Journal of Optical Communications and Networking · 2021 · 13 citations · 20 references
EngineeringHeterogeneous MemoryHigh Performance Computer NetworkComputer ArchitectureData Center NetworkEmbedded SystemsHardware SystemsMulti-channel Memory ArchitectureComputer MemoryHigh-performance ArchitectureComputing SystemsMemory CardMemory DevicesParallel ComputingData ManagementOptical NetworkingMemory ResourcesData Center SystemOptical NetworkComputer EngineeringComputer ScienceMemory ArchitectureData Center ManagementOptical MemoryHigh Bandwidth MemoryCloud ComputingIn-memory Computing
The memory over optical network (MONet) system is a disaggregated data center architecture where serial (HMC)/parallel (DDR4) memory resources can be accessed over optically switched interconnects within and between racks. An FPGA/ASIC-based custom hardware IP (ReMAT) supports heterogeneous memory pools, accommodates optical-to-electrical conversion for remote access, performs the required serial/parallel conversion, and hosts the necessary local memory controller. An optically interconnected HMC-based (serial I/O type) memory card is accessed by a memory controller embedded in the compute card, simplifying the hardware near the memory modules. This substantially reduces overheads on latency, cost, power consumption, and space. We characterize CPU–memory performance by experimentally demonstrating the impact of distance, number of switching hops, transceivers, channel bonding, and bit rate per transceiver on the bit error rate, power consumption, additional latency, sustained remote memory bandwidth/throughput (using industry standard benchmark STREAMS), and cloud workload performance (such as operations per second, average added latency, and retired instructions per second memcached with YCSB cloud workloads). MONet pushes the CPU–memory operational limit from a few centimeters to tens of meters, yet applications can experience as low as 10% performance penalty (at 36 m) compared to a direct-attached equivalent. Using the proposed parallel topology, a system can support up to 100,000 disaggregated cards.
20
Benchmarking cloud serving systems with YCSB
Brian F. Cooper, Adam Silberstein, Erwin Tam et al. · 2010 · 3.6K citations
The Case for Energy-Proportional Computing
Luiz André Barroso, Urs Hölzle · Computer · 2007 · 2.5K citations
Disaggregated memory for expansion and sharing in blade servers
Kevin Lim, Jichuan Chang, Trevor Mudge et al. · 2009 · 364 citations
Engineering, Shared Memory, High-performance Architecture +12
Network support for resource disaggregation in next-generation datacenters
Sangjin Han, Norbert Egi, Aurojit Panda et al. · 2013 · 162 citations