 |
|
DAILY NEWS AND INFORMATION
FOR THE GLOBAL GRID COMMUNITY /
|

LINUX NETWORX VP LAYS OUT NEWFOUND
GRID STRATEGY
GRIDtoday recently interviewed Eric Pitcher, vice president of product
marketing for Linux Networx, about his company's announcement that, after a
successful run in the cluster community, Linux Networx is now entering the
Grid market.
GRIDtoday: What is Linux Networx's strategy for entering
the
Grid market?
ERIC PITCHER: The next generation of supercomputing will
involve new computing paradigms-virtualization, on-demand provisioning,
collaborative computing and secure information exchange crossing
organizational boundaries -- which is where Grid computing comes in.
Linux Networx recognizes this emerging era in computing and is prepared to
meet these demands by developing a Grid-enabled cluster solution. We've
partnered with the leading Grid software provider, Platform Computing, to
deploy Grid software solutions on our clusters. The result is a turnkey,
Grid-enabled cluster solution.
Gt: Why offer Grid-enabled solutions now?
EP: Customers are looking for Grid-enabled solutions that
have added value, support and services so they can hit the ground running and
maximize their ROI. Linux Networx has already proven itself as a leader in the
cluster community by offering added value for high productivity computing. The
cluster expertise from Linux Networx coupled with Platform's industry-leading
Grid software provides a proven solution to help our customers meet this
emerging wave in high performance computing.
Gt: Explain the concept of high-productivity computing?
EP: High productivity computing takes the focus off of
theoretical peak performance and emphasizes the importance of maximizing the
number of jobs that are able to run on a system during its lifetime. Linux
Networx customers have demanding organizational goals associated with their
computing systems and are more concerned about the actual production
capabilities of their machine rather than theoretical performance. Linux
Networx delivers clusters that greatly accelerate product development, improve
product reliability and overall save customers significant time and money.
Gt: What value does Linux Networx bring to Grid
computing?
EP: With Linux Networx, customers receive a Grid-enabled
cluster with the peace-of-mind that comes from knowing their cluster comes
fully integrated with our cluster management tools and other services we
provide to ensure a high productivity machine. These capabilities include
rigorous Q/A and validation, full pre-ship system buildup and testing, and
optimized applications.
Now than ever before, organizations have demanding organizational goals and
rely on their computing systems for intensive research efforts. We've
consistently been able to provide high productivity systems to improve our
customers' product reliability and save them significant time and money.
Gt: What kind of impact has Linux Networx had on the cluster
community?
EP: Linux Networx has proven its leadership position in
the
high performance computing market through innovative technology and services
that provide customers with high productivity computing. Since being the first
company to deliver a commercial cluster, Linux Networx has demonstrated its
vision in cluster computing by accomplishing many "firsts." We delivered the
first commercial Linux cluster and were the first to develop cluster
management tools. Linux Networx also made waves in the HPC industry when it
was the first company to design, build and patent vertically mounted nodes. In
2002, Linux Networx got world-wide attention for designing and building the
world's fastest Linux cluster-the first Linux cluster to be ranked as a top
five supercomputer.
Gt: Linux Networx has demonstrated its ability to
successfully deploy the latest clustering technology to its customers. What
impact does this have on the cluster market, in general?
EP: Organizations are looking for cluster vendors that
contribute value to their systems rather than just deliver hardware. Customers
have learned from previous experiences that buying cheap hardware results in
additional time and money spent trying to manage the system and get jobs to
successfully run on the cluster. Cluster vendors with the ability to validate
the latest components and integrate them successfully, design the optimal
system to meet the customer's applications profile, integrate management
tools, and offer cluster expertise and training are some of the items that
allow organizations to maximize their cluster investment.
The real value of an HPC system is the amount of production an organization
is
able to receive from the cluster during its lifetime. For example, the
Lightning system (2,816 processors) Linux Networx delivered to Los Alamos
National Lab was delivered in just three months from when it was ordered to
when the Lab was running production jobs on the cluster. To guarantee high
productivity on your cluster and maximum sustained performance, organizations
need to look for vendors with these added value items and Linux and clustering
expertise.
Gt: What should organizations look for in a cluster
vendor?
EP: Selecting the right cluster vendor is crucial in
helping
an organization meet its computational goals. Work with a cluster vendor with
proven cluster expertise in designing and building turnkey solutions, as well
as rigorous Q/A and validation processes, pre-ship system buildup and testing,
optimized applications, the latest cluster technologies, open source tools,
and cluster management solutions. The right vendor can reduce the risk of
installing a high-powered computing system-but also can deliver a highly
productive computing machine that helps bring a product to market quicker or
advance important research. Optimized systems running independent software
vendor (ISV) application code like Fluent or Nastran, for example, is very
important. Linux Networx offers Total Cluster Management from one interface,
technical training and professional services, all of which are key to
delivering a highly productive computing system for maximum sustained
performance.
|