GRIDS@SC:

INTERVIEW WITH FORCE10
By Alan Beck, Editor-in-Chief, HPCwire

Recently, HPCwire conducted an interview with a representative of Force10. The questions and answers follow below:

HPC: Force10 had been quiet for a while, but recently there has been new announcements regarding the TeraGrid project, CERN and a global relationship with

IBM: What has changed?

As you know, Force10 delivered line rate 10 Gigabit Ethernet switch/routers in October 2002 more than a year ahead of the competition. Since then, the E-Series family of switch/routers has seen tremendous traction in the market. Our efforts have been focused on building the business, extending the product line, developing new features, winning key customers.

Guided by the leadership of CEO Marc Randall, Force10 is committed to this tradition. We are focused on engineering execution first and marketing second. At Force10 we are committed to extraordinary products for high performance network environments. To do this end we think about the entire customer experience of buying, owning and operating a high performance network. So, for example, our product represents a new level of performance and resiliency, helping to make Ethernet based network more predictable through high availability architecture, massive port densities and Terabit scalability in our backplane and switch fabric designs. Let me give you another example. In order to simplify network operation, we have devised unique proactive service and support capabilities aimed at first helping customers avoid problems and over time enhancing our ability to support customers during down time. We are committed to working with existing and new customers to continue to push the limits of what people think is possible. We want to keep surprising this community as the HPC market has really been the foundation for why we are now a growing enterprise.

HPC: Can you tell me how other customers outside of the HPC market are leveraging your solutions?

IBM: The Force10 customer base is growing in all segments, and across all geographies. We are seeing the concept of Grid computing is now entering the mainstream enterprise network environment. Whether the company's focus is manufacturing, banking or portals, the movement to localized but distributed computational resources -- be they PC's or blade servers -- is now very much the norm.

For banks and portals, the need to data mine and process transactions has driven the need for a more resilient and scalable aggregation layer in the network. On these dimensions -- scalability and resiliency -- Force10 leads the industry.

For manufacturing companies, as well as biomedical and geophysical markets, we see very similar needs as in the HPC market. The need to process massive amounts of information rapidly and with as low cost as possible is core to the business. As a result the combination of high speed computing and high performance networking is a perfect match.

HPC: What are you announcing at SC2003?

IBM: We are excited to announce the latest addition to our family of line-rate 10 GbE and 1 GbE resilient switch/routers. The E300 complements the E600 and E1200 by extending the product line to reach mid-size clusters. The E-Series E300 Gigabit/10 Gigabit Ethernet switch/router -- a chassis-based system that supports up to 6 10GbE and 72 GbE ports. Extending features and capabilities you'd expect from all Force10 Networks products, the E300 delivers, in a compact 6-RU platform, 100 percent line rate, non-blocking performance across every port with all high-touch services enabled. Moreover, the E300 is the first small form factor 10 Gigabit Ethernet switch/router and as a result it dramatically lowers the entry cost for a customer interested in deploying 10 Gigabit Ethernet

HPC: What are you demonstrating at SC2003?

IBM: Our systems are known for delivering resiliency, scalability and performance.

It is easy to demonstrate speeds and feeds. And, as you know, the Force10 E-Series remains the industry leader in line rate Gigabit and 10 Gigabit densities. But customers demand more than speeds and feeds. They demand resiliency. So we will be demonstrating both industry leading performance and best in class resiliency in our Show booth number 1711.

The E300, E600, and E1200 will demonstrate resiliency technology that increases Grid and cluster stability and availability. Leveraging the EtherScale "High Availability" design, the E-Series switch/routers will demonstrate redundant Route Processor Module hitless fail-over of Layer 2 and Layer 3 traffic as well as graceful and concurrent OSPF and BGP restarts -- capability expected for truly resilient and "Hitless" Layer 3 fail-over. All of this will be done at line rate in a 10 Gigabit network.

HPC: How can readers get more information about Force10 Networks.

IBM: Beyond readers attending the show stopping by our booth -- 1711 -- we continue to expand the information available on our Web site: http://www.force10networks.com or they can call us at (866) 571-2600.