Informacja

Drogi użytkowniku, aplikacja do prawidłowego działania wymaga obsługi JavaScript. Proszę włącz obsługę JavaScript w Twojej przeglądarce.

Wyszukujesz frazę "GPU" wg kryterium: Temat


Tytuł:
Accelerating computation of a reduced order model of a structural system resulting from Craig–Bampton reduction using GPU programming
Autorzy:
Górecki, Piotr
Kalinowski, Miłosz
Jeziorek, Łukasz
Broniszewski, Jakub
Koziara, Tomasz
Powiązania:
https://bibliotekanauki.pl/articles/38706110.pdf
Data publikacji:
2024
Wydawca:
Instytut Podstawowych Problemów Techniki PAN
Tematy:
GPU
CPU
reduced order model
structural model
CuPy array library
model zredukowanego zamówienia
model strukturalny
CuPy
Opis:
The Craig–Bampton (CB) method is a well-known substructuring technique that reduces the size of a finite element model (FEM) using a set of vibration modes. For large FEA models, the reduction process could be computationally expensive since it requires algebra operations on FEM mode shapes and FEM system sparse matrices. In this paper, we investigate the potential of usage of GPU parallel processing to speed up solving the system of linear equations that results from the CB reduction process made for a model of cyclic structures. A Python based high-level approach, employing the CuPy, GinkGo and STRUMPACK libraries on the GPU, is compared with an optimized Fortran code. In side-to-side comparisons, employing the same inputs, the Python-GPU code is run on a single GPU device and the Fortran code is run on a multi-core compute node. The CB reduction process was split into several parts, each dealing with different kind of algebraic formulation of the problem. Performance comparisons were focused on the sparse system linear solver, since it turned out to be the most time-consuming part. The results suggest that the current GPU-based linear sparse solvers do not surpass the state-of-the-art CPU-based MKL PARDISO solver (at least up to 1M DOFs).
Źródło:
Computer Assisted Methods in Engineering and Science; 2024, 31, 1; 51-66
2299-3649
Pojawia się w:
Computer Assisted Methods in Engineering and Science
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
Metody działania rosyjskich służb specjalnych w świetle afery Olega Kulinicza
Working methods of the Russian secret services in the light of the Oleg Kulinich case
Autorzy:
Świerczek, Marek
Powiązania:
https://bibliotekanauki.pl/articles/23379226.pdf
Data publikacji:
2023-12-06
Wydawca:
Agencja Bezpieczeństwa Wewnętrznego
Tematy:
FSB
SBU
WCzK
GPU
kontrwywiad ofensywny
dezorganizacja
infiltracja
konflikt rosyjsko-ukraiński
VChK
offensive counterintelligence
disorganisation
infiltration
Russian-Ukrainian conflict
Opis:
Autor dokonuje analizy przypadku zatrzymania oficera Służby Bezpieczeństwa Ukrainy Olega Kulinicza pod zarzutem szpiegostwa na rzecz Federalnej Służby Bezpieczeństwa Federacji Rosyjskiej. Na podstawie analizy zadań stawianych temu rosyjskiemu agentowi dochodzi do wniosku, że sposób działania rosyjskiego kontrwywiadu diametralnie różni się od metod zachodnich służb. Główną różnicą jest przeniesienie środka ciężkości działań operacyjnych z pracy rozpoznawczo-informacyjnej na próby agenturalnego przejęcia kontroli nad instytucjami przeciwnika, głównie cywilnymi i wojskowymi służbami specjalnymi, oraz na realizację infiltracji wywiadowczej przez ludzi dysponujących tym samym habitusem co kandydaci do werbunku. Korzystając z dorobku psychologii poznawczej oraz badań z zakresu historii najnowszej, autor udowadnia, że od ponad 100 lat rosyjskie służby wykorzystują i udoskonalają te metody.
The author analyses the case of the detention of Ukrainian Security Service officer Oleg Kulinich on suspicion of espionage for the Federal Security Service of the Russian Federation. On the basis of the analysis of the tasks posed to this Russian agent, he concludes that the modus operandi of Russian counterintelligence is diametrically opposed to the methods of Western services. The main difference is the shifting of the centre of gravity of operational activities from reconnaissance-information work to attempts at agentic seizure of control over enemy institutions, mainly civilian and military special services, and the realisation of intelligence infiltration by people with the same habitus as recruitment candidates. Drawing on the achievements of cognitive psychology and research in recent history, the author demonstrates that the Russian services have been using and refining these methods for more than 100 years.
