Zobrazeno 1 - 10
of 93
pro vyhledávání: '"Jung, Myoungsoo"'
We present GraphTensor, a comprehensive open-source framework that supports efficient parallel neural network processing on large graphs. GraphTensor offers a set of easy-to-use programming primitives that appreciate both graph and neural network exe
Externí odkaz:
http://arxiv.org/abs/2305.17469
Supporting atomic durability of updates for persistent memories is typically achieved with Write-Ahead Logging (WAL). WAL flushes log entries to persistent memory before making the actual data persistent to ensure that a consistent state can be recov
Externí odkaz:
http://arxiv.org/abs/2302.13394
This paper proposes TRAININGCXL that can efficiently process large-scale recommendation datasets in the pool of disaggregated memory while making training fault tolerant with low overhead. To this end, i) we integrate persistent memory (PMEM) and GPU
Externí odkaz:
http://arxiv.org/abs/2301.07492
Graph neural networks (GNNs) process large-scale graphs consisting of a hundred billion edges. In contrast to traditional deep learning, unique behaviors of the emerging GNNs are engaged with a large set of graphs and embedding data on storage, which
Externí odkaz:
http://arxiv.org/abs/2201.09189
Autor:
Zhang, Jie, Jung, Myoungsoo
Traditional graphics processing units (GPUs) suffer from the low memory capacity and demand for high memory bandwidth. To address these challenges, we propose Ohm-GPU, a new optical network based heterogeneous memory design for GPUs. Specifically, Oh
Externí odkaz:
http://arxiv.org/abs/2109.05430
Autor:
Zhang, Jie, Kwon, Miryeong, Gouk, Donghyun, Koh, Sungjoon, Kim, Nam Sung, Kandemir, Mahmut Taylan, Jung, Myoungsoo
Large persistent memories such as NVDIMM have been perceived as a disruptive memory technology, because they can maintain the state of a system even after a power failure and allow the system to recover quickly. However, overheads incurred by a heavy
Externí odkaz:
http://arxiv.org/abs/2106.14241
Host-side page victimizations can easily overflow the SSD internal buffer, which interferes I/O services of diverse user applications thereby degrading user-level experiences. To address this, we propose FastDrain, a co-design of OS kernel and flash
Externí odkaz:
http://arxiv.org/abs/2006.08966
Autor:
Zhang, Jie, Jung, Myoungsoo
We propose ZnG, a new GPU-SSD integrated architecture, which can maximize the memory capacity in a GPU and address performance penalties imposed by an SSD. Specifically, ZnG replaces all GPU internal DRAMs with an ultra-low-latency SSD to maximize th
Externí odkaz:
http://arxiv.org/abs/2006.08975
Emerging storage systems with new flash exhibit ultra-low latency (ULL) that can address performance disparities between DRAM and conventional solid state drives (SSDs) in the memory hierarchy. Considering the advanced low-latency characteristics, di
Externí odkaz:
http://arxiv.org/abs/1912.06998
Autor:
Koh, Sungjoon, Zhang, Jie, Kwon, Miryeong, Yoon, Jungyeon, Donofrio, David, Kim, Nam Sung, Jung, Myoungsoo
Publikováno v:
IEEE Transactions on Parallel and Distributed Systems ( Volume: 30 , Issue: 6 , June 1 2019 )
Large-scale systems with all-flash arrays have become increasingly common in many computing segments. To make such systems resilient, we can adopt erasure coding such as Reed-Solomon (RS) code as an alternative to replication because erasure coding i
Externí odkaz:
http://arxiv.org/abs/1906.08602