologies, 19th International Conference, PCT 2025, Moscow, Russia, April 8–10, 2025, Revised Selected Papers. (CCIS, volume 2891)Vol. 2891.: Springer, 2026. P. 513–531.
Modern supercomputer systems play a crucial role in scientific and engineering research. To ensure their effectiveness, these fields require reliable methods for evaluating supercomputer performance. Although benchmarking is a fundamental tool, current ranking systems often inadequately represent real-world performance in high-performance computing (HPC) applications. As a result, employing actual scientific software packages provides a more ...
Added: May 19, 2026
Hardware-Software Complex for Network-on-Chip Prototyping Using Multiple FPGAs
Mikhail Y. Romashikhin, Aleksandr Y. Romanov, IEEE Access 2026 Vol. 14 P. 7921–7931
This paper presents a hardware-software multi-FPGA complex designed for hardware prototyping of networks-on-chip (NoCs). The rationale for the use of multiple FPGAs for NoC prototyping is given. The architecture of the complex and its components–the software part generating top-level files and configuration files describing the NoC for several FPGAs, hardware part consisting of interfacing switches ...
Added: January 22, 2026
GEMM Algorithm for Multi-GPU Platforms with Regular Uneven Data Transfer Links
Choi Y. R., Malkovsky S., Stegailov V., , in: 11th Russian Supercomputing Days, RuSCDays 2025, Moscow, Russia, September 29–30, 2025, Revised Selected Papers.: Springer, 2026. Ch. 3 P. 32–47.
Multi-GPU servers often exhibit uneven characteristics. For instance, the data transfer bandwidth between four NVIDIA V100 GPUs can vary due to the NVLink connecting these devices to a specific CPU in servers with IBM POWER 9 processors, which means that the communication bandwidth between other devices is comparably slower. To address this issue, the Multi-GPU ...
Added: January 3, 2026
Tracing of GPU-Aware MPI Applications: First Benchmarks for the Angara Interconnect
Ismagilov T., Mukosey A., Smirnov F. et al., , in: Parallel Processing and Applied Mathematics. 15th International Conference, PPAM 2024, Ostrava, Czech Republic, September 8–11, 2024, Revised Selected Papers, Part IVol. 15579: Lecture Notes in Computer Science.: Springer, 2025. P. 256–270.
The efficiency of data transfer is one of the most important issues of supercomputer development in the post-Moore era. The rise of heterogeneous computing systems introduces such complicated patterns of data transfers as, for instance, the GPU-aware MPI technology. The practical deployment of this technology in applications requires the development of the dedicated system software ...
Added: December 26, 2025
«Cтройка» – компьютерная игра для знакомства с параллельным программированием
Воронова К. Д., Plaksin M. A., В кн.: Актуальные проблемы математики, механики и информатики 2022: Сборник статей по материалам студенческой конференции (г. Пермь, ПГНИУ, 25 мая – 10 июня 2022 г.).: Пермь: ПГНИУ, 2022. С. 25–29.
The rapid development of parallel computing technologies makes it urgent to include propaedeutics of parallel computing in the school computer science course. Since this topic is not yet included in the school curriculum, it can be done through extracurricular activities, in particular, through Internet contests. Since 2013, parallel computing tasks have become a compulsory part ...
Added: February 29, 2024
Methods for Changing Parallelism in the Process of High-Level VLSI Synthesis
Ryzhenko I. N., Nepomnyaschy O. V., A. I. Legalov et al., Automatic Control and Computer Sciences 2023 Vol. 57 No. 7 P. 696–705
In this paper, methods for increasing the efficiency of VLSI development based on the method of architecture-independent design are proposed. The route of high-level VLSI synthesis is considered. The principle of constructing a VLSI hardware model based on the functional-flow programming paradigm is stated. The results of the development of methods and algorithms for the ...
Added: February 27, 2024
GPU-Accelerated Matrix Exponent for Solving 1D Time-Dependent Schrödinger Equation
Choi Y. R., Stegailov V., , in: Supercomputing: 9th Russian Supercomputing Days, RuSCDays 2023, Moscow, Russia, September 25–26, 2023, Revised Selected Papers, Part I.: Springer, 2023. P. 100–113.
Non-adiabatic electron-ion quantum dynamics is still an area of many unresolved problems even for such simple systems as the H2+ molecular ion. Mathematical modelling based on time-dependent Schrödinger equation (TDSE) is an important method that can provide better understanding of these phenomena. In this work, we present TDSE solution for 1D TDSE that describes non-adiabatic electron-ion ...
