Precision Medicine pushes demand for HPC at the Edge: AI on the Fly ® Delivers

In this special guest feature, Tim Miller from One Stop Systems writes that by bringing specialized, high performance computing capabilities to the edge through AI on the Fly, OSS is helping the industry deliver on the enormous potential of precision medicine.

Source: Shutterstock

Advances in high performance computing equipped with rapid diagnostic tools, advanced imaging devices, and genetic sequencers are enabling a growing trend toward precision medicine where healthcare is personalized based on an individual’s genetic make-up. Artificial intelligence can deliver transformative insights that enhance and accelerate this trend. This is a dramatic example of how high performance computing at the edge can improve the quality of life of millions of people around the world.

To enable this revolution, the most powerful, high performance computing technologies historically associated with centralized enterprise and cloud data centers need to move out to the edge, often embedded directly in specialized medical devices or co-located with the primary data sources in the field at hospitals and clinics.

Three key elements are fundamental in many of these edge devices; from MRI or CT imaging devices, to genetic sequencers and cell analysis systems. First, there is the requirement to acquire and shape massive amounts of high speed incoming data. Second, edge devices need to efficiently store this raw data and then move it to compute and analysis engines. Third, these engines then transform the transferred data to actionable intelligence. All of these capabilities need to be designed and integrated to meet the specialized size, power, and environmental constraints of the edge based application.

Increasingly, the power required in these compute engines is being delivered through NVIDIA Tensor Core GPUs where each NVIDIA GPU has thousands of computational cores that can process massive data in parallel, as well as the NVIDIA end-to-end software solution stack. An illustrative example is secondary genomic analysis used in precision medicine. Modern genomics involves the rapid production of vast amounts of raw sequencing data using next-generation sequencers (NGS) coupled with massive computing for secondary analysis that converts the data into useful results. The most popular toolkit for doing this secondary analysis is GATK4 Best Practices pipeline. Traditionally, this work has been done by large numbers of CPUs. Parabricks, a company recently acquired by NVIDIA, has changed the paradigm and developed a GPU based solution that executes genomic analysis best practices workflows on NVIDIA GPUs. Parabricks is built using NVIDIA CUDA-Z and benefits from CUDA, cuDNN and Tensor TR inference software and is now available through the NVIDIA GPU Cloud (NGC), a software hub for accelerated computing applications. Deploying this capability integrated with sequencing platforms in the hospital and clinic environment is a critical requirement for making precision medicine a reality.

One Stop Systems, Inc. (OSS) has developed the technology and expertise required to work with OEMs to build this next generation of highly intelligent medical devices ready for field deployment. In fact, its AI on the Fly ® platforms and building blocks are being used by OEMs to build and deliver these medical solutions to the market today. AI on the Fly puts computing and storage resources for AI and HPC workflows, not in the datacenter, but on the edge near the sources of data. Applications are emerging for this new paradigm not only in precision medicine, but in diverse areas including autonomous vehicles, battlefield command and control, and industrial automation. The common elements of these solutions are high data rate acquisition, high speed low latency storage, and efficient high performance compute analytics. With OSS, all of these building block elements are connected seamlessly with memory mapped PCI Express interconnect configured and customized as appropriate, to meet the specific environmental requirements of ‘in the field’ installations.

OSS building blocks include high slot count PCI Express expansion systems capable of acquiring 100 GB per second of data, NVMe storage nodes providing up to a petabyte of high speed, low latency solid state storage, as well as platforms capable of housing up to 16 of the latest NVIDIA GPUs or other HPC/AI accelerators for high end compute engine requirements. OSS also provides the customization capabilities to provide all of these elements in unique form factors or designed for specialized environmental and ruggedized conditions.

OSS is working with genomic sequencer OEMs to embed its data acquisition, storage, and compute engine platforms directly in genomic analysis solutions deployed in medical facilities worldwide. In 2019, OSS worked with Parabricks to benchmark genomic analysis results on its compute accelerator utilizing 16 NVIDIA V100S GPUs in parallel. When executed on the OSS platform, execution speed for secondary analysis was increased by a factor of 40X with a full human genome analysis reduced from taking two-days to taking less than one hour. OSS is working with other medical OEMs to deploy AI on the Fly solutions in live cell analysis systems which captures and analyzes images for cell therapy, oncology, and immunology research. Additional deployed applications include robotic eye surgery and high speed blood analysis systems.

By bringing specialized, high performance computing capabilities to the edge through AI on the Fly, OSS is helping the industry deliver on the enormous potential of precision medicine.

Disclaimer: This article may contain forward-looking statements based on One Stop Systems’ current expectations and assumptions regarding the company’s business and the performance of its products, the economy and other future conditions and forecasts of future events, circumstances and results.

Tim Miller from OSS

About the Author

Tim Miller is Vice President of Strategy at One Stop Systems. Tim has over 33 years of experience in high tech operations, management, marketing, business development, and sales. He previously was the CEO of Dolphin Interconnect Solutions and CEO and founder of StarGen, Inc. Tim holds a Bachelor of Science in Engineering from Cornell University, a Masters of Business Administration from Wharton, and a Masters in Computer Science from the University of Pennsylvania.