Virtual memory mapped network interface for the SHRIMP multicomputer

Blumrich, Matthias A.; Li, K.; Alpert, Richard; Dubnicki, Cezary; Felten, Edward W.; Sandberg, Jonathan S.

doi:10.1145/192007.192024

Cited by 20 publications

(12 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, modifying the routing algorithm in order to introduce adaptivity does not add any extra delay, since the basic operation (accessing a table) is the same, and routing tables are programmed at the network initialization phase. 3 With respect to the selection of a single output channel from a set of candidates, since up Ã /down Ã routing is partially adaptive, it requires making a selection in case that the routing table returns several outgoing channels. When the degree of adaptivity is increased by applying this design methodology, the selection function has to select among a larger set of choices, increasing its delay.…”

Section: Applying the First Methodology: Increasing Adaptivitymentioning

confidence: 99%

See 1 more Smart Citation

High-performance routing in networks of workstations with irregular topology

Silla

Duato

2000

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

ÐNetworks of workstations are rapidly emerging as a cost-effective alternative to parallel computers. Switch-based interconnects with irregular topology allow the wiring flexibility, scalability, and incremental expansion capability required in this environment. However, the irregularity also makes routing and deadlock avoidance on such systems quite complicated. In current proposals, many messages are routed following nonminimal paths, increasing latency and wasting resources. In this paper, we propose two general methodologies for the design of adaptive routing algorithms for networks with irregular topology. Routing algorithms designed according to these methodologies allow messages to follow minimal paths in most cases, reducing message latency and increasing network throughput. As an example of application, we propose two adaptive routing algorithms for AN1 (previously known as Autonet). They can be implemented either by duplicating physical channels or by splitting each physical channel into two virtual channels. In the former case, the implementation does not require a new switch design. It only requires changing the routing tables and adding links in parallel with existing ones, taking advantage of spare switch ports. In the latter case, a new switch design is required, but the network topology is not changed. Evaluation results for several different topologies and message distributions show that the new routing algorithms are able to increase throughput for random traffic by a factor of up to 4 with respect to the original up Ã /down Ã algorithm, also reducing latency significantly. For other message distributions, throughput is increased more than seven times. We also show that most of the improvement comes from the use of minimal routing.

show abstract

Section: Applying the First Methodology: Increasing Adaptivitymentioning

confidence: 99%

“…This overhead traditionally accounted for a high percentage of message latency [28], [24]. However, we did not consider such an overhead because some recent proposals reduce and/or hide that overhead, thus exposing hardware latency [18], [3], [25], [13]. We only considered network hardware latency in this study.…”

Section: Performance Evaluationmentioning

confidence: 99%

High-performance routing in networks of workstations with irregular topology

Silla

Duato

2000

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

show abstract

“…SHRIMP [6] uses a specialized NIC to provide userlevel messaging by allowing processes to directly write the memory of other processes through hardware support. Cashmere [56] leverages DEC's Memory Channel [22], a remote-write network, to implement a software DSM.…”

Section: Related Workmentioning

confidence: 99%

Scale-out NUMA

et al. 2014

View full text Add to dashboard Cite

Emerging datacenter applications operate on vast datasets that are kept in DRAM to minimize latency. The large number of servers needed to accommodate this massive memory footprint requires frequent server-to-server communication in applications such as key-value stores and graph-based applications that rely on large irregular data structures. The fine-grained nature of the accesses is a poor match to commodity networking technologies, including RDMA, which incur delays of 10-1000x over local DRAM operations. We introduce Scale-Out NUMA (soNUMA)-an architecture, programming model, and communication protocol for low-latency, distributed in-memory processing. soNUMA layers an RDMA-inspired programming model directly on top of a NUMA memory fabric via a stateless messaging protocol. To facilitate interactions between the application, OS, and the fabric, soNUMA relies on the remote memory controller-a new architecturally-exposed hardware block integrated into the node's local coherence hierarchy. Our results based on cycle-accurate full-system simulation show that soNUMA performs remote reads at latencies that are within 4x of local DRAM, can fully utilize the available memory bandwidth, and can issue up to 10M remote memory operations per second per core.

show abstract

“…It simulates large-scale ocean movements based on eddy and boundary currents. Shpatial is a modified version of the Splash-2 spatial water application that we obtained from the SHRIMP project [8]. The primary differences between Shpatial and WaterSp are that data locality has been increased and synchronization decreased.…”

Section: Application Suitementioning

confidence: 99%

Per-node multithreading and remote latency

Thitikamol

Keleher

1998

IEEE Trans. Comput.

View full text Add to dashboard Cite

This paper evaluates the use of per-node multi-threading to hide remote memory and synchronization latencies in software DSMs. As with hardware systems, multi-threading in software systems can be used to reduce the costs of remote requests by running other threads when the current thread blocks. We added multi-threading to the CVM software DSM and evaluated its impact on the performance of a suite of common shared memory programs. Multi-threading resulted in speed improvements of at least 20% in two of the applications, and better than 15% for several other applications. However, we also found that good performance can not always be achieved transparently for non-trivial applications. Also, the characteristics of the underlying DSM protocol can have a large effect on multi-threading's utility.

show abstract

Virtual memory mapped network interface for the SHRIMP multicomputer

Cited by 20 publications

References 20 publications

High-performance routing in networks of workstations with irregular topology

High-performance routing in networks of workstations with irregular topology

Scale-out NUMA

Per-node multithreading and remote latency

Contact Info

Product

Resources

About