Nanosecond optical switching and control system for data center networks
Electrical switching based data center networks have an intrinsic bandwidth bottleneck and,
require inefficient and power-consuming multi-tier switching layers to cope with the rapid …
require inefficient and power-consuming multi-tier switching layers to cope with the rapid …
Semeru: A {Memory-Disaggregated} managed runtime
Resource-disaggregated architectures have risen in popularity for large datacenters.
However, prior disaggregation systems are designed for native applications; in addition, all …
However, prior disaggregation systems are designed for native applications; in addition, all …
SiP-ML: high-bandwidth optical network interconnects for machine learning training
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 …
bandwidth ML training clusters with strong scaling properties. Our design, called SiP-ML …
One-sided {RDMA-Conscious} extendible hashing for disaggregated memory
Memory disaggregation is a promising technique in datacenters with the benefit of improving
resource utilization, failure isolation, and elasticity. Hashing indexes have been widely used …
resource utilization, failure isolation, and elasticity. Hashing indexes have been widely used …
{ROLEX}: A Scalable {RDMA-oriented} Learned {Key-Value} Store for Disaggregated Memory Systems
Disaggregated memory systems separate monolithic servers into different components,
including compute and memory nodes, to enjoy the benefits of high resource utilization …
including compute and memory nodes, to enjoy the benefits of high resource utilization …
{FORD}: Fast one-sided {RDMA-based} distributed transactions for disaggregated persistent memory
Persistent memory (PM) disaggregation improves the resource utilization and failure
isolation to build a scalable and cost-effective remote memory pool. However, due to offering …
isolation to build a scalable and cost-effective remote memory pool. However, due to offering …
Corundum: An open-source 100-gbps nic
Corundum is an open-source, FPGA-based prototy** platform for network interface
development at up to 100 Gbps and beyond. The Corundum platform includes several core …
development at up to 100 Gbps and beyond. The Corundum platform includes several core …
Gimbal: enabling multi-tenant storage disaggregation on SmartNIC JBOFs
Emerging SmartNIC-based disaggregated NVMe storage has become a promising storage
infrastructure due to its competitive IO performance and low cost. These SmartNIC JBOFs …
infrastructure due to its competitive IO performance and low cost. These SmartNIC JBOFs …
Fast, scalable, and programmable packet scheduler in hardware
V Shrivastav - Proceedings of the ACM special interest group on data …, 2019 - dl.acm.org
With increasing link speeds and slowdown in the scaling of CPU speeds, packet scheduling
in software is resulting in lower precision and higher CPU utilization. By offloading packet …
in software is resulting in lower precision and higher CPU utilization. By offloading packet …
{MemLiner}: Lining up Tracing and Application for a {Far-Memory-Friendly} Runtime
Far-memory techniques that enable applications to use remote memory are increasingly
appealing in modern datacenters, supporting applications' large memory footprint and …
appealing in modern datacenters, supporting applications' large memory footprint and …