Źródło:
Przegląd Bezpieczeństwa Wewnętrznego; 2023, 15, 29; 63-93
2080-1335
2720-0841
Pojawia się w:
Przegląd Bezpieczeństwa Wewnętrznego
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
Working methods of the Russian secret services in the light of the Oleg Kulinich case
Metody działania rosyjskich służb specjalnych w świetle afery Olega Kulinicza
Autorzy:
Świerczek, Marek
Powiązania:
https://bibliotekanauki.pl/articles/23378895.pdf
Data publikacji:
2023-12-06
Wydawca:
Agencja Bezpieczeństwa Wewnętrznego
Tematy:
FSB
SBU
VChK
GPU
oefnsive counterintelligence
disorganisation
infiltration
Russian-Ukrainian conflict
WCzK
kontrwywiad ofensywny
dezorganizacja
infiltracja
konflikt rosyjsko-ukraiński
Opis:
The author analyses the case of the detention of Ukrainian Security Service oficer Oleg Kulinich on suspicion of espionage for the Federal Security Service of the Russian Federation. On the basis of the analysis of the tasks posed to this Russian agent, he concludes that the modus operandi of Russian counterintelligence is diametrically opposed to the methods of Western services. The main diefrence is the shifting of the centre of gravity of operational activities from reconnaissance-information work to attempts at agentic seizure of control over enemy institutions, mainly civilian and military special services, and the realisation of intelligence infiltration by people with the same habitus as recruitment candidates. Drawing on the achievements of cognitive psychology and research in recent history, the author demonstrates that the Russian services have been using and refining these methods for more than 100 years.
Autor dokonuje analizy przypadku zatrzymania oficera Służby Bezpieczeństwa Ukrainy Olega Kulinicza pod zarzutem szpiegostwa na rzecz Federalnej Służby Bezpieczeństwa Federacji Rosyjskiej. Na podstawie analizy zadań stawianych temu rosyjskiemu agentowi dochodzi do wniosku, że sposób działania rosyjskiego kontrwywiadu diametralnie różni się od metod zachodnich służb. Główną różnicą jest przeniesienie środka ciężkości działań operacyjnych z pracy rozpoznawczo-informacyjnej na próby agenturalnego przejęcia kontroli nad instytucjami przeciwnika, głównie cywilnymi i wojskowymi służbami specjalnymi, oraz na realizację infiltracji wywiadowczej przez ludzi dysponujących tym samym habitusem co kandydaci do werbunku. Korzystając z dorobku psychologii po-znawczej oraz badań z zakresu historii najnowszej, autor udowadnia, że od ponad 100 lat rosyjskie służby wykorzystują i udoskonalają te metody.
Źródło:
Przegląd Bezpieczeństwa Wewnętrznego; 2023, 15, 29; 291-322
2080-1335
2720-0841
Pojawia się w:
Przegląd Bezpieczeństwa Wewnętrznego
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
A Novel FE/MC-based Mathematical Model of Mushy Steel Deformation with GPU Support
Autorzy:
Hojny, Marcin
Dębiński, Tomasz
Powiązania:
https://bibliotekanauki.pl/articles/2106557.pdf
Data publikacji:
2022
Wydawca:
Polska Akademia Nauk. Czytelnia Czasopism PAN
Tematy:
FEM
Monte Carlo
extra-high temperatures
soft-reduction
GPU
Opis:
The paper presents the results of work leading to the construction of a spatial hybrid model based on finite element (FE) and Monte Carlo (MC) methods allowing the computer simulation of physical phenomena accompanying the steel sample testing at temperatures that are characteristic for soft-reduction process. The proposed solution includes local density variations at the level of mechanical solution (the incompressibility condition was replaced with the condition of mass conservation), and at the same time simulates the grain growth in a comprehensive resistance heating process combined with a local remelting followed by free/controlled cooling of the sample tested. Simulation of grain growth in the entire computing domain would not be possible without the support of GPU processors. There was a 59-fold increase in the computing speed on the GPU compared to single-threaded computing on the CPU. The study was complemented by examples of experimental and computer simulation results, showing the correctness of the adopted model assumptions.
