Zobrazeno 1 - 10
of 110
pro vyhledávání: '"Kandemir, Mahmut T."'
Autor:
Jain, Rishabh, Bhasi, Vivek M., Jog, Adwait, Sivasubramaniam, Anand, Kandemir, Mahmut T., Das, Chita R.
Personalized recommendation is a ubiquitous application on the internet, with many industries and hyperscalers extensively leveraging Deep Learning Recommendation Models (DLRMs) for their personalization needs (like ad serving or movie suggestions).
Externí odkaz:
http://arxiv.org/abs/2410.22249
Autor:
Sharma, Aakash, Bhasi, Vivek M., Singh, Sonali, Kesidis, George, Kandemir, Mahmut T., Das, Chita R.
We propose a novel GPU-cluster scheduler for distributed DL (DDL) workloads that enables proximity based consolidation of GPU resources based on the DDL jobs' sensitivities to the anticipated communication-network delays. Our scheduler consists of th
Externí odkaz:
http://arxiv.org/abs/2401.16492
Autor:
Ramezani, Morteza, Cong, Weilin, Mahdavi, Mehrdad, Kandemir, Mahmut T., Sivasubramaniam, Anand
Despite the recent success of Graph Neural Networks (GNNs), training GNNs on large graphs remains challenging. The limited resource capacities of the existing servers, the dependency between nodes in a graph, and the privacy concern due to the centra
Externí odkaz:
http://arxiv.org/abs/2111.08202
Autor:
Sarma, Anup, Singh, Sonali, Jiang, Huaipan, Pattnaik, Ashutosh, Mishra, Asit K, Narayanan, Vijaykrishnan, Kandemir, Mahmut T, Das, Chita R
Machine/deep-learning (ML/DL) based techniques are emerging as a driving force behind many cutting-edge technologies, achieving high accuracy on computer vision workloads such as image classification and object detection. However, training these mode
Externí odkaz:
http://arxiv.org/abs/2109.07710
Recurrent Neural Networks (RNNs), more specifically their Long Short-Term Memory (LSTM) variants, have been widely used as a deep learning tool for tackling sequence-based learning tasks in text and speech. Training of such LSTM applications is compu
Externí odkaz:
http://arxiv.org/abs/2106.12089
Graph-based data structures have drawn great attention in recent years. The large and rapidly growing trend on developing graph processing systems focuses mostly on improving the performance by preprocessing the input graph and modifying its layout.
Externí odkaz:
http://arxiv.org/abs/2104.10039
Autor:
Gunasekaran, Jashwant Raj, Thinakaran, Prashanth, Chidambaram, Nachiappan, Kandemir, Mahmut T., Das, Chita R.
Datacenters are witnessing a rapid surge in the adoption of serverless functions for microservices-based applications. A vast majority of these microservices typically span less than a second, have strict SLO requirements, and are chained together as
Externí odkaz:
http://arxiv.org/abs/2008.12819
Autor:
Ausavarungnirun, Rachata, Ghose, Saugata, Kayıran, Onur, Loh, Gabriel H., Das, Chita R., Kandemir, Mahmut T., Mutlu, Onur
In a modern GPU architecture, all threads within a warp execute the same instruction in lockstep. For a memory instruction, this can lead to memory divergence: the memory requests for some threads are serviced early, while the remaining requests incu
Externí odkaz:
http://arxiv.org/abs/1804.11038
Autor:
Jung, Myoungsoo, Kandemir, Mahmut T.
Resource utilization is one of the emerging problems in many-chip SSDs. In this paper, we propose Sprinkler, a novel device-level SSD controller, which targets maximizing resource utilization and achieving high performance without additional NAND fla
Externí odkaz:
http://arxiv.org/abs/1705.04627
In this paper, we present a novel cache design based on Multi-Level Cell Spin-Transfer Torque RAM (MLC STTRAM) that can dynamically adapt the set capacity and associativity to use efficiently the full potential of MLC STTRAM. We exploit the asymmetri
Externí odkaz:
http://arxiv.org/abs/1704.05044