“Enterprises are having considerable difficulty and incurring extraordinary expense adapting to the changing and evolving small file and mixed workloads dominating the HPC and AI landscape,” says Robert Murphy, Director of Product Marketing, Panasas, Inc., in an exclusive Hotseat Interview with Enterprise Talk.
ET Bureau: How has the relevance of plug-and-play data storage appliances changed in the current pandemic scenario?
Robert Murphy: One of the most striking observations occurred at the beginning of the most severe lockdowns in late March of 2020. That also coincided with many Life Sciences organizations immediately pivoting their research to studying COVID-19. During that time, these organizations needed to quickly add more storage to cope with the increased specimen analysis throughput and storage requirements, in some cases increasing by two orders of magnitude overnight.
These organizations’ campuses were off-limits to all but essential internal personnel, so when new Panasas storage appliances were delivered, trained Panasas personnel could not enter the site to install the equipment. Panasas’ plug and play ActiveStor Ultra data storage appliance allowed the customers to install these systems, on their own, without vendor assistance. Even today, the border between the US and Canada is still closed to non-essential traffic, and customers in Canada are able to self-install these plug and play appliances to get their business back to full productivity.
Read More: The Grave Side of Big Data Mistakes and Approaches to Prevent Them
ET Bureau: How are enterprises adapting to the changing and evolving small file and mixed workloads dominating today’s and tomorrow’s HPC and AI landscape?
Robert Murphy: Enterprises are having considerable difficulty and incurring extraordinary expense adapting to the changing and evolving small file and mixed workloads dominating the HPC and AI landscape. For those that can afford it, a few enterprises are deploying very expensive All-Flash storage to deal with the small-file onslaught. Which solves the problem, but it is very, very expensive.
And for those who are more cost-conscious, enterprises are deploying hybrid flash-HDD based systems that deploy a small capacity of flash in a “hot” tier for working data and migrate it to a “cold” tier for non-working set data. This hot-cold migration limits the available performance to a very small flash tier and incurs management headaches and potential data loss as data needs to be moved back and forth between tiers.
ET Bureau: What, according to you, is the biggest roadblock to the adoption of Dynamic Data Acceleration technology?
Robert Murphy: The beauty of Dynamic Data Acceleration (DDA) technology is that there are absolutely no roadblocks to adoption. Dynamic Data Acceleration is completely automatic, and adaptively maximizes the performance of any workload. DDA intelligently places new files on the optimal storage media based on predicted usage and size thresholds that adjust dynamically as the workload evolves.
Read More: Nurturing Big Data with the Power of Artificial Intelligence
ET Bureau: How has HPC data storage solutions helped startups map their business growth journeys?
Robert Murphy: A good example of the benefits of HPC data storage solutions is observed across the Life Science startups that need to analyze growing amounts of genomic information very quickly, but also need to do so cost-effectively.
These startups can get their products (vaccines, drugs, reagents, etc.) to market much quicker, and at less cost, with greater efficacy and safety, by simulating molecular, biochemical, and biomedical interactions computationally, using HP data storage solutions, as opposed to relying exclusively on “in real life” research in wet labs that is a more expensive and slower approach.
Robert Murphy leads global product marketing for Panasas ActiveStor storage solutions and the PanFS parallel file system. Robert has more than 30 years of experience in technology marketing with a focus on data storage and data management software platforms. Prior to joining Panasas, Robert held the position of Big Data Program Manager at General Atomics, an energy and defense corporation specializing in research and technology development, and prior to that, he was responsible for high-performance data and analytics in IBM’s Software-Defined Environments organization. Other appointments include various product marketing and product management positions with increasing levels of responsibility at Hewlett Packard, Silicon Graphics, Oracle, and Sun Microsystems. Robert holds a Biomedical Engineering degree from Purdue University.
Debjani Chaudhury works as an Associate Editor with OnDot Media. In this capacity, she contributes editorial articles for two platforms, focusing on the latest global technology and trends.Debjani is a seasoned Content Developer who comes with 3 years of experience with Fashion, IT, and International Marketing industries. She has represented India in International trade forums like Hannover Messe, Germany.