Źródło:
Archives of Metallurgy and Materials; 2022, 67, 2; 735--742
1733-3490
Pojawia się w:
Archives of Metallurgy and Materials
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
Execution time prediction model for parallel GPU realizations of discrete transforms computation algorithms
Autorzy:
Puchala, Dariusz
Stokfiszewski, Kamil
Wieloch, Kamil
Powiązania:
https://bibliotekanauki.pl/articles/2173537.pdf
Data publikacji:
2022
Wydawca:
Polska Akademia Nauk. Czytelnia Czasopism PAN
Tematy:
graphics processing unit
GPU
execution time prediction model
discrete wavelet transform
DWT
lattice structure
convolution-based approach
orthogonal transform
orthogonal filter banks
time effectiveness
prediction accuracy
procesor graficzny
model przewidywania czasu wykonania
dyskretna transformata falkowa
struktura sieciowa
podejście oparte na splotach
przekształcenia ortogonalne
ortogonalne banki filtrów
efektywność czasowa
dokładność przewidywania
Opis:
Parallel realizations of discrete transforms (DTs) computation algorithms (DTCAs) performed on graphics processing units (GPUs) play a significant role in many modern data processing methods utilized in numerous areas of human activity. In this paper the authors propose a novel execution time prediction model, which allows for accurate and rapid estimation of execution times of various kinds of structurally different DTCAs performed on GPUs of distinct architectures, without the necessity of conducting the actual experiments on physical hardware. The model can serve as a guide for the system analyst in making the optimal choice of the GPU hardware solution for a given computational task involving particular DT calculation, or can help in choosing the best appropriate parallel implementation of the selected DT, given the limitations imposed by available hardware. Restricting the model to exhaustively adhere only to the key common features of DTCAs enables the authors to significantly simplify its structure, leading consequently to its design as a hybrid, analytically–simulational method, exploiting jointly the main advantages of both of the mentioned techniques, namely: time-effectiveness and high prediction accuracy, while, at the same time, causing mutual elimination of the major weaknesses of both of the specified approaches within the proposed solution. The model is validated experimentally on two structurally different parallel methods of discrete wavelet transform (DWT) computation, i.e. the direct convolutionbased and lattice structure-based schemes, by comparing its prediction results with the actual measurements taken for 6 different graphics cards, representing a fairly broad spectrum of GPUs compute architectures. Experimental results reveal the overall average execution time and prediction accuracy of the model to be at a level of 97.2%, with global maximum prediction error of 14.5%, recorded throughout all the conducted experiments, maintaining at the same time high average evaluation speed of 3.5 ms for single simulation duration. The results facilitate inferring the model generality and possibility of extrapolation to other DTCAs and different GPU architectures, which along with the proposed model straightforwardness, time-effectiveness and ease of practical application, makes it, in the authors’ opinion, a very interesting alternative to the related existing solutions.
Źródło:
Bulletin of the Polish Academy of Sciences. Technical Sciences; 2022, 70, 1; e139393, 1--30
0239-7528
Pojawia się w:
Bulletin of the Polish Academy of Sciences. Technical Sciences
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
Execution time prediction model for parallel GPU realizations of discrete transforms computation algorithms
Autorzy:
Puchala, Dariusz
Stokfiszewski, Kamil
Wieloch, Kamil
Powiązania:
https://bibliotekanauki.pl/articles/2173635.pdf
Data publikacji:
2022
Wydawca:
Polska Akademia Nauk. Czytelnia Czasopism PAN
Tematy:
graphics processing unit
GPU
execution time prediction model
discrete wavelet transform
DWT
lattice structure
convolution-based approach
orthogonal transform
orthogonal filter banks
time effectiveness
prediction accuracy
procesor graficzny
model przewidywania czasu wykonania
dyskretna transformata falkowa
struktura sieciowa
podejście oparte na splotach
przekształcenia ortogonalne
ortogonalne banki filtrów
efektywność czasowa
dokładność przewidywania
Opis:
Parallel realizations of discrete transforms (DTs) computation algorithms (DTCAs) performed on graphics processing units (GPUs) play a significant role in many modern data processing methods utilized in numerous areas of human activity. In this paper the authors propose a novel execution time prediction model, which allows for accurate and rapid estimation of execution times of various kinds of structurally different DTCAs performed on GPUs of distinct architectures, without the necessity of conducting the actual experiments on physical hardware. The model can serve as a guide for the system analyst in making the optimal choice of the GPU hardware solution for a given computational task involving particular DT calculation, or can help in choosing the best appropriate parallel implementation of the selected DT, given the limitations imposed by available hardware. Restricting the model to exhaustively adhere only to the key common features of DTCAs enables the authors to significantly simplify its structure, leading consequently to its design as a hybrid, analytically–simulational method, exploiting jointly the main advantages of both of the mentioned techniques, namely: time-effectiveness and high prediction accuracy, while, at the same time, causing mutual elimination of the major weaknesses of both of the specified approaches within the proposed solution. The model is validated experimentally on two structurally different parallel methods of discrete wavelet transform (DWT) computation, i.e. the direct convolutionbased and lattice structure-based schemes, by comparing its prediction results with the actual measurements taken for 6 different graphics cards, representing a fairly broad spectrum of GPUs compute architectures. Experimental results reveal the overall average execution time and prediction accuracy of the model to be at a level of 97.2%, with global maximum prediction error of 14.5%, recorded throughout all the conducted experiments, maintaining at the same time high average evaluation speed of 3.5 ms for single simulation duration. The results facilitate inferring the model generality and possibility of extrapolation to other DTCAs and different GPU architectures, which along with the proposed model straightforwardness, time-effectiveness and ease of practical application, makes it, in the authors’ opinion, a very interesting alternative to the related existing solutions.
