Pregled bibliografske jedinice broj: 1209236
Batched matrix operations on distributed GPUs with application in theoretical physics
Batched matrix operations on distributed GPUs with application in theoretical physics // Proceedings of 45th Jubiilee International Convention on Information, Communication and Electronic Technology (MIPRO)
Opatija, Hrvatska: Institute of Electrical and Electronics Engineers (IEEE), 2022. str. 293-299 doi:10.23919/mipro55190.2022.9803591 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 1209236 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Batched matrix operations on distributed GPUs with application in theoretical physics
Autori
Mijić, Nenad ; Davidovic, Davor
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
Proceedings of 45th Jubiilee International Convention on Information, Communication and Electronic Technology (MIPRO)
/ - : Institute of Electrical and Electronics Engineers (IEEE), 2022, 293-299
ISBN
978-953-233-103-5
Skup
45th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO 2022)
Mjesto i datum
Opatija, Hrvatska, 23.05.2022. - 27.05.2022
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
matrix multiplication ; batched operations ; GPU ; MPI ; HPC
Sažetak
One of the most important and commonly used operations in many linear algebra functions is matrix-matrix multiplication (GEMM), which is also a key component in obtaining high performance of many scientific codes. It is a computationally intensive function requiring O(n3) operations, and its high computational intensity makes it well-suited to be significantly accelerated with GPUs. Today, many research problems require solving a very large number of relatively small GEMM operations that cannot utilise the entire GPU. To overcome this bottleneck, special functions have been developed that pack several GEMM operations into one and then compute them simultaneously on a GPU, which is called a batch operation. In this research work, we have proposed a different approach based on linking multiple GEMM operations to Message Passing Interface (MPI) processes and then binding multiple MPI processes to a single GPU. To increase GPU utilisation, more MPI processes (i.e. GEMM operations) are added. We implement and test this approach in the field of theoretical physics to compute entanglement properties through simulated annealing Monte Carlo simulation of quantum spin chains. For the specific use case, we were able to simulate a much larger spin system and achieve a speedup of up to 35× compared to the parallel CPU-only version.
Izvorni jezik
Engleski
Znanstvena područja
Matematika, Računarstvo
POVEZANOST RADA
Projekti:
--UIP-2020-02-4559 - Skalabilni algoritmi visokih performansi za buduće heterogene distribuirane računalne sustave (HybridScale) (Davidović, Davor) ( CroRIS)
Ustanove:
Institut "Ruđer Bošković", Zagreb
Profili:
Davor Davidović
(autor)