Last year, NVIDIA helped the Energy Department build the fastest supercomputer in the world — Summit at Oak Ridge National Laboratory. The supercomputer runs on over 27,000 NVIDIA GPUs, which power more than 90 percent of the system’s performance. Summit allows the world’s top scientists to harness AI to drive discovery and to work on daunting challenges like addiction patterns and fusion energy.
“I am continuously amazed by what we’ve seen the government accomplish in the areas of AI and accelerated computing, though there is much more to be done,” said Anthony Robbins, vice president of the North America public sector.
NVIDIA has also taken a leading role in educating the federal government on the vast possibilities of AI. Its leadership testified before a U.S. congressional committee last February and participated in a White House AI Summit in May to advocate for the importance of funding AI research, driving agency AI adoption and opening access to data.
“It has been an honor helping policymakers across the federal government understand this revolutionary technology and how it can improve government,” Robbins said.
NVIDIA also has driven education through the largest AI conference in the D.C. area — NVIDIA’s GPU Technology Conference, or GTC D.C. More than 2,600 registered for the event, which informed federal leaders of AI possibilities and provided data scientists and government technologists access to AI resources.
CPUs have been an important part of the federal government’s processing engines for decades. As a result of Moore’s Law coming to an end, the federal government needs alternatives for tomorrow’s data and information. NVIDIA’s GPUs offer an option better suited to process the government’s rapidly growing and valuable pools of data.
“There’s been tremendous progress in AI and its deployment across enterprises of all kinds,” Robbins said. “As it relates to adoption of AI in the federal government, this is a large-scale effort and transition.”
For example, Robbins said that applying AI to government projects will save taxpayer dollars, transform health care, improve citizen services and help keep the nation at the forefront of innovation.
To address some of the challenges AI presents, NVIDIA focuses on education through initiatives like NVIDIA’s Deep Learning Institute, which has already trained more than 120,000 data scientists with practical techniques to harness AI.
AI was a key theme at NVIDIA in 2018 — one that will only expand. Robbins said this theme will also play out across the whole federal government and expects it to accelerate in the years to come.
“To progress on its AI journey, the federal government will need to continue to address at least four things, he said. “Government needs more well-trained data engineers and data scientists, better access to data, improved infrastructure both on-prem and in the cloud, and use cases and proof points reflecting early wins. NVIDIA is helping the government in each of these areas. One of the most important themes for this coming year is helping the government move faster.”
Why Watch:
The government’s AI journey has begun. The adoption of AI into the government enterprise of the future is likely going to be the largest technology transformation ever undertaken. This effort will continue for more than a decade and will require help from the industry, including large federal systems integrators like Lockheed Martin and Booz Allen Hamilton and cloud providers like AWS and Microsoft Azure, as well as commercial companies, startups and universities like Carnegie Mellon.
“In 2019, we plan on working across this ecosystem to help the government with its overall change and transformation effort, enabling innovation at greater speed,” Robbins said.
AI will touch all aspects of government, ultimately reducing costs and improving citizen services.
Today, we see AI from the data center to the edge and in use cases for health care, cybersecurity, platform sustainment, and autonomous systems and robotics. In 2019, NVIDIA plans to continue bringing innovation around GPUs, CUDA, software development kits and frameworks to government, combined with training via the DLI — all at scale through its ecosystem in service of one of the largest markets in the world.