Źródło:
Bulletin of the Polish Academy of Sciences. Technical Sciences; 2022, 70, 1; art. no. e139393
0239-7528
Pojawia się w:
Bulletin of the Polish Academy of Sciences. Technical Sciences
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
Execution time prediction model for parallel GPU realizations of discrete transforms computation algorithms
Autorzy:
Puchala, Dariusz
Stokfiszewski, Kamil
Wieloch, Kamil
Powiązania:
https://bibliotekanauki.pl/articles/2173530.pdf
Data publikacji:
2022
Wydawca:
Polska Akademia Nauk. Czytelnia Czasopism PAN
Tematy:
graphics processing unit
GPU
execution time prediction model
discrete wavelet transform
DWT
lattice structure
convolution-based approach
orthogonal transform
orthogonal filter banks
time effectiveness
prediction accuracy
Opis:
Parallel realizations of discrete transforms (DTs) computation algorithms (DTCAs) performed on graphics processing units (GPUs) play a significant role in many modern data processing methods utilized in numerous areas of human activity. In this paper the authors propose a novel execution time prediction model, which allows for accurate and rapid estimation of execution times of various kinds of structurally different DTCAs performed on GPUs of distinct architectures, without the necessity of conducting the actual experiments on physical hardware. The model can serve as a guide for the system analyst in making the optimal choice of the GPU hardware solution for a given computational task involving particular DT calculation, or can help in choosing the best appropriate parallel implementation of the selected DT, given the limitations imposed by available hardware. Restricting the model to exhaustively adhere only to the key common features of DTCAs enables the authors to significantly simplify its structure, leading consequently to its design as a hybrid, analytically–simulational method, exploiting jointly the main advantages of both of the mentioned techniques, namely: time-effectiveness and high prediction accuracy, while, at the same time, causing mutual elimination of the major weaknesses of both of the specified approaches within the proposed solution. The model is validated experimentally on two structurally different parallel methods of discrete wavelet transform (DWT) computation, i.e. the direct convolutionbased and lattice structure-based schemes, by comparing its prediction results with the actual measurements taken for 6 different graphics cards, representing a fairly broad spectrum of GPUs compute architectures. Experimental results reveal the overall average execution time and prediction accuracy of the model to be at a level of 97.2%, with global maximum prediction error of 14.5%, recorded throughout all the conducted experiments, maintaining at the same time high average evaluation speed of 3.5 ms for single simulation duration. The results facilitate inferring the model generality and possibility of extrapolation to other DTCAs and different GPU architectures, which along with the proposed model straightforwardness, time-effectiveness and ease of practical application, makes it, in the authors’ opinion, a very interesting alternative to the related existing solutions.