Added: January 26, 2024
GPU-based molecular dynamics of fluid flows: Reaching for turbulence
Pavlov D., Galigerov V., Kolotinskii D. et al., International Journal of High Performance Computing Applications 2024 Vol. 38 No. 1 P. 34–49
Fluid dynamics is a ubiquitous problem that arises in different branches of science and industry. It is usually tackled by numerically solving differential equations on a finite grid. Molecular dynamics was not a feasible tool to approach fluid dynamics until very recently due to its disproportional computational complexity. In this paper we propose a new ...
Added: July 18, 2023
ЗНАКОМСТВО С ПАРАЛЛЕЛЬНЫМИ ВЫЧИСЛЕНИЯМИ В РАМКАХ ДИСТАНЦИОННОГО КОНКУРСА «ТРИЗФОРМАШКА-2022»
Воронова К. Д., Plaksin M. A., В кн.: Дистанционное обучение – образовательная среда XXI века : материалы XII Междунар. науч.-метод. конф. (Республика Беларусь, Минск, 26 мая 2022 года).: Мн.: БГУИР, 2022. С. 163–163.
It is proposed to introduce schoolchildren and students to the basics of parallel computing using the distance competition "TRIZformashka". For the competition "TRIZformashka-2022" the computer game "Builder" was specially developed for teaching how to build parallel algorithms. A description of the game and a download link are given. ...
Added: October 31, 2022
Supercomputing: 7th Russian Supercomputing Days, RuSCDays 2021, Moscow, Russia, September 27–28, 2021, Revised Selected Papers
Springer, 2021.
Added: October 19, 2022
Tuning of a Matrix-Matrix Multiplication Algorithm for Several GPUs Connected by Fast Communication Links
Choi Y. R., Nikolskiy V., Stegailov V., , in: Parallel Computational Technologies: 16th International Conference, PCT 2022, Dubna, Russia, March 29–31, 2022, Revised Selected Papers.: Springer, 2022. Ch. 12 P. 158–171.
Added: August 11, 2022
Algorithm for Adaptive Mesh Redistribution in Lattice Boltzmann Simulations
Ziganurova L., Shchur L., Lobachevskii Journal of Mathematics 2022 Vol. 43 No. 2 P. 513–518
The Lattice Boltzmann method (LBM) is the alternative approach for hydrodynamic equation solving. Two factors make it a favorite approach nowadays. Firstly, the attractive feature of LBM is that it is intrinsic for parallel simulations due to the linear structure of the algorithm. Secondly, what makes LBM special for the research, it is well applicable to the simulations ...
Added: May 25, 2022
Методы преобразования параллелизма в процессе высокоуровневого синтеза СБИС
Рыженко И. Н., Непомнящий О. В., Легалов А. И. et al., Моделирование и анализ информационных систем 2022 Т. 29 № 1 С. 60–72
In this paper methods for increasing the efficiency of VLSI development based on the method of architecture-independent design are proposed. The route of high-level VLSI synthesis is considered. The principle of constructing a VLSI hardware model based on the functional-flow programming paradigm is stated. The results of the development of methods and algorithms for transformation functional-parallel ...
Added: March 18, 2022
  • About
  • About
  • Key Figures & Facts
  • Sustainability at HSE University
  • Faculties & Departments
  • International Partnerships
  • Faculty & Staff
  • HSE Buildings
  • HSE University for Persons with Disabilities
  • Public Enquiries
  • Studies
  • Admissions
  • Programme Catalogue
  • Undergraduate
  • Graduate
  • Exchange Programmes
  • Summer University
  • Summer Schools
  • Semester in Moscow
  • Business Internship
  • Research
  • International Laboratories
  • Research Centres
  • Research Projects
  • Monitoring Studies
  • Conferences & Seminars
  • Academic Jobs
  • Yasin (April) International Academic Conference on Economic and Social Development
  • Media & Resources
  • Publications by staff
  • HSE Journals
  • Publishing House
  • iq.hse.ru: commentary by HSE experts
  • Library
  • Economic & Social Data Archive
  • Video
  • HSE Repository of Socio-Economic Information
  • HSE1993–2026
  • Contacts
  • Copyright
  • Privacy Policy
  • Site Map
Edit