Visit additional Tabor Communication Publications
February 11, 2009
With a focus on system manageability and the science its users do on their hardware, Penguin Computing is charging hard at the small- to mid-sized cluster end of the HPC business. Once focused on the Linux enterprise, the company has spent the past five years growing its HPC business, starting with the acquisition of Scyld Software, and most recently bringing on former IBM executive Charles Wuischpard as its CEO. HPCwire talked with Wuischpard to get an idea of where the company is today, and what makes it tick.
Penguin Computing is a privately-held company based in San Francisco, specializing in Linux-based clusters for high performance computing and enterprise computing. Penguin was founded in 1998 by Sam Ockman (who is also one of the authors of Open Sources: Voices from the Open Source Revolution) and was one of the first companies to build a server business solely based on Linux.
Today Penguin Computing has a broad base of traditional HPC and enterprise customers, including Fortune 1000 firms and Web companies that you've heard of, like Facebook and Digg. Their hardware includes the suite of AMD- and Intel-based clusters that you would expect in this part of the market. In 2003 Penguin acquired Scyld Software, the cluster management software company created by Donald Becker. Becker, who is currently CTO of Penguin, has long been a part of the Linux community, and helped found the Beowulf computing effort along with Thomas Sterling while he was at NASA.
Wuischpard describes the Penguin of today as "radically different" than it was even just two years ago. He joined the company after a diverse career in IT that included 16 years in sales at IBM. When he came to Penguin in January 2007, its business was all based on "inside sales" (guys in the office selling computers over the phone) and comprised of sub-$50,000 orders. Today, 70 percent of Penguin's sales are of systems that cost $200,000 and up, and the approach to sales is much more in line with the rest of the HPC business. Wuischpard says that the company built five systems last year that would have qualified for the TOP500 -- had the customers allowed the systems to be submitted (they demurred, citing competitive advantage).
Penguin's niche is in providing turnkey, engineered HPC systems for customers who are relatively new to HPC. They've been down the path of building their own clusters, and have proven to themselves that HPC has real value. Now they want to grow that capability, without building all of the infrastructure and expertise it takes to successfully build larger and larger systems on their own.
The acquisition of Scyld was part of a strong push by Penguin to grow its presence in high performance computing. "Scyld's software added commercial-strength provisioning and management software, making our clusters substantially easier to use for customers than if they built their own systems with some of the free software out there, like Rocks," says Wuischpard. He describes Penguin's focus in HPC as being from the application down, not from the hardware up. "We win business because we understand the science that is being done first." Over the past several years the company has hired discipline-specific researchers to make sure that its cluster designers have a deep understanding of the computational requirements of their customers and are building systems to meet their needs.
With a few exceptions, today's HPC market is based on mostly interchangeable hardware. Penguin knows this, and builds its value for customers on integrating that hardware with a Linux stack and management software that make the systems productive for their customers. As an example of its success in this area, Penguin's Scyld Clusterware solution is the reference image for the SSI cluster [PDF] in the Intel Cluster Ready. Of course, Penguin isn't the only vendor to offer a provisioning and management solution, and others, like SGI (with its Industrial Strength Linux Environment), are focusing on this as well.
Wuischpard describes the company's relatively small size as a key advantage for customers. "Our core engineering team is small, but the entire company is oriented around Scyld as a differentiator. We are all focused on working with our customers to improve the software, and we can take unfiltered requirements directly to engineering."
Penguin also emphasizes the reliability and post-installation stability of its machines, something you'll read about in articles going back to the company's founding. Penguin ties all of its manufacturing and order support systems into a single company-wide process that manages the build process for customer systems. "Everything in this process is based around making sure a system is buildable -- and will work as expected for our customers -- before we accept the order," says Wuischpard. All systems are also assembled and burned in at the factory before they are shipped out for installation at customer sites, an activity that benefits customers and provides an ideal training ground for Penguin's new customer support engineers. Of course, as Penguin goes after much larger ("hero class") HPC business, it will have a harder time affording the infrastructure to do a full system burn in prior to shipment.
Penguin has business in a number of markets, but its leading customer segments in HPC are in the traditional biggies: CAE and manufacturing, life sciences, and government business with DoD and the intelligence communities. Refreshingly, its is also careful about how it grows, consciously avoiding going after big business in the DoE, for example, that might bring significant risk to a company its size.
One of the benefits of this business strategy is that Penguin's customer base is broad enough to provide some insulation in the face of economic shocks and Wuischpard describes 2008 as a record year. "Revenues were up 25%, margins were up 30%, and operating expenses were down 28%." This is almost sure to change as the economic crisis deepens, but this is a pretty good position to be in right now. In fact, if changing market conditions create a realignment in the HPC vendor community, Penguin is one of the companies I will be looking for to step up and become a more significant force in HPC.
May 22, 2013 |
At some point in the not-too-distant future, building powerful, miniature computing systems will be considered a hobby for high schoolers, just as robotics or even Lego-building are today. That could be made possible through recent advancements made with the Raspberry Pi computers.
May 16, 2013 |
When it comes to cloud, long distances mean unacceptably high latencies. Researchers from the University of Bonn in Germany examined those latency issues of doing CFD modeling in the cloud by utilizing a common CFD and its utilization in HPC instance types including both CPU and GPU cores of Amazon EC2.
May 15, 2013 |
Supercomputers at the Department of Energy’s National Energy Research Scientific Computing Center (NERSC) have worked on important computational problems such as collapse of the atomic state, the optimization of chemical catalysts, and now modeling popping bubbles.
May 10, 2013 |
Program provides cash awards up to $10,000 for the best open-source end-user applications deployed on 100G network.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
In this demonstration of SGI DMF ZeroWatt disk solution, Dr. Eng Lim Goh, SGI CTO, discusses a function of SGI DMF software to reduce costs and power consumption in an exascale (Big Data) storage datacenter.
The Cray CS300-AC cluster supercomputer offers energy efficient, air-cooled design based on modular, industry-standard platforms featuring the latest processor and network technologies and a wide range of datacenter cooling requirements.