Źródło:
Bulletin of the Polish Academy of Sciences. Technical Sciences; 2022, 70, 1; e139393, 1--30
0239-7528
Pojawia się w:
Bulletin of the Polish Academy of Sciences. Technical Sciences
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
Execution time prediction model for parallel GPU realizations of discrete transforms computation algorithms
Autorzy:
Puchala, Dariusz
Stokfiszewski, Kamil
Wieloch, Kamil
Powiązania:
https://bibliotekanauki.pl/articles/2173636.pdf
Data publikacji:
2022
Wydawca:
Polska Akademia Nauk. Czytelnia Czasopism PAN
Tematy:
graphics processing unit
GPU
execution time prediction model
discrete wavelet transform
DWT
lattice structure
convolution-based approach
orthogonal transform
orthogonal filter banks
time effectiveness
prediction accuracy
procesor graficzny
model przewidywania czasu wykonania
dyskretna transformata falkowa
struktura sieciowa
podejście oparte na splotach
przekształcenia ortogonalne
ortogonalne banki filtrów
efektywność czasowa
dokładność przewidywania
Opis:
Parallel realizations of discrete transforms (DTs) computation algorithms (DTCAs) performed on graphics processing units (GPUs) play a significant role in many modern data processing methods utilized in numerous areas of human activity. In this paper the authors propose a novel execution time prediction model, which allows for accurate and rapid estimation of execution times of various kinds of structurally different DTCAs performed on GPUs of distinct architectures, without the necessity of conducting the actual experiments on physical hardware. The model can serve as a guide for the system analyst in making the optimal choice of the GPU hardware solution for a given computational task involving particular DT calculation, or can help in choosing the best appropriate parallel implementation of the selected DT, given the limitations imposed by available hardware. Restricting the model to exhaustively adhere only to the key common features of DTCAs enables the authors to significantly simplify its structure, leading consequently to its design as a hybrid, analytically–simulational method, exploiting jointly the main advantages of both of the mentioned techniques, namely: time-effectiveness and high prediction accuracy, while, at the same time, causing mutual elimination of the major weaknesses of both of the specified approaches within the proposed solution. The model is validated experimentally on two structurally different parallel methods of discrete wavelet transform (DWT) computation, i.e. the direct convolutionbased and lattice structure-based schemes, by comparing its prediction results with the actual measurements taken for 6 different graphics cards, representing a fairly broad spectrum of GPUs compute architectures. Experimental results reveal the overall average execution time and prediction accuracy of the model to be at a level of 97.2%, with global maximum prediction error of 14.5%, recorded throughout all the conducted experiments, maintaining at the same time high average evaluation speed of 3.5 ms for single simulation duration. The results facilitate inferring the model generality and possibility of extrapolation to other DTCAs and different GPU architectures, which along with the proposed model straightforwardness, time-effectiveness and ease of practical application, makes it, in the authors’ opinion, a very interesting alternative to the related existing solutions.
Źródło:
Bulletin of the Polish Academy of Sciences. Technical Sciences; 2022, 70, 1; art. no. e139393
0239-7528
Pojawia się w:
Bulletin of the Polish Academy of Sciences. Technical Sciences
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
GPU implementation of atomic fluid MD simulation.
Autorzy:
Dawid, Aleksander
Powiązania:
https://bibliotekanauki.pl/articles/2197547.pdf
Data publikacji:
2022
Wydawca:
Politechnika Gdańska
Tematy:
MD simulabon
GPU
atomic fluid
MD parallel algorithm
Opis:
A computer simulation of an atomic fluid on a GPU was implemented using the CUDA architecture. It was shown that the programming model for efficient numerical computing applications was changing with the development of the CUDA architecture. The introduction of the L2 cache decreased the latency between the global GPU memory and the registers. The performed MD simulation using the global memory and registers showed that the average acceleration relative to the CPU reached 80 times for single-precision calculations. Usually, the shared block memory gives much better results for this kind of calculation. We have found that using the shared memory gives acceleration over 116 times in comparison to the CPU. It is about 49% faster than using the global memory and registers. It is shown here that the performance of generally available graphics cards for double-precision calculations is significantly lower than for single-precision calculations. The recorded double-precision acceleration relative to the CPU in our experiment averaged 6 and 7 times for the global and shared memory, respectively. We performed these calculations on two different CUDA enable device systems.
