Zobrazeno 1 - 10
of 220
pro vyhledávání: '"Gomez-Luna P"'
Autor:
Yuksel, Ismail Emir, Tugrul, Yahya Can, Bostanci, F. Nisa, Oliveira, Geraldo F., Yaglikci, A. Giray, Olgun, Ataberk, Soysal, Melina, Luo, Haocong, Gómez-Luna, Juan, Sadrosadati, Mohammad, Mutlu, Onur
We experimentally analyze the computational capability of commercial off-the-shelf (COTS) DRAM chips and the robustness of these capabilities under various timing delays between DRAM commands, data patterns, temperature, and voltage levels. We extens
Externí odkaz:
http://arxiv.org/abs/2405.06081
Autor:
Gogineni, Kailash, Dayapule, Sai Santosh, Gómez-Luna, Juan, Gogineni, Karthikeya, Wei, Peng, Lan, Tian, Sadrosadati, Mohammad, Mutlu, Onur, Venkataramani, Guru
Reinforcement Learning (RL) trains agents to learn optimal behavior by maximizing reward signals from experience datasets. However, RL training often faces memory limitations, leading to execution latencies and prolonged training times. To overcome t
Externí odkaz:
http://arxiv.org/abs/2405.03967
Autor:
Rhyner, Steve, Luo, Haocong, Gómez-Luna, Juan, Sadrosadati, Mohammad, Jiang, Jiawei, Olgun, Ataberk, Gupta, Harshita, Zhang, Ce, Mutlu, Onur
Modern Machine Learning (ML) training on large-scale datasets is a very time-consuming workload. It relies on the optimization algorithm Stochastic Gradient Descent (SGD) due to its effectiveness, simplicity, and generalization performance. Processor
Externí odkaz:
http://arxiv.org/abs/2404.07164
Processing-using-DRAM (PUD) architectures impose a restrictive data layout and alignment for their operands, where source and destination operands (i) must reside in the same DRAM subarray (i.e., a group of DRAM rows sharing the same row buffer and r
Externí odkaz:
http://arxiv.org/abs/2403.04539
Autor:
Oliveira, Geraldo F., Olgun, Ataberk, Yağlıkçı, Abdullah Giray, Bostancı, F. Nisa, Gómez-Luna, Juan, Ghose, Saugata, Mutlu, Onur
Processing-using-DRAM (PUD) is a processing-in-memory (PIM) approach that uses a DRAM array's massive internal parallelism to execute very-wide data-parallel operations, in a single-instruction multiple-data (SIMD) fashion. However, DRAM rows' large
Externí odkaz:
http://arxiv.org/abs/2402.19080
Autor:
Yuksel, Ismail Emir, Tugrul, Yahya Can, Olgun, Ataberk, Bostanci, F. Nisa, Yaglikci, A. Giray, Oliveira, Geraldo F., Luo, Haocong, Gómez-Luna, Juan, Sadrosadati, Mohammad, Mutlu, Onur
Processing-using-DRAM (PuD) is an emerging paradigm that leverages the analog operational properties of DRAM circuitry to enable massively parallel in-DRAM computation. PuD has the potential to reduce or eliminate costly data movement between process
Externí odkaz:
http://arxiv.org/abs/2402.18736
Publikováno v:
Tecnología y ciencias del agua, Vol 15, Iss 6, Pp 255-310 (2024)
En la cuenca Guantánamo-Guaso, ubicada en la región oriental de Cuba, los recursos hídricos están sometidos a fuertes presiones debido al desarrollo poblacional, agropecuario e industrial, constatándose el deterioro del ecosistema con la consigu
Externí odkaz:
https://doaj.org/article/aca5bfd87a7942e4a3500991ebb4e015
To ease the programmability of PIM architectures, we propose DaPPA(data-parallel processing-in-memory architecture), a framework that can, for a given application, automatically distribute input and gather output data, handle memory management, and p
Externí odkaz:
http://arxiv.org/abs/2310.10168
Data movement between memory and processors is a major bottleneck in modern computing systems. The processing-in-memory (PIM) paradigm aims to alleviate this bottleneck by performing computation inside memory chips. Real PIM hardware (e.g., the UPMEM
Externí odkaz:
http://arxiv.org/abs/2310.01893
Computing on encrypted data is a promising approach to reduce data security and privacy risks, with homomorphic encryption serving as a facilitator in achieving this goal. In this work, we accelerate homomorphic operations using the Processing-in- Me
Externí odkaz:
http://arxiv.org/abs/2309.06545