Nam Sung Kim
6 papers · 2018–2024 · 4 conferences · across top CS/AI conferences
Achievements
Jump to papers ↓+4 more ↓ Show less ↑
π£ Hot Topic Early Bird π Interdisciplinary Bridge πΊοΈ Taxonomy Completionist (11) π§ Keyword Pioneer π Conference Polyglot (4)
π
Academic Marathon
(6)
π
Cross-Pollinator
(3)
π
Keyword Champion
π
Trend Setter
Conferences
NIPS (2)
NSDI (2)
ICLR (1)
OSDI (1)
Top co-authors
Research topics
Keywords
distributed training
(3)
gradient aggregation
(1)
distributed learning
(1)
neural network optimization
(1)
vector quantization
(1)
rate limiting
(1)
convolutional neural network
(1)
gradient compression
(1)
performance isolation
(1)
data center
(1)
programmable switch
(1)
bandwidth efficiency
(1)
storage system
(1)
i/o scheduler
(1)
nvme driver
(1)
firmware optimization
(1)
gradient quantization
(1)
parallel training
(1)
floating point operation
(1)
network acceleration
(1)
Papers
Harmonic: Hardware-assisted RDMA Performance Isolation for Public Clouds
NSDI 2024
PipeGCN: Efficient Full-Graph Training of Graph Convolutional Networks with Pipelined Feature Communication
ICLR 2022
Unlocking the Power of Inline Floating-Point Operations on Programmable Switches
NSDI 2022
Pipe-SGD: A Decentralized Pipelined SGD Framework for Distributed Deep Net Training
NIPS 2018
GradiVeQ: Vector Quantization for Bandwidth-Efficient Gradient Aggregation in Distributed CNN Training
NIPS 2018
FlashShare: Punching Through Server Storage Stack from Kernel to Firmware for Ultra-Low Latency SSDs
OSDI 2018