APAC life sciences firms face HPC challenges

Pharmaceutical and biotech companies in Asia-Pacific are accelerating genomics research, but their high-performance computing (HPC) infrastructure that supports workloads such as precision medicine and drug discovery remains a work in progress.

According to an IDC study commissioned by Lenovo, 54% of 150 respondents in Japan, Korea, Singapore, Thailand and India expect more than 10% growth in their genomic workloads in the next two years.

However, just over 50% are at a startup stage with respect to their genomics HPC infrastructure, having less than a year of experience. A third indicated having one to three years of experience, and only 15% have more than three years of experience.

Some life sciences firms are also facing challenges with their HPC infrastructure, such as the lack of scalability of existing HPC systems, too much customisation and complexity, undesirable pricing models, as well as support issues and high support costs.

As such, about half of respondents are planning to acquire new HPC systems, with 25% expecting to do so within three to six months, and 10% in less than three months. The systems include next-generation sequencing (NGS) machines, which are being acquired by those in more developed and better-funded markets to improve research productivity.

However, an NGS machine with the ability to create 1 TB of data a day will also increase the storage requirements for genomic data dramatically. Currently, about half of APAC respondents have a genomic data storage capacity of 1-5 PB, and a fifth have a capacity of more than 5 PB.

IDC noted that the inability to scale storage capacity could pose other challenges, such as regulatory compliance in a highly regulated industry such as life sciences.

Sinisa Nikolic, Lenovo’s infrastructure solutions group director and segment leader for HPC and artificial intelligence in Asia-Pacific, said: “The volume and type of genomics data generated is unimaginable and to make accurate decisions based on this data requires huge computing power. This gets even more difficult with complex and unscalable solutions, which organisations across Asia-Pacific ranked as top challenges when looking for genomics solutions.”

To speed up genomics research, some research institutions like the Genome Institute of Singapore (GIS) have turned to public cloud services to access HPC resources.

In 2015, it started using Amazon Web Services (AWS) to make sense of Singapore’s group B streptococcus outbreak at the time and analyse genomics data in a cost-effective and scalable manner, said Swaine Chen, group leader at GIS’s laboratory of bacterial genomics.

“We manage our own compute environment, but our ability to scale as an institution bears no comparison to what we can access after moving over to AWS,” Chen said. “It also allows us to focus on the biology and collaborate with researchers in other parts of the world.”

According to IDC’s study, almost three-quarters of respondents are participating in a national, regional or global genome initiative, reflecting global efforts to leverage genomic data for game-changing innovation in healthcare and life sciences.