Nvidia is building a supercomputer specifically designed to help with the increasingly computationally complex area that is medical imaging.
Unveiled at Nvidia’s annual GPU Technology Conference in San Jose this week, Clara — named after Clara Barton who founded the America Red Cross — was labelled by Nvidia founder and CEO Jensen Huang as “a datacentre-virtualised, multimodality, multi-user, medical computational medical instrument”.
With around 3 million instruments installed in hospitals throughout the world, and 100,000 new ones installed each year, Huang predicts it will take about 30 years before the installed base is replaced with upgraded equipment.
“The question is how do we solve this problem? We can’t wait for 30 years for doctors to be able to have this early detection and for doctors to have this technology in their hands,” he said. “Because of our technology, it’s possible for us to virtually upgrade every single medical instrument.”
According to Huang, supercomputing is now a fundamental pillar of science.
“We need larger computers; even then we need larger computers; the world needs larger computers, because there’s serious work to be done, there’s serious ground-breaking work to be done,” he said.
“Science needs super-charged computers and that’s the reason why we’re building super-charged computers.
“Our strategy at Nvidia is to accelerate GPU computing at the speed of light.”
Speaking with ZDNet about the future a machine like Clara promises, Nvidia VP of healthcare and AI business development Kimberly Powell said the GPU giant has been focusing on healthcare for over 10 years, alongside the medical imaging industry, which has been driven by computing for at least as long.
“If you think about what medical imaging does, it does computation, it does processing, digitalisation, and artificial intelligence,” she said.
As Huang detailed, Nvidia started by performing image reconstruction inside instruments and serving up the volume rendering and image processing required for the interpretation of such images.
For the past five years, Nvidia has also worked with the research and startup communities exploring the future of AI from a healthcare lens.
“We’re at a very fortunate time in medical imaging in that all the technology Nvidia has been developing over the past five years, we can now literally virtualise medical imaging pipelines, these medical imaging instruments,” she explained.
“We can take data that may be coming from a 10-year-old machine and apply computation to it remotely and be able to do all of this really innovative computation, essentially bringing the capability to every instrument that’s on the planet.
“And so that’s what Clara is; it’s really a platform for doing this medical imaging supercomputing workload.”
While Clara is still being built by Nvidia, what the supercomputer will be capable of is no pipedream.
At its most basic, Clara is a computing platform that allows the virtualisation of workloads — a virtualised datacentre for all the computation that happens in medical imaging.
“It could be an on-prem resource, it could be on AWS, and it will leverage the technology stack that we’ve put together over the years,” Powell continued. “So of course it has virtualised GPU, it has Nvidia GPU containers, and it containerises applications, and eventually it will make use of Kubernetes on GPUs to efficiently manage these microservices over the resources that are needed.”
According to Powell, the dream for Nvidia is to provide Clara to regions where radiologists or specialists, for example, aren’t as common, or where the high-level training that is required to become a radiologist isn’t as accessible as it is in many Western regions.
“One of the ways we think about it is, let’s make sure it’s everywhere accessible to everyone and as quickly as we can bring these tools and put it into the hands of the people that can use it,” she said. “You can spin up a very cheap instance in the cloud to get access to it, you could buy gaming GPUs to get access.”
Pointing to AlphaGo, Powell said that upon reflection, in the first three games the computer beat the human — but in the fourth game, the human beat the computer because he was taught by the computer something he didn’t always know how to do.
“So it is not an elimination thing, this is an augmentation thing,” she said. “I think we all get better by technology and so this technology is giving the ability to see things and they can learn from that and build that into their own internal brain and neural network and experience.
“I’m completely speculating, but you could imagine you might become an expert in 10 years’ time versus 30 years’ time, because you’ve had the computer be able to point these things out.”
Discussing some potential use cases, Powell said she came across a case recently that saw 23,000 unread X-rays in the UK; similarly, she said in Japan there’s a very “disorienting number” of instruments and medical exams that are performed when placed in a ratio against the number of practicing radiologists.
“This, I think, is an opportunity to balance some of that out,” she said.
Powell said being able to sequence genomes requires a lot of data and computational power to process such data, and that the data is only growing thanks to the increasing cost efficiency of genomics.
“And so the pharmaceutical companies, they’re all trying to figure out how do we integrate this data and perform some intelligent analysis on top of it to get on that journey to true precision medicine,” she explained.
“Now we’re reading everything about the body that’s happening on a daily basis — that data integration I think is going to transform the pharmaceutical industry and totally transform the healthcare continuum, because you have a lot more information about a person before they became symptomatic.
“It’s going to be so powerful and that digitisation … consolidation where we want to get all that data in one place is going to require a new computing architecture and a new computing paradigm — a lot of it coming from software.”
Powell expects more of Clara to be unveiled at the Radiological Society of North America conference in November.
Disclaimer: Asha McLean travelled to GTC as a guest of Nvidia
MORE FROM GTC 2018
The processing power Nvidia’s GPUs are capable of has made them a useful tool for those dealing in ether, but the company’s CEO would prefer the GPUs be kept for use in areas such as gaming or high-performance computing.
The Drive Constellation simulation environment allows for sensor data to be processed as if it were coming in from sensors on a real car cruising the streets.
At its annual GPU conference in San Jose, the company announced that the Nvidia Quadro GV100 GPU with Nvidia RTX technology will provide artists of all kinds with access to the tech usually reserved for cinematic productions.
At its GPU Technology Conference in San Jose, CA, Nvidia announces hardware and interconnect upgrades, data center GPU inference improvements and a new IoT partnership with ARM.
Announced at GTC 2018, the partnership will see the Nvidia Deep Learning Accelerator architecture integrated into Arm’s Project Trillium.