Źródło:
TASK Quarterly. Scientific Bulletin of Academic Computer Centre in Gdansk; 2022, 26, 1; 25-37
1428-6394
Pojawia się w:
TASK Quarterly. Scientific Bulletin of Academic Computer Centre in Gdansk
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
Oskarżenie o współpracę z wywiadem polskim jako narzędzie GPU/NKWD USRS w czystkach w Armii Sowieckiej w latach dwudziestych i trzydziestych XX wieku. Przypadek Henryka Wieczffińskiego i polski wątek sprawy kryptonim „Wiesna”
Accusation of Collaboration with Polish Intelligence as a Tool of the GPU/NKVD of the Ukrainian SSR in the Soviet Army Purges in the 1920s and 1930s. The Case of Henryk Wieczffiński and a Polish Theme in the "Vesna" Case
Autorzy:
Kokin, Serhii
Powiązania:
https://bibliotekanauki.pl/articles/20265847.pdf
Data publikacji:
2022
Wydawca:
Instytut Pamięci Narodowej, Komisja Ścigania Zbrodni przeciwko Narodowi Polskiemu
Tematy:
Henryk Wieczffiński
Józef Piłsudski
Ukrainian Soviet Socialist Republic
OGPU
GPU of Ukraine
"Vesna" case
Ukraińska Socjalistyczna Republika Sowiecka
GPU Ukrainy
sprawa „Wiesna”
Opis:
Porażka Sowietów w wojnie polsko-bolszewickiej w 1920 r. oznaczała upadek ich nadziei na szybki i udany eksport rewolucji proletariackiej do Europy Środkowej i Zachodniej. Symbolem tej ciężkiej klęski geopolitycznej na wiele lat stał się dla bolszewików Józef Piłsudski, toteż walka z jego dziedzictwem w Ukraińskiej Socjalistycznej Republice Sowieckiej (USRS) należała do najważniejszych zadań komunistycznych służb bezpieczeństwa w latach dwudziestych i trzydziestych XX w. Pokazowe pod tym względem były liczne rozpracowania operacyjne i duże grupowe sprawy karne wszczęte w tym okresie przez organy OGPU, w taki czy inny sposób zawsze bowiem był w nich obecny „polski ślad”. Przykładem takiej działalności jest likwidacja przez GPU Ukrainy Wszechzwiązkowej Wojskowo-Oficerskiej Organizacji Kontrrewolucyjnej (sprawa „Wiesna”), w której ramach został aresztowany, a następnie rozstrzelany (12 lipca 1931 r. w Charkowie) por. Henryk Wieczffiński, lotnik obserwator, szef sztabu 20 Brygady Lotniczej Ukraińskiego Okręgu Wojskowego.
The defeat of the Soviets in the Polish-Bolshevik War in 1920 marked the collapse of their hopes for the rapid and successful export of the proletarian revolution to Central and Western Europe. Józef Piłsudski became a long-time symbol of this grave geopolitical defeat for the Bolsheviks, so the fight against his legacy in the Ukrainian Soviet Socialist Republic (USSR) was one of the most important tasks of the Communist security services in the 1920s and 1930s. Demonstrative in this respect were the numerous operational examinations and large collective criminal cases initiated during this period by the OGPU authorities, for in one way or another there was always a "Polish trace" present. An example of such activity is the liquidation of the All-Union Military-Officer Counterrevolutionary Organisation by the GPU of Ukraine (the "Vesna" case), as part of which Lt. Henryk Wieczffiński, an observer pilot, chief of staff of the 20th Aviation Brigade of the Ukrainian Military District, was arrested and subsequently executed (12 July 1931 in Kharkov).
Źródło:
Aparat Represji w Polsce Ludowej 1944–1989; 2022, 20; 375-392
1733-6996
Pojawia się w:
Aparat Represji w Polsce Ludowej 1944–1989
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
An optimized parallel implementation of non-iteratively trained recurrent neural networks
Autorzy:
El Zini, Julia
Rizk, Yara
Awad, Mariette
Powiązania:
https://bibliotekanauki.pl/articles/2031147.pdf
Data publikacji:
2021
Wydawca:
Społeczna Akademia Nauk w Łodzi. Polskie Towarzystwo Sieci Neuronowych
Tematy:
GPU implementation
parallelization
Recurrent Neural Network
RNN
Long-short Term Memory
LSTM
Gated Recurrent Unit
GRU
Extreme Learning Machines
ELM
non-iterative training
Opis:
Recurrent neural networks (RNN) have been successfully applied to various sequential decision-making tasks, natural language processing applications, and time-series predictions. Such networks are usually trained through back-propagation through time (BPTT) which is prohibitively expensive, especially when the length of the time dependencies and the number of hidden neurons increase. To reduce the training time, extreme learning machines (ELMs) have been recently applied to RNN training, reaching a 99% speedup on some applications. Due to its non-iterative nature, ELM training, when parallelized, has the potential to reach higher speedups than BPTT. In this work, we present Opt-PR-ELM, an optimized parallel RNN training algorithm based on ELM that takes advantage of the GPU shared memory and of parallel QR factorization algorithms to efficiently reach optimal solutions. The theoretical analysis of the proposed algorithm is presented on six RNN architectures, including LSTM and GRU, and its performance is empirically tested on ten time-series prediction applications. Opt- PR-ELM is shown to reach up to 461 times speedup over its sequential counterpart and to require up to 20x less time to train than parallel BPTT. Such high speedups over new generation CPUs are extremely crucial in real-time applications and IoT environments.
