HPCwire: Mellanox Builds Bridges to Ethernet, Fibre Channel

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

HPCwire >> Topic >> Networks

February 17, 2009

Mellanox Builds Bridges to Ethernet, Fibre Channel

by Michael Feldman, HPCwire Editor

Page: 1 of 2
1 | 2 All »

Leveraging the momentum it has achieved with 40 Gbps InfiniBand, Mellanox Technologies is entering the network gateway business. Today the company introduced BridgeX, a technology that offers fabric convergence for SAN and LAN infrastructure in multi-protocol datacenters.

In a nutshell, BridgeX is a silicon device that consolidates gateway communication for InfiniBand, Ethernet and Fibre Channel fabrics. The concept is the same as for the company's ConnectX-based adapter products: use a single piece of silicon to handle multiple network protocols. The BridgeX chip is meant to be incorporated into gateway hardware -- bridges or switch blade modules -- that can carry Ethernet or Fibre Channel traffic over lossless 40 Gbps InfiniBand or 10GbE fabrics. The goal is for BridgeX-equipped gateways to reduce network infrastructure complexity and cost by replacing multiple switches and bridges that handle only single protocols.

Each BridgeX chip has the capacity to manage two 40 Gbps InfiniBand or six 10GbE uplinks and deliver a combination of up to six 10GbE ports and up to eight 2/4/8Gb Fibre Channel ports on the downlink side. The exact mix of downlink ports for Fibre Channel and Ethernet can be configured dynamically, enabling a good deal of flexibility as datacenter resources grow or get moved around.

Since Ethernet is available on both the uplink and downlink sides, some of the downlink ports can serve as a 10GbE pass-through. For InfiniBand bridging, there appears to be no pass-through capability. So if InfiniBand-based storage needs to be accessed, that traffic would have to be routed through an external switch or adapter connected directly to a server.

The silicon can be used to build two flavors of gateways -- 40 Gbps InfiniBand or 10GbE. The former enables Gigabit Ethernet, 10GbE, and Fibre Channel traffic to run over an InfiniBand network, while the latter version can be used for Fibre Channel over Ethernet (FCoE) bridging.

Mellanox itself is launching its own InfiniBand bridge product line, the BX4000 series, which incorporates both types of gateways (may be configured for either InfiniBand or Ethernet as the convergence fabric). A BX4010 gateway contains a single BridgeX chip, while the BX4020 contains two bridge chips, enabling four 40Gbps InfiniBand or 12 10GbE uplinks and the corresponding number of downlinks. Multiple boxes can be stacked if even denser bridging is required.

Mellanox's foray into building network boxes is relatively recent. While mainly known for offering switch silicon (InfiniScale) and network adapter cards (ConnectX, InfiniHost), six to nine months ago the company quietly launched its own line of InfiniBand switches -- the MTX36xx series. Now with the BX4000 boxes, Mellanox seems intent on expanding its ecosystem footprint even further.

At the same time, Mellanox will be offering BridgeX silicon on the open market for other OEMs to incorporate into gateway boxes. Although no partners were announced at the product launch, Mellanox is in discussion with both InfiniBand and Ethernet vendors looking to support unified fabric environments.

The initial deployments of BridgeX will likely be for InfiniBand setups, especially since Mellanox is offering its own InfiniBand bridge capability in the BX4000. This could be a big deal in the high performance computing space, since a lot of storage is attached to InfiniBand clusters via Ethernet. Fibre Channel is basically a no-show in HPC, but to the extent InfiniBand is used in commercial enterprises, this technology could offer a cost-effective solution for connecting legacy storage and Ethernet LANs with InfiniBand-based clusters.

Since InfiniBand is much less widespread than either of the other two protocols, the larger market for BridgeX is for Ethernet to Fibre Channel gateways. Conveniently, the FCoE standard is expected to be ratified this year just as the 10GbE ecosystem should be shifting into high gear. It remains to be seen if traditional Ethernet vendors like Cisco and Brocade will be tempted by BridgeX or if their smaller rivals will see the Mellanox technology as a way to gain a competitive edge in the FCoE market.

According to Gilad Shainer, director of technical marketing at Mellanox, most of the bridge solutions on the market today rely on a big powerful CPU inside a server, along with a lot of management software. Bridges implemented with this model perform protocol termination at the uplink side and initiate the new protocol on the downlink end.

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

Feature Articles

Moscow State University Supercomputer Has Petaflop Aspirations

The Moscow State University supercomputer, Lomonosov, has been selected for a high-performance makeover, with the goal of tripling its processing power to achieve petaflop-level performance in 2010. T-Platforms, who developed and manufactured the supercomputer, is the odds-on favorite to lead the project.
Read More...

Intel Ups Performance Ante with Westmere Server Chips

Right on schedule, Intel has launched its Xeon 5600 processors, codenamed "Westmere EP." The 5600 represents the 32nm sequel to the Xeon 5500 (Nehalem EP) for dual-socket servers. Intel is touting better performance and energy efficiency, along with new security features, as the big selling points of the new Xeons.
Read More...

The Week in Review

The ACM Turing Award goes to the creator of the modern personal computer; and Voltaire announces a mid-range InfiniBand switch and new technology that accelerates distributed applications. We recap those stories and more in our weekly wrapup.
Read More...

Read more HPCwire features...

Top Headlines

AMD: OEMs primed for Opteron 6100s

Mar 17 | The Register | But what about the tier ones? Read more...

Arrival of the Desktop Supercomputer

Mar 17 | Cadalyst Magazine | A new generation of workstations is changing the nature of technical computing. Read more...

Scheduling HPC In The Cloud

Mar 17 | Linux Magazine | Latest iteration of Sun Grid Engine able to tap into Cloud. Read more...

Tailoring Medicine with Supercomputers

Mar 16 | Bio-IT World | Biotech firm builds genetic models from patient data. Read more...

Gelsinger Stuns Analysts and Colleagues with Storage Pool Plan

Mar 15 | The Register | EMC's grand vision for unified global storage. 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.

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