Disaggregated memory for expansion and sharing in blade servers

502 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Trends in big data analytics (2014, methodology): “Flash-based NVMs can be exposed as block devices either through Serial Attached SCSI (SAS) and Serial Advanced Technology Attachment (SATA) or PCI express interfaces or even combined with dis-aggregated memory to provide cost-effective solutions [61].”
Buddy Compression: Enabling Larger Memory for Deep Learning and HPC Workloads on GPUs (2019, background): “com/ 2017/02/17/tsubame3-0-points-future-hpe-pascal-nvlink-opa-server/ [54] K.”
Who Limits the Resource Efficiency of My Datacenter: An Analysis of Alibaba Datacenter Traces (2019, background): “[16] has pointed out that the increasing imbalance between computation and memory may cause memory capacity per-core dropping by 30% every two years.”
CACTI 7 (2017, methodology): “…products have emerged recently or will soon be available, including buffer-on-board (BoB) [Intel 2014], LRDIMM [Micro 2015a], HyperCloud DIMMS (HC-DIMMs) [Netlist 2012], NVDIMMs [SanDisk 2014], HMC [Pawlowski 2011], NVMs [Burr et al. 2010; Strukov et al. 2008], and memory blades [Lim et al. 2009].”
Processing Data Where It Makes Sense: Enabling In-Memory Computation (2019, background): “Unfortunately, it has become increasingly difficult in recent years to scale all of these dimensions [1, 2, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31], and the main memory bottleneck has instead been worsening.”