Źródło:
Journal of Artificial Intelligence and Soft Computing Research; 2021, 11, 1; 33-50
2083-2567
2449-6499
Pojawia się w:
Journal of Artificial Intelligence and Soft Computing Research
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
Performance enhancement of CUDA applications by overlapping data transfer and Kernel execution
Autorzy:
Raju, K.
Chiplunkar, Niranjan N
Powiązania:
https://bibliotekanauki.pl/articles/1956064.pdf
Data publikacji:
2021
Wydawca:
Polskie Towarzystwo Promocji Wiedzy
Tematy:
CPU-GPU
high-performance computing
kernel
data transfer
CUDA streams
obliczenia wysokiej wydajności
jądro
transfer danych
strumienie CUDA
Opis:
The CPU-GPU combination is a widely used heterogeneous computing system in which the CPU and GPU have different address spaces. Since the GPU cannot directly access the CPU memory, prior to invoking the GPU function the input data must be available on the GPU memory. On completion of GPU function, the results of computation are transferred to CPU memory. The CPU-GPU data transfer happens through PCIExpress bus. The PCI-E bandwidth is much lesser than that of GPU memory. The speed at which the data is transferred is limited by the PCI-E bandwidth. Hence, the PCI-E acts as a performance bottleneck. In this paper two approaches are discussed to minimize the overhead of data transfer, namely, performing the data transfer while the GPU function is being executed and reducing the amount of data to be transferred to GPU. The effectiveness of these approaches on the execution time of a set of CUDA applications is realized using CUDA streams. The results of our experiments show that the execution time of applications can be minimized with the proposed approaches.
Źródło:
Applied Computer Science; 2021, 17, 3; 5-18
1895-3735
Pojawia się w:
Applied Computer Science
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
Fenomen sowieckich służb specjalnych
The phenomens of the soviet special services
Autorzy:
Świerczek, Marek
Powiązania:
https://bibliotekanauki.pl/articles/121538.pdf
Data publikacji:
2020
Wydawca:
Towarzystwo Wiedzy Obronnej
Tematy:
sowieckie służby specjalne
WCzK
GPU
OGPU
Ochrana
instrumentarium pracy operacyjnej
terror
infiltracja
prowokacja
dezinformacja
Soviet special services
VChK
Okhrana
operational work instruments
infiltration
provocation
disinformation
Opis:
Autor artykułu, analizując fenomen skuteczności sowieckich służb specjalnych, wysuwa hipotezę, że była ona skutkiem splotu kilku czynników. Wśród głównych elementów składających się na niebywałą sprawność służb sowieckich w pierwszej dekadzie po przewrocie październikowym, zdaniem autora znajdują się: fuzja doświadczeń instytucjonalnych Ochrany z wiedzą środowisk kryminalno-rewolucyjnych tworzących WCzK oraz rozwiązania operacyjne niespotykane wśród cywilizowanych społeczeństw. Zdaniem autora, sowieckie służby zostały de facto wyjęte spod prawa i nakazów moralności (uznawanych za przeżytek burżuazyjny), co umożliwiło stosowanie metod i form pracy niemożliwych w Europie Zachodniej. Tak rozumiane instrumentarium sowieckich służb składało się z mieszanki terroru, infiltracji, prowokacji i dezinformacji.
The article author, analyzing the phenomenon of the effectiveness of the Soviet special services, puts forward the hypothesis that it was the result of a combination of several factors. Among the main elements that made up the incredible efficiency of the Soviet services in the first decade after the October coup, according to the author there were: a fusion of institutional experiences of the Okhrana with the knowledge of criminal-revolutionary environments creating the VChK and operational solutions unheard of in civilized societies. According to the author, the Soviet services acted under no law and moral obligations (considered by revolutionaries as bourgeois remnant), which allowed the use of methods and forms of operational work unthinkable in Western Europe. The instruments of the Soviet services consisted of a mixture of terror, infiltration, provocation and disinformation.
