Zobrazeno 1 - 10
of 26
pro vyhledávání: '"Eiman Ebrahimi"'
Autor:
Zaid Qureshi, Vikram Sharma Mailthody, Wen-mei W. Hwu, Jinjun Xiong, Eiman Ebrahimi, Seungwon Min
Publikováno v:
Proceedings of the VLDB Endowment. 14:114-127
Modern analytics and recommendation systems are increasingly based on graph data that capture the relations between entities being analyzed. Practical graphs come in huge sizes, offer massive parallelism, and are stored in sparse-matrix formats such
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::3f2da832b6ef0cd6a1730d2e63a31189
https://doi.org/10.14778/3425879.3425883
https://doi.org/10.14778/3425879.3425883
Autor:
Benjamin Klenk, Madeleine Glick, Eiman Ebrahimi, Mehrdad Khani, Manya Ghobadi, Ziyi Zhu, Mohammad Alizadeh, Keren Bergman, Amin Vahdat
Publikováno v:
SIGCOMM
This paper proposes optical network interconnects as a key enabler for building high-bandwidth ML training clusters with strong scaling properties. Our design, called SiP-ML, accelerates the training time of popular DNN models using silicon photonics
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::bde66cf5e93c80306d4a9ce2f110011a
https://doi.org/10.1145/3452296.3472900
https://doi.org/10.1145/3452296.3472900
Autor:
Arslan Zulfiqar, David Nellans, Victor Zhang, Eiman Ebrahimi, Yaosheng Fu, Saptadeep Pal, Szymon Migacz, Puneet Gupta
Publikováno v:
IEEE Micro. 39:91-101
Deploying deep learning (DL) models across multiple compute devices to train large and complex models continues to grow in importance because of the demand for faster and more frequent training. Data parallelism (DP) is the most widely used paralleli
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::b5e5cf72396926de185947f87ea35c17
https://doi.org/10.1109/mm.2019.2935967
https://doi.org/10.1109/mm.2019.2935967
Publikováno v:
ACM Transactions on Architecture and Code Optimization. 16:1-24
Conventional on-chip TLB hierarchies are unable to fully cover the growing application working-set sizes. To make things worse, Last-Level TLB (LLT) misses require multiple accesses to the page table even with the use of page walk caches. Consequentl
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::a5578c0309fd683b8b20e3bc7ce509eb
https://doi.org/10.1145/3309710
https://doi.org/10.1145/3309710
Autor:
Kun Wu, Wen-mei W. Hwu, Eiman Ebrahimi, Mert Hidayetoglu, Seungwon Min, Sitao Huang, Deming Chen, Jinjun Xiong
Graph Convolutional Networks (GCNs) are increasingly adopted in large-scale graph-based recommender systems. Training GCN requires the minibatch generator traversing graphs and sampling the sparsely located neighboring nodes to obtain their features.
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::1a09b0218295585f51b97fbd4152ca6c
Planaria: Dynamic Architecture Fission for Spatial Multi-Tenant Acceleration of Deep Neural Networks
Autor:
Hadi Esmaeilzadeh, Mohammad Alian, Eiman Ebrahimi, Sean Kinzer, Navateja Alla, Joon Kyung Kim, Brahmendra Reddy Yatham, Byung Hoon Ahn, Cliff Young, Soroush Ghodrati, Hardik Sharma, Nam Sung Kim
Publikováno v:
MICRO
Deep Neural Networks (DNNs) have reinvigorated real-world applications that rely on learning patterns of data and are permeating into different industries and markets. Cloud infrastructure and accelerators that offer INFerence-as-a-Service (INFaaS) h
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::991fd19948d5b55726891c0e4651f829
https://doi.org/10.1109/micro50266.2020.00062
https://doi.org/10.1109/micro50266.2020.00062
Autor:
Vikram Sharma Mailthody, Carl Pearson, Mert Hidayetoglu, Wen-mei W. Hwu, Eiman Ebrahimi, Jinjun Xiong, Rakesh Nagi
Publikováno v:
HPEC
This paper presents GPU performance optimization and scaling results for inference models of the Sparse Deep Neural Network Challenge 2020. Demands for network quality have increased rapidly, pushing the size and thus the memory requirements of many
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::fd2daa876b1cbd0f469f48aca37e021c
http://arxiv.org/abs/2007.14152
http://arxiv.org/abs/2007.14152
Autor:
Onur Mutlu, Niladrish Chatterjee, Stephen W. Keckler, Gwangsun Kim, Mike O'Connor, Eiman Ebrahimi, Kevin Hsieh, Nandita Vijaykumar
Publikováno v:
ISCA
Main memory bandwidth is a critical bottleneck for modern GPU systems due to limited off-chip pin bandwidth. 3D-stacked memory architectures provide a promising opportunity to significantly alleviate this bottleneck by directly connecting a logic lay
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::3616917e8041cdaf787718e56bbf8d99
https://doi.org/10.1145/3007787.3001159
https://doi.org/10.1145/3007787.3001159
Publikováno v:
MICRO
Historically, improvement in GPU performance has been tightly coupled with transistor scaling. As Moore's Law slows down, performance of single GPUs may ultimately plateau. To continue GPU performance scaling, multiple GPUs can be connected using sys
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::ea3fd042e610d5acac011c3433b99ec6
https://doi.org/10.1109/micro.2018.00035
https://doi.org/10.1109/micro.2018.00035
Autor:
Onur Mutlu, Abhilasha Jain, Phillip B. Gibbons, Eiman Ebrahimi, Nandita Vijaykumar, Gennady Pekhimenko, Nastaran Hajinazar, Diptesh Majumdar, Kevin Hsieh
Publikováno v:
ISCA
This paper makes a case for a new cross-layer interface, Expressive Memory (XMem), to communicate higher-level program semantics from the application to the system software and hardware architecture. XMem provides (i) a flexible and extensible abstra
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::ffbdcf3a77c39c9df80772e9b62ee285
https://doi.org/10.1109/isca.2018.00027
https://doi.org/10.1109/isca.2018.00027