Visit additional Tabor Communication Publications
October 16, 2008
Bob Graybill, whose high-profile roles have included heading the DARPA High Productivity Computing Systems (HPCS) program and working with USC-ISI, the Council on Competitiveness and half a dozen big defense contractors, is now CEO and president of a start-up. Nimbis Services aims to expand HPC use in manufacturing by brokering cycles, storage and expertise.
HPCwire: What was the rationale for founding Nimbis Services?
Bob Graybill: The company's direction is based on more research data and market experience than most new companies have at their disposal. That includes three years of comprehensive studies, IDC surveys and conferences funded in part by DARPA, DOE, NNSA, and NSF that were led by the Council on Competitiveness and University of Southern California Information Sciences Institute. The research included in-depth discussions with the Council's HPC Advisory Committee comprised of senior industry, academic, laboratories and government representatives.
HPCwire: What was the gist of the findings?
Graybill: There's a huge benefit from using HPC to do modeling and simulation, or what we call digital analysis computing, in order to accelerate competitiveness. Larger companies have been benefiting from this for years, but there's a huge gap between the P&W and Procter & Gambles of this world, and their supply chains. The studies identified major barriers preventing smaller firms and business units from moving beyond their desktop computers to exploit HPC. The barriers include the cost of acquiring HPC systems and ISV software, as well a lack of expertise in how to use HPC resources. The studies also showed that in larger companies it's not unusual for users to go outside of their divisions to get HPC resources because the resources within the division are often fully utilized.
HPCwire: What happens when HPC isn't available?
Graybill: In companies of any size, when people don't have access to HPC they typically do one of three things, none of them ideal. They might just ignore the problem, or they might downsize the problem so it can run in a reasonable period of time on a desktop computer. That allows you to run a lot more jobs, but it could limit the quality of the product because you're not modeling fully. Sometimes people also revert to more physical experimentation if they can't model the problem on the desktop. But building multiple physical prototypes is significantly more expensive from an engineering, manufacturing and time-to-market perspective. It's also much more labor intensive, so it disadvantages countries like the U.S. that have higher labor rates.
HPCwire: Did the company name come from Harry Potter's turbo-broomstick?
Graybill: Both names come from the Latin nimbus, meaning "cloud." Ours is the plural form, nimbis. We wanted to find a unique, memorable name. I hope we succeeded.
HPCwire: You've been in HPC leadership positions in government and academia. Why did you decide to pursue the goal of helping to boost industrial competitiveness though a private company?
Graybill: Our first approach was to try a non-profit model like MOSIS, which USC-ISI has run over the past 25 years to aggregate multiple semiconductor chip designs onto a single wafer run, but we got strong feedback from this target industry segment that they wanted to deal with an organization that has a real stake in the game. This suggested a for-profit approach. The business model Nimbis Systems adopted is based on discussions with major OEMs, their suppliers, and ISVs.
HPCwire: Which part of your background was most helpful for your role with Nimbis Services?
Graybill: I'm not sure if there was any single one. It was a combination. There was my Lockheed-Martin experience managing large engineering organizations and the challenge of consolidating and reorganizing groups for business reasons. Probably the most relevant experiences were DARPA and the Council on Competitiveness. As a DARPA program manager, you were expected to come up with new programs and present them like a VC pitch, and you'd come away with either a check or no check. At DARPA I worked with a very diverse community of academic researchers, industry thought leaders and senior government representatives. It was a fantastic experience. The Council also gave me an opportunity to work with senior folks in academia, industry and the labs but from a slightly different perspective: competitiveness and ROI instead of R&D. I got a good feel for what drives their decisions and the factors that might make them interested in a potential product or service.
HPCwire: What is the business model for Nimbis?
Graybill: To be a business-to-business brokerage or clearinghouse. The idea is to provide pre-negotiated access to cycles, software and expertise on an on-demand, pay-as-you-go basis. We won't own any equipment or do consulting ourselves. We're simply a clearinghouse that builds a menu of quality services and then brings the buyers and the sellers of those services together. Our targets are periodic and experimental users, initially in the manufacturing sector. These are people who don't want to jump over huge hurdles to get the benefits of modeling and simulation using HPC. We're an aggregator of services. We also help our partners, our service providers, by reaching out to a brand new community on their behalf.
HPCwire: What is your menu of services today, and how do you expect that to grow?
Graybill: Our initial focus is primarily in the manufacturing area, including computing cycles, storage, and ISV software. Over the next year or two, we plan to expand to other areas, and from a capabilities perspective we plan to adopt a more automated approach to connecting to expertise, almost like a business social network.
One thing that's clear from the past three years of in-depth surveys and heavy interactions with users and vendors is that there is no one simple solution to learning how to expand the use of modeling and simulation analysis. It takes an ecosystem, and Nimbis will be an enabler in this. We recently announced a partnership with the Ohio Supercomputer Center. OSC was one of the first HPC academic centers to recognize that there is a gap between high-end users and desktop scientific and engineering users. OSC and their Blue Collar Computing initiative are a perfect fit with Nimbis' strategy. They provide regional outreach to companies in their area, including compute cycles and expertise. They provide a path so that these users can migrate, if they wish, to longer-term production cycles from Nimbis. So, OSC provides transitional outreach. OSC benefits from Nimbis Services' ability to provide on-demand commercial licenses as driven by their industry R&D partners.