Źródło:
Wiedza Obronna; 2020, 2; 63-74
0209-0031
2658-0829
Pojawia się w:
Wiedza Obronna
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
„Płonąca granica”. Działalność szpiegowska w kontekście nielegalnych migracji na wschodniej granicy II Rzeczypospolitej (na przykładzie województwa tarnopolskiego)
“A Border in Flames”. Espionage in the Context of Illegal Migrations along the Eastern Border of the Second Republic of Poland (Based on the Example of the Tarnopolskie Voivodeship)
Autorzy:
Humennyi, Serhii
Powiązania:
https://bibliotekanauki.pl/articles/478484.pdf
Data publikacji:
2019
Wydawca:
Instytut Pamięci Narodowej, Komisja Ścigania Zbrodni przeciwko Narodowi Polskiemu
Tematy:
spying and sabotage activities
illegal migration
smuggling
USRR
Second Republic of Poland
Tarnopolskie Voivodeship
Soviet and Polish border
State Political Board (GPU)
Border Protection Corps (BPC)
Opis:
Z jawisko działalności szpiegowskiej w kontekście nielegalnych migracji na terenach przygranicznych II Rzeczypospolitej pozostaje nadal mało zbadane. Celem niniejszego artykułu jest analiza procederu nielegalnego przekraczania granicy państwowej i wyjaś- nienie wpływu tego zjawiska na sytuację polityczną i gospodarczą w przygranicznym pasie województwa tarnopolskiego, a ściślej w jego powiatach: borszczowskim, czortkow- skim, kopyczynieckim, skałackim i zbaraskim, w latach 1921–1939. Najprawdopodobniej jest to pierwsza próba przybliżenia zależności między stanem ochrony granicy pań- stwowej, konfliktem ukraińsko-polskim w warunkach istnienia państwa polskiego i działalnością dywersyjno-wywiadowczą władz radzieckich na obszarze jednego ze wschodnich województw ІІ Rzeczypospolitej.
T he phenomenon of espionage in the context of illegal migrations in the borderlands of the Second Republic of Poland has not been thoroughly examined so far. This article aims to analyse the procedure of crossing the state border illegally and explaining the impact of the phenomenon on the political and economic situation in the borderlands of the Tarnopolskie Voivodeship, specifically in the Poviats of Borszczów, Czortków, Kopyczyńce, Skałat and Zbaraż in the years 1921–1939. This it most probably the first attempt to explain the relationship between the status of the protection of the state bor- der, the Ukrainian and Polish conflict during the existence of the Polish state and the sabotage and intelligence activities of the Soviet authorities in the territories of one of the Eastern voivodeships of the Second Republic of Poland.
Źródło:
Pamięć i Sprawiedliwość; 2019, 33; 327-344
1427-7476
Pojawia się w:
Pamięć i Sprawiedliwość
Dostawca treści:
Biblioteka Nauki
Artykuł
Tytuł:
Exploiting multi-core and many-core parallelism for subspace clustering
Autorzy:
Datta, Amitava
Kaur, Amardeep
Lauer, Tobias
Chabbouh, Sami
Powiązania:
https://bibliotekanauki.pl/articles/331126.pdf
Data publikacji:
2019
Wydawca:
Uniwersytet Zielonogórski. Oficyna Wydawnicza
Tematy:
data mining
subspace clustering
multicore processor
many core processor
GPU computing
eksploracja danych
procesor wielordzeniowy
obliczenia GPU
Opis:
Finding clusters in high dimensional data is a challenging research problem. Subspace clustering algorithms aim to find clusters in all possible subspaces of the dataset, where a subspace is a subset of dimensions of the data. But the exponential increase in the number of subspaces with the dimensionality of data renders most of the algorithms inefficient as well as ineffective. Moreover, these algorithms have ingrained data dependency in the clustering process, which means that parallelization becomes difficult and inefficient. SUBSCALE is a recent subspace clustering algorithm which is scalable with the dimensions and contains independent processing steps which can be exploited through parallelism. In this paper, we aim to leverage the computational power of widely available multi-core processors to improve the runtime performance of the SUBSCALE algorithm. The experimental evaluation shows linear speedup. Moreover, we develop an approach using graphics processing units (GPUs) for fine-grained data parallelism to accelerate the computation further. First tests of the GPU implementation show very promising results.
Źródło:
International Journal of Applied Mathematics and Computer Science; 2019, 29, 1; 81-91
1641-876X
2083-8492
Pojawia się w:
International Journal of Applied Mathematics and Computer Science
Dostawca treści:
Biblioteka Nauki
Artykuł

Ta witryna wykorzystuje pliki cookies do przechowywania informacji na Twoim komputerze. Pliki cookies stosujemy w celu świadczenia usług na najwyższym poziomie, w tym w sposób dostosowany do indywidualnych potrzeb. Korzystanie z witryny bez zmiany ustawień dotyczących cookies oznacza, że będą one zamieszczane w Twoim komputerze. W każdym momencie możesz dokonać zmiany ustawień dotyczących cookies