While cloud services are becoming more sophisticated and customizable, they may fall short in handling the “data avalanche” that the life sciences industry is grappling with.
Data sets that are too large and complex for traditional data processing methods to handle are a sticky issue in the sciences, perhaps no more so than in the life sciences industry. From biotechnology to pharmaceuticals to medical device manufacturers, private-sector organizations and research institutions spend their days performing research and development tasks that produce terabytes of raw data gleaned from clinical results, disease states, scientific studies, and individual patient data.
This “data avalanche” must then be sorted, analyzed, and distilled into actionable information so that new medications, biomedical devices, and other products can be developed, refined, tested, and approved for sale as quickly as possible. Because of the gravity of the decisions that depend on this information — human lives are literally at stake – life sciences companies operate under some of the closest scrutiny and strictest regulations in the world.
To cope with the data avalanche and compete in an increasingly hypercompetitive global marketplace, many life sciences organizations are considering cloud computing solutions to increase their data processing power and storage capabilities. However, cloud solutions can often fall short in big data processing capabilities and result in cybersecurity issues, eye-popping monthly bills, and other problems.
Big Data = Big Cloud Bills
Cloud providers are quick to tout the alleged cost savings of using their services over purchasing in-house IT infrastructure. They claim that customers do not have to come up with large capital investments upfront and pay only for what they use, as they go. This works out well for some companies, particularly small start-up firms that are cash-poor and that aren’t dealing with very large data sets or highly complex computations.
However, cloud services such as AWS are notorious for sky-high monthly bills filled with hidden “gotchas,” especially for companies that require a lot of computing power. Deep Value, which develops complex research-driven trading algorithms, decided to run some numbers once their AWS bills began to exceed $70,000/month. In the end, they discovered that using AWS was 380% more expensive than purchasing their own high-performance computing equipment.
In addition to unexpected line items on their monthly invoice, cloud customers can also be hit hard by the indirect costs of performance problems and cyberattacks.
Cloud Performance May Not Be Up to Par
In the life sciences, the reliability and uptime of mission-critical systems are paramount, and a key marketing point of cloud services is the idea that customers don’t have to worry about maintaining their own equipment. Yet as cloud computing grows in popularity, cracks are appearing in its foundation. In February 2017, AWS suffered an outage that was so bad, it couldn’t get into its own systems to communicate with the throngs of customers that were knocked offline – all due to a misconfiguration on the part of an AWS employee.
The cloud doesn’t necessarily beat in-house infrastructure in the performance category, either, especially when processing enormous data sets. Cloud service providers typically run multiple servers in different locations, which can cause very serious latency issues when transferring large data sets and performing the highly complex calculations that life sciences companies run all day long.
The Dark Cloud of Cybersecurity Concerns
Over the past few months, an epidemic of AWS breaches impacted organizations large and small, including Verizon, the Republican National Committee, and a company called Talent Pen that processed job applications containing the personal information of thousands of Americans who held Top Secret security clearances. All of these breaches were due to the affected organizations (or their third-party vendors) not having configured their AWS security settings properly.
Cloud security settings can be very tricky, but even if an organization gets them right, they can still be hacked through no fault of their own. Because so much valuable data from so many different organizations is being migrated to the cloud, data centers have become highly attractive targets for hackers. Financial regulators and the tech industry are so concerned about the possibility of a major attack on AWS that in the wake of the February outage, they sounded the alarm over what they deem an over-reliance on the AWS service by the organizations that form the bedrock of American society, particularly financial companies.
The cloud security risk to life sciences organizations is three pronged: a cyberattack that brings down their cloud could leave them unable to operate; invaluable market research and digital intellectual property could be stolen by competitors or foreign governments; and a hack could mean running afoul of a myriad of government regulations and being hit with millions of dollars of fines and lawsuits.
In-House IT Infrastructure Means Optimum Customization & Control
Finally, cloud services suffer from customization and control issues. When using a cloud service, the ability to make changes is quite limited. Services such as AWS offer a menu of items that fit most organizations’ needs – unless those needs are highly specialized. Organizations that own their own computing equipment have complete control over their data environments and can act quickly to make adjustments or implement new features.
Rather than rushing to migrate to the cloud, certain life sciences organizations might be better served by investing in their own high-performance computing equipment to reduce their costs, improve their cyber security, and make the data avalanche work for them to accelerate innovation. Only a true comparison between the two can shed light on which option works best.
