HPCwire: Ethernet Fabrics Offer Way Forward for Seismic Applications

The Leading Source for Global News and Information Covering the Ecosystem of High Productivity Computing

HPCwire >> Features

March 24, 2009

Ethernet Fabrics Offer Way Forward for Seismic Applications

by Joseph Ammirato, Vice President of Marketing, Woven Systems

Page: 1 of 2
1 | 2 All »

Exploring new frontiers, driving higher efficiencies, and shortening time-to-production. This is the lifeblood of the oil and gas industry. Businesses in this industry are constantly finding new and more efficient ways to interpret and model data and ascertain risk through cutting-edge technologies to maintain a competitive advantage.

With ever increasing simulation model sizes, locating, extracting, and producing energy has become an extremely data-intensive undertaking. Examples of this burgeoning data demand include massive parallel computation and simulation models from geophysicists; 3D visualization and interpretation of geographic data from reservoir simulation engineers, and mining historic seismic data of current reservoirs with up-to-date algorithms. IT managers are in a continuous race to stay ahead of these massive compute and data flow tasks.

Figure 1: Visualizing oil reservoir from Engineering Simulation and Scientific Software's Cyclops software

Deploying more servers, network storage, and workstations in the datacenter can temporarily relieve the demand for increased compute and storage capacity. However, new problems quickly arise as IT managers are confronted with the challenge of interconnecting massive amounts of devices that constantly produce and exchange huge data sets. Without a network that can maintain performance as compute and storage capacity grows, sustaining a competitive edge becomes more difficult as the time-to-result lengthens.

Traditional Datacenter Networks Become Bottlenecks

Ethernet technology has served the industry well over the past few decades. But given the growing demands on oil and gas datacenters -- particularly in terms of compute, storage, and bandwidth requirements -- traditional datacenter networks are becoming inhibitors in the race to stay ahead.

Figure 2: Traditional network diagram

These key inhibitors include:

Spanning Tree Protocol (STP)

STP leads to an underutilized network infrastructure. It was designed to prevent looping traffic in a traditional multi-tier network. STP ensures all network forwarding paths are in a primary-backup model. The end result is a very high rate of inefficiency with up to 50 percent of the available network bandwidth sitting idle in a hot-standby mode.

Figure 3: Spanning tree network diagram

around

Best Effort Delivery Mechanism Leads to Data Loss

Ethernet relies on "best-effort" data transmission. Traditional network devices are assigned buffers to deal with temporary network bursts. Once the buffer is exhausted, the network switch simply drops further incoming packets and relies on higher level networking protocols to recover and retransmit. Ethernet's approach, while sufficient for normal day-to-day IT applications and traffic patterns, is grossly inadequate for large-scale computing models in the oil and gas industry. Compute servers and storage infrastructures drive increased model sizes and data sets at high speed. Geophysicists and network architects have found that these heavy traffic patterns result in constant data loss that requires additional computing and storage cycles just to recover from the network data loss.

Store-and-forward Switching Leads to Unpredictable Jitter

Traditional Ethernet switching infrastructures are built on "store-and-forward" architectures. Incoming traffic is stored, analyzed, and forwarded to the destination. Each and every forwarding stage of the network repeats this forwarding process as data traffic propagates through the network. Due to the variable size of network data sets, a traditional Ethernet network always has performance variability or "jitter" (the variance in latency among all data flows). While simple and fast enough for traditional IT applications, such as PC desktop connectivity, jitter actually causes unpredictable performance degradation in the oil and gas datacenter.

Typical applications in oil and gas -- such as simulation, location assessment, risk management, visualization and others -- all require data input from a network storage source which feeds into a Message Passing Interface (MPI)-based parallel application environment. These MPI applications are designed to constantly process data sets with variable sizes in real-time, and in a parallel and pipe-lined computing environment. Due to the unpredictable nature of traditional Ethernet jitter, these parallel oil and gas applications are forced to spend significant idle cycles waiting for the completion of the slowest process before moving to the next stage of their pipe-lined computing process. The end result is longer time-to-result.

All in all, it is clear that the network interconnect is lengthening the oil and gas industry's time-to-result. A new architecture and a new delivery mechanism are needed. Both must be ubiquitous, standards-based, and low-cost while driving a higher level of performance. In this way, oil and gas datacenters can facilitate faster time-to-result.

