Transformative Solutions for Accelerating AI, Analytics and Deep Learning at NVIDIA #GTC19

Print Friendly, PDF & Email

One pivotal message received by attendees of this week’s NVIDIA GPU Technology Conference (GTC) in Silicon Valley is the importance of game-changing storage solutions and applications that empower users to accomplish their most challenging AI objectives.

Take DataDirect Networks (DDN®) for example. The company featured and demonstrated innovations in its A³I® architecture and new customer use cases in autonomous driving, life sciences and healthcare, retail and financial services. DDN also today announced a partnership with Robovision.

The company showcased the next generation of A³I reference architectures, which include NVIDIA’s DGX POD™, DGX-2™, and DDN’s AI400™ parallel storage appliance. Through industry standard benchmarks and customer use cases using the new reference architectures, DDN demonstrated how it establishes higher levels of application acceleration when combining the power of GPUs with parallel storage systems. Featured use cases included testing results from the Max Delbrück Center (MDC) for Molecular Medicine in conjunction with Zuse Institute Berlin (ZIB). MDC is a leading molecular biology and clinical research center in Germany, and ZIB is an interdisciplinary research institute for applied mathematics and data-intensive high-performance computing. MDC originally partnered with DDN in 2018 to greatly accelerate its applications and to efficiently manage large volumes of data. Using an A³I architecture with an NVIDIA DGX-1 system and an AI200™ from DDN, MDC was able to improve accuracy and precision within analyzed images while simultaneously accelerating training performance by 240 percent. More recent testing with an A³I architecture comprised of NVIDIA’s DGX-2 system and DDN’s AI400 revealed results that more than doubled previous improvements.

Our main discovery using DDN hardware was noticing its ability to scale,” said Alf Wachsmann, head of scientific computing department at MDC.  “In our testing, we observed ‘native’ performance of the DGX-1 and DGX-2 for really large data sets, which was not possible without the A³I hardware. ‘Native’ performance means that we see all eight DGX-1 and 16 DGX-2 GPUs fully utilized, thus, we are able to make perfect use of these resources and speed up our deep learning workload from weeks to days.”

DDN to Demonstrate Frictionless-Retail Solution with XXII

DDN featured a frictionless-retail experience in a live, interactive booth demonstration at NVIDIA #GTC2019.  Frictionless retail employs concepts such as autonomous check-out and real-time analytics to improve the shopper experience from creating seamless, in-store interactions to enabling real-time customer service. XXII, a leading provider of real-time video analysis solutions, worked with DDN to bring the seamless in-store experience to the GTC show floor. Using proprietary computer vision algorithms, the XXII solution allows customers to shop in stores without any friction and without having to wait in lines.

We strongly believe in the power of AI to transform our lives concretely,” explains Dam Mulhem, co-founder of XXII. “With DDN’s impressive storage solutions in the cloud and on premise, we are privileged to partner with the DDN team to help us achieve our goals. DDN is also one of the leaders of AI-oriented calculators for inference purposes, with high storage capacities and high-volume video management, and is assisting us with overcoming challenges in that arena as well.”

DDN Partnership with Robovision

With many customers expressing concerns around implementing AI successfully, DDN is excited to partner with Robovision and its self-service deep learning platform, which enables organizations to bring concrete, maintainable AI applications live in mere weeks.

Recent testing in our lab of the combination of NVIDIA’s DGX-2 along with DDN’s AI400 has reinforced our view that infrastructure considerations are fundamental to the realization of business transformation and life-improving AI applications,” said Jonathan Berte, CEO at Robovision. “DDN’s commitment to developing highly parallel and scalable architectures matches well with the high performance requirements of compute and AI applications.  We look forward to upcoming projects with DDN and leveraging them to help customers extract the most value from their complex data.”

New DDN PartnerLink Program Tier

During the GTC conference, DDN launched a new Platinum Tier within the DDN PartnerLink program, exclusively for NVIDIA DGX-certified resellers of DDN A³I with NVIDIA DGX-1, DGX POD, and DGX-2. This specialized tier is designed to provide added support to the resellers who work closely with AI and deep learning customers and best understand their needs, and to support them with the highest quality of service.

The new tier will provide resellers with exclusive offerings from DDN, including access to proof of concept units for qualified opportunities, special pricing for demo units, guaranteed margins for registered deals, and special pricing on DDN A³I with NVIDIA DGX systems.

DDN’s renewed focus on the channel over the last three years has strengthened our relationships with both customers and partners. We are excited about offering the Platinum Tier and to continuing to differentiate our channel program with focus and expertise,” said Dean Koester, senior vice president of field sales for Americas and Europe at DDN.

DDN Presentation at GTC:  Edge to Core: A Meta Study of Data Complexity in AI

On Tuesday, DDN’s Dr. James Coomer, senior vice president of product, presented “Edge to Core: A Meta Study of Data Complexity in AI,” which discussed the challenges uncovered in AI and deep learning workloads, the most efficient approaches to handling data, and use cases in autonomous vehicles, retail, health care, finance and other markets. The presentation covered the complete requirements of the data life cycle including initial acquisition, processing, inference, long-term storage, and driving data back into the field to sustain continuous improvement. Dr. Coomer also provided examples of data life cycles in production that trigger diverse architectures from turnkey reference systems with NVIDIA DGX systems and DDN A³I to tailor-made solutions.

Contributed by Daniel D. Gutierrez, Managing Editor and Resident Data Scientist for insideBIGDATA. In addition to being a tech journalist, Daniel also is a consultant in data scientist, author, educator and sits on a number of advisory boards for various start-up companies. 


Sign up for the free insideBIGDATA newsletter.


Speak Your Mind