We are in the process of formalizing relationships with HPC centers in other parts of the country. It's important to have good geographic coverage because companies have shown a strong preference for working with HPC centers in their own regions.
HPCwire: What are the advantages for HPC services buyers and sellers of working with a company like Nimbis?
Graybill: Buyers usually have a short time window to come up with a solution to a problem. They have a limited budget and are not planning any in-house digital analysis computing capability. We lower the barriers so they can try out the software packages and run them on different cycle providers. We will give them a choice of ISVs, cycle providers and whatever additional expertise they need in a one-stop-shopping experience through a single Web portal. Without this, it might take them many months to investigate one element of the solution at a time. With Nimbis, they can turn the faucet on and off as they wish. We're about on-demand computing. Nimbis also gives potential users an opportunity to explore the use of modeling and simulation beyond the desktop without investing in the hardware and the software licenses.
For sellers, Nimbis is a new marketing channel, a feeder channel for potential long-term users.
HPCwire: What other business relationships can you talk about?
Graybill: We're about choice rather than exclusivity, so we welcome discussions with any potential partner that can provide quality HPC services within our format and requirements. As our Web site shows, as of today we have agreements for cycles and storage with IBM, Amazon Web Services and R-Systems. We also have agreements for R&D cycles, storage and consulting services from OSC. We have arrangements with the National Center for Manufacturing Science for expertise in commercializing products, with Wolfram Research for ISV and consulting services, and with Decision Incite for consulting services. We're far along in discussions with additional ISVs and HPC centers. We expect to add a lot more ISVs in the next 12 months.
HPCwire: The Council on Competitiveness studies you referred to concluded that lowering the barriers to HPC use for desktop users would boost U.S. industrial competitiveness. Is that part of Nimbis' mission?
Graybill: Boosting U.S. industrial competitiveness isn't part of our mission as a private, for-profit company, but if we succeed we expect that to happen along the way. Nimbis is creating a focal point for connecting buyers and sellers in the digital analysis computing community. We're an enabler that can help service providers to grow and users to become more competitive. We are trying to provide a way for everyone to grow.
HPCwire: Will Nimbis offer services to non-U.S. organizations?
Graybill: Our partner network of cycle providers is U.S.-based, and anyone can use these services. We're talking with non-U.S.-based ISVs with the aim of making their products and expertise available through our U.S.-based cycle providers.
HPCwire: What's next for Nimbis? Will we see you at SC08?
Graybill: We plan to announce additional partnerships with ISVs and HPC centers in the coming months. We'll enter beta status in the first quarter of 2009 and offer beta opportunities early next year. That means we're looking for beta users now. We'll go into production in the first half of 2009. One ideal type of beta user is a partnership that includes a motivated OEM, a supplier, and an ISV with a specific application they want to address right away. Another ideal combination is a tier 1 firm with a supplier and an ISV. A third combination is an HPC center and a supplier.
At SC08, Nimbis will be featured in partner booths and we'll have one or more demos available in partner booths.
HPCwire: Is there anything important I didn't ask about?
Graybill: As I look back over the past three years and the Council on Competitiveness and government survey findings, what strikes me is that the democratization of HPC, or digital analysis computing, really requires the engagement of all parties within the ecosystem: academia, HPC centers, ISVs, OEMs and users. We see Nimbis as one of the tools that will help accelerate the use of digital analysis computing and help more competitive products to be developed. There is a role everyone needs to play, and we want to be a key facilitator to make this happen. We want to be a rallying point. We're not a whole solution, but an enabler. The immediate need is to lower the barriers, and then the ongoing business model is to provide these services and expertise to businesses that need it for the long haul.
Jun 18, 2013 |
The world's largest supercomputers, like Tianhe-2, are great at traditional, compute-intensive HPC workloads, such as simulating atomic decay or modeling tornados. But data-intensive applications--such as mining big data sets for connections--is a different sort of workload, and runs best on a different sort of computer.
Jun 18, 2013 |
Researchers are finding innovative uses for Gordon, the 285 teraflop supercomputer housed at the San Diego Supercomputer Center (SDSC) that has a unique Flash-based storage system. Since going online, researchers have put the incredibly fast I/O to use on a wide variety of workloads, ranging from chemistry to political science.
Jun 17, 2013 |
The advent of low-power mobile processors and cloud delivery models is changing the economics of computing. But just as an economy car is good at different things than a full size truck, an HPC workload still has certain computing demands that neither the fastest smartphone nor the most elastic cloud cluster can fulfill.
Jun 14, 2013 |
For all the progress we've made in IT over the last 50 years, there's one area of life that has steadfastly eluded the grasp of computers: understanding human language. Now, researchers at the Texas Advanced Computing Center (TACC) are utilizing a Hadoop cluster on its Longhorn supercomputer to move the state of the art of language processing a little bit further.
Jun 13, 2013 |
Titan, the Cray XK7 at the Oak Ridge National Lab that debuted last fall as the fastest supercomputer in the world with 17.59 petaflops of sustained computing power, will rely on its previous LINPACK test for the upcoming edition of the Top 500 list.
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.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?
Join our webinar to learn how IT managers can migrate to a more resilient, flexible and scalable solution that grows with the data center. Mellanox VMS is future-proof, efficient and brings significant CAPEX and OPEX savings. The VMS is available today.