“Taming Big Data can be challenging,” says Nadia Short, Vice President and General Manager, Cyber Systems, General Dynamics Mission Systems. “The challenge is that much of the actionable patterns and insights reside in unstructured data and the traditional linear scale computing and storage was no longer sufficient to process them,” she adds. With its more than three decades of experience and expertise in the field of HPC, General Dynamics Mission Systems deploys new forms of HPC configurations to unlock insights housed in unimaginably large stores of unstructured data. The company utilizes HPC’s multiple compute units and storage servers that are interconnected using fast and efficient networking – InfiniBand – and leverages technologies such as Apache Hama, an open source analytic framework that maximizes the HPC system architecture design to facilitate better data movement.
Due to our flexibility, rethinking and open architecture (OA), General Dynamics Mission Systems is a leader in next generation prototype development
General Dynamics Mission Systems combined HPC with Big Data by porting the Apache Hama graph analytic framework to an HPC InfiniBand Cluster. The open source analytic framework with its distributed programming model is famous for dividing individual tasks for individual computing units in an HPC environment. “However, this linear model fails to query against a huge set of data hampering the speed of the overall process,” says Nadia. In wake of such an obstacle, General Dynamics Mission Systems reengineered Apache Hama from the distributed programming model to the parallel programming model. The new model divided the single task into many small tasks and assigned those to different computing units so that the task gets executed in short time. “By optimizing Apache Hama, we exploit the absolute speed and power of HPC,” extols Nadia. General Dynamics Mission Systems achieved 10 times more performance than traditional systems while analyzing an 8 million vertices graph derived from Twitter data—which is a best example of High Performance Data Analytics (HPDA). There lies great value in the ability to access large data sets, but there is greater value if it’s done at extreme speed. It would keep the data analyst engaged in the problem, prompting more and varied queries. This allows the data analyst to test new methods, leading to new insights into the data sets. General Dynamics Mission Systems uses its Internal Research and Development (IRAD) to create afore mentioned HPC optimization solutions. “As we bring our expertise into the HPDA arena, we are leading the way towards the next generation of HPC technologies,” says Nadia.
Fuelled by an understanding of HPC’s potential to reduce ‘time to insight’ and accelerate time-to-market, government enterprises and commercial organizations from engineering firms to geosciences companies are expressing interest in HPC.
Additionally, as HPDA can perform real-time analytic computations on live data leading to real-time intelligence, the performance level would enable a new concept of operations in Activity Based Intelligence (ABI) and similar missions. “Accelerating discovery and insight can change the game when decisions and actions are made based on intelligence enabling organizations to respond and execute on dynamic and critical mission requirements,” states Nadia. “We are a leader in next generation HPC prototype development that eases the burden of data intensive tasks,” she adds. Due to the company’s flexibility, rethinking and open architecture, General Dynamics Mission Systems supports enterprise missions by architecting and integrating solutions to optimize the client’s existing HPC systems.
Partnering with industry leaders to stay abreast with HPC technology
“For General Dynamics Mission Systems to provide relevant, mission-specific solutions, it is imperative that we understand the HPC landscape-the key players, developers of next-generation technology,” mentions Nadia. General Dynamics Mission Systems has developed a viable solution to this obstacle, GDNexus. Through GDNexus, a community of technology providers, General Dynamics Mission Systems is able to better understand the technologies and innovation that HPC providers are delivering today. GDNexus connects technology experts with customer missions to create innovation, best-in-industry solutions.
” HPDA is the marriage of Big Data with HPC that promises to keep pace with the high volume of constantly changing data, transforming it into a competitive and mission advantage”
Customers benefit in a number of ways through these partnerships. Not only do they get access to mission relevant technologies but the flexible model accelerates identification and delivery of solutions. Additionally, customers can reduce technology acquisition risk as they can test drive the performance and effectiveness of proposed solutions in a controlled environment that imitates the real ones. “GDNexus helps us maintain our responsiveness; ensuring that customers have what they need to execute their goals successfully,” she adds.
The company has also partnered with key HPC vendors to further its inroads into HPC innovations. As a systems integrator, General Dynamics Mission Systems’ job is to create a solution by pulling together a team of partners who add value to the solution. One such company is Data VortexTM, which features advanced computing systems built around the revolutionary, patent-protected Data Vortex Network Switch and Vortex Interface Cards (VICs). Designed and optimized for data movement and real parallel computing, Data Vortex systems enable better performance on applications that require massive data communications by eliminating network congestion thus keeping the processors fully utilized. Other such partnerships range across U.S. defense, intelligence, cyber and homeland security communities. “Partnering with technology providers like this is a key success factor for us. Our focus is on what our customer needs to accomplish their mission and we will bring relevant resources to bear,” cites Nadia. “We’re an honest broker, teaming with other companies to get solutions to customers more quickly and more cost-effectively,” she adds.
“As HPC and Big Data continue to converge, we will continue to invest in the future through smart IRAD projects and strategic partnerships,” says Nadia. “The strides we have made in HPDA is just the beginning. We are leading the way towards next generation HPC for our customers and their missions.”