Why High-Performance Computing Matters to Healthcare

Though unequivocally delicious, one thing chocolate cake isn’t considered is  healthy, right? Actually, depending on your gut’s microbiome, it may be healthier than fresh fruit according to Research. Researchers from Israel’s Weizmann Institute of Science studied the effect of different foods on 800 people to learn more about the body’s glycemic response mechanisms.

Raw data including demographics, height,  weight, BMI index,  sleep and exercise patterns, as well as blood and gut microbiome  was collected for the study. In all, over 50,000 meals were assessed, and over 1.5 million glucose measurements were collected. This project—with its massive datasets—was made possible by high-performance computing (HPC), a form of supercomputing that is able to digest staggering workloads and provide lucid results. For this study, researchers concluded that people respond differently to various foods so that indeed, for some people, cheesecake may be healthier than a grapefruit!

What unites these independent Research and studies  is their shared requirement to process Big data. The potential of HPC brings a tsunami like innovation to the Healthcare Industry and the opportunity to Transform Healthcare is clear,  but for your HPC infrastructure to run reliably and smoothly, it must be built on storage that can keep up with very high compute speeds. If storage is slow or unavailable, research will slow down, stop altogether, or—worst of all—data could be lost, destroying years of work.

Even if the HPC began to boom well before cloud computing with the development of huge supercomputers starting in the ’90s, the rise of cloud computing greatly democratized HPC. For example, a researcher studying a new material does not need to send a proposal to a supercomputer facility to calculate the electronic properties of its highly correlated material, he can just fire up 200 spot instances on his AWS account, run the workload, download the results, and finally switch off the cluster. All this for just a few bucks.

Going away with the lengthy and frustrating proposal review process needed to access a supercomputer facility, the scientist can obtain the same results in a fraction of the time with minimal effort, and act more quickly on the results: should a new experiment find out a slightly different composition with better properties than the original sample, the researcher can simply rerun the analysis overnight without any additional overhead.

Having the compute power you need, exactly when you need it, and pay only for what you use has a huge value in these situations. High Performance systems provide a fast, reliable, cost-effective storage solution that helps keep your healthcare operations running as fast as your data scientists’ projects evolve. You can seamlessly scale systems to accommodate a vast amount of data, using a granular, building-block approach to growth. You can auto-scale with a pay-as-you-go strategy, adding one or more storage drives  at a time. Along with speed, reliability, and elastic scaling, the system helps reduce your Total cost of ownership because it helps bring down infrastructure costs.

It is a bit more like a collaboration not a competition and the particularity in the health system and medicine, is that we have all these new forms of data, genomics, their social information, and we need these tools to help us synthesize the data and the information that the doctor can use or can use as a patient. So, instead of being afraid of this technology, it should empower us. We must see this as a tool and technologies of power, but taking care of who owns the data (what we share, how or when a smartwatch is used) and how they are sharing it, but these technologies can give a much better vision of personal health, and even helping doctors and nurses make smarter and more personalized diagnoses and therapies.