Ethernet Fabrics Break the Bottleneck

While traditional datacenter networks are becoming inhibitors for oil and gas companies in their race to stay ahead, Ethernet fabrics present a new way for oil and gas companies to speed time-to-result and maintain a competitive edge.

Page: 1 of 2
1 | 2 All »

Article Tools

Share Options

(Digg, Technorati, more)

Subscribe

Subscribe to HPCwire

Discussion

There are 0 discussion items posted.

Join the Discussion

You must be a registered user in order to comment on HPCwire stories.

Members enjoy:

	Access to Special Content, including Podcasts, Webcasts, and Whitepapers and more!
	Ability to take an active role in the discussions that are shaping the HPC technology discipline!
	Special offers available only to HPCwire members!

Become a Registered User Today!

Registered Users Log in Here to Comment

Email Address:	Password (case sensitive)

Remember me	Forgot Password?

RSS Feeds

Feeds by Topic

Feeds by Industry

Feeds by Content Type

Top Headlines

IBM Releases Energy Efficient Power7 System

Feb 09 | eWeek Europe | Company says new high-end servers will deliver "intelligent performance." Read more...

Inductive Coupling Packs Flash Drive in a Chip

Feb 09 | EE Times | Wireless technology promises energy-efficient chip-to-chip communication. Read more...

IBM, Microsoft Help Create Montana Supercomputer

Feb 08 | eWeek | A new kind of Rocky Mountain high. Read more...

AMD Aims for GPUs in Mainstream Servers Starting 2012

Feb 08 | Computerworld | Chip maker hopes to bring CPU-GPU processors to servers in two years. Read more...

Graphene Transistors That Work at Blistering Speeds

Feb 05 | Technology Review | IBM has created graphene transistors that leave silicon ones in the dust. Read more...

Read more headlines...

Featured Whitepapers

Virtualization for Aggregation And The vSMP Architecture™

Jan 12 | | In-depth look at vSMP Foundation server virtualization technology, technical implementation, use cases and capabilities. The technical whitepaper provides an architectural overview and details on the three vSMP Foundation products: vSMP Foundation for SMP, vSMP Foundation for Cluster and vSMP Foundation for Cloud.

Copper Cable Technologies for High Performance Computing

Jan 18 | | This white paper discusses Gore’s copper cable assemblies, and how they continue to exceed the standards for providing reliable, cost-effective solutions for high-performance computer applications.

Appro Assists LLNL with Cluster Designed for Extreme Scale Visualization

Jan 11 | | LLNL is home to some of the fastest computers in the world. In 2012, LLNL expects to have the Sequoia supercomputing cluster operational with a projected performance of over 20 PFLOP/s. These systems will focus on strengthening the foundations of predictive simulation through running large suites of complex simulations and then comparing model predictions with experimental data. To visualize this project’s large amount of data, LLNL requested an Appro Supercomputing Cluster specifically designed to support interactive data analysis.

View the White Paper Library

Multimedia

Webcast: Virtualized Data Center Roundtable

Join this online panel discussion for live Q&A with leading industry experts, analysts, and end-users to discuss the latest innovations, best practices, barriers to implementation, and measurable benefits of server virtualization with a particular focus on today's real world solutions.

Webcast: Watch SC09 Birds of a Feather Video: Scalable Fault-Tolerant HPC Supercomputers

Learn about scalable fault-tolerant architectures and examples of energy efficient and scalable supercomputing clusters using dual QDR InfiniBand to combine capacity computing with network failover capabilities with the help of programming languages such as MPI and a robust Linux cluster management package.

Webcast: High Performance Computing for a Smarter Planet

LIVE@SCO9: The IBM team discusses new innovations in hardware, software and services that help clients better understand their workloads and get insight from their R&D efforts. Technology demonstrations include the soon-to-be-released Power7 HPC processor, the DCS990 system with 2.4 petabytes of storage, the xCAT management tool, secure HPC cloud computing and more. Winners of two HPCwire Readers' and Editors’ Choice Awards! Take the IBM virtual tour at SC09 or more information go online to: http://www-03.ibm.com/systems/deepcomputing/sc09.html