Scalable storage becomes vital for complex models and extensive datasets as generative AI expands: Sameer Bhatia, Seagate Technology

Sameer Bhatia, Director of Asia Pacific Consumer Business Group and Country Manager for India & SAARC, Seagate Technology,

In an exclusive interview with CRN India, Sameer Bhatia, Director of Asia Pacific Consumer Business Group and Country Manager for India & SAARC, Seagate Technology, shares his insights on the profound impact of generative AI on storage products and solutions. Bhatia emphasizes that as companies strive to train their AI tools and accelerate their AI offerings, the need for mass-data storage will take center stage. To meet this demand, businesses will require robust data storage strategies that effectively handle both unstructured and structured data. Furthermore, Bhatia highlights how enterprises in India can achieve business resilience by implementing an efficient and comprehensive data storage strategy. Read the full interview

How effective data management strategy can help businesses to thrive in the AI-led future?’
Data is essential for AI, powering successful machine learning and accurate predictions. AI algorithms rely on ample data, making data storage crucial. According to a recent IDC report, 84% of enterprise data created in 2022 was useful for analysis, but only 24% of it was being analysed or fed directly into AI or machine-learning algorithms.

This means companies are failing to tap 60% of available data for analysis. That’s lost business value. If the data is not stored, not even the smartest of AI models can help. It’s like having an electric car in your garage and needing to drive somewhere: if the battery isn’t charged, the car won’t get you where you need to go. The bigger and better the data set, the better-trained the AI model will be—delivering clearer insights and more business value. Scalable storage becomes vital for complex models and extensive datasets as generative AI expands. Storing input data and generating insights enhances business value and future utility.

Organisations are transitioning to an Edge-with-Cloud architecture in 2023, how is it adding to the relevance of multi-cloud environments in India?

The adoption of the Edge-with-Cloud architecture is a representation of how multicloud environment in India is diversifying. Multicloud relates to deploying multiple cloud environments (public or private) from more than one cloud vendor. With unprecedented data growth, companies are implementing multicloud strategies for seamless access to data and to avoid risk of a single point of failure.

However, data growth is outstripping storage capacity and network bandwidth and competing in today’s data economy is predicated on storing, mobilising, and analysing it at unprecedented volume and speed — all in an increasingly distributed IT ecosystem. Companies are taking edge computing as a cloud deployment approach that brings their applications closer to the internet of things (IoT) devices and necessary servers. By placing servers closer to users, businesses can quickly turn around large amounts of data.

With the dawn of 5G in India, how will organisations transition to an Edge-with-Cloud architecture in 2023 in order to keep up with IT infrastructure?
The transformative potential of 5G technology across sectors is substantial. It will offer accelerated data speeds, reduced latency, and heightened connectivity, facilitating seamless real-time transfer of substantial data volumes. The surge in device and data adoption driven by 5G will establish IT infrastructure as a premium location—equivalent to prime beachfront property—for edge computing. The Edge-with-Cloud model will seamlessly integrate, process, and store data closer to its origin. 5G’s speed and the edge model’s proximity will jointly reduce latency.

5G will propel the demand for infrastructure edge computing, spurring investments in servers, storage, and networking. Through embedding computing resources at the infrastructure edge, 5G networks can efficiently capture real-time data, cost-effectively processing it for insights, in contrast to discarding data in the cloud. This entails boosting computing resources at edge locations. Realising 5G’s full potential necessitates prioritising edge data, significantly improving application performance, and enabling real-time processing of extensive data volumes.

What is Seagate doing to enhance its partner ecosystem in India to the next level?
We attribute our success in India to our partners and strive to empower them with essential knowledge and tools. Our partner programme focuses on tailored incentives for sales initiatives and customer acquisition, catering to specific industry needs. We support our partners with optimal storage solutions and services, and one great example is our award-winning Seagate SkyHawk Partner App for video and imaging applications. This app aids partners in real-time storage calculations, online warranty service booking, and connectivity to SeaCare centers.

Additionally, our Seagate Partner Programme connects with customers and partners through training sessions, seminars, and co-marketing initiatives. These activities help us to gather insights, reinforce our market leadership, and channel growth in the storage, software, and system industries.

In your experience, what are the key considerations for enterprises when it comes to choosing a storage solution for their data?

A robust data management model begins with DataOps. It combines the processes of data analytics and the work of data engineers with a variety of data sources to make results clear for business users—the “data consumers.”

Enterprises gather a huge amount of data. An organised, rigorous set of business processes is vital if large companies hope to sort through data from thousands or even hundreds of thousands of consumers. Without a set of defined operations, information sits idle in a data lake, where it takes up valuable and expensive space without adding any real value to company operations. A set of clear enterprise data operations standards makes it easy to govern a company’s gathered information. Instead of a loose, disorganised pile of information, data operations help manage gathered information and take advantage of it in a useful, meaningful way.

With data operations at its core, enterprises can define how data is collected, integrated, transformed, stored, and utilised within their enterprise framework. When designing distributed storage systems, leaders must prioritise governance and interoperability. Identifying business goals and insights shapes intelligent data collection, guiding data type choices.

How does your company address the challenges of scalability and data growth when it comes to storage solutions for enterprise clients?

We efficiently address organisational scalability and data expansion concerns by providing customised storage solutions that seamlessly adapt to changing data needs. Our product line includes mass-capacity data storage systems, enterprise hard drives, and SSDs, which offer the best value with industry-leading capacity, firmware, and multi-core capabilities.
The intelligent storage management features in our products optimise resource utilisation. For example, data tiering automatically shifts frequently used data to faster tiers and less-used data to less-expensive storage, creating a balance between speed and efficiency.
Our solutions contribute to cloud-ready designs and technologies which improve scalability even further. We support enterprises to effortlessly extend storage into the cloud, allowing for uninterrupted data expansion. This interface enables dynamic data growth and flexible scalability on demand.

What measures does your company take to ensure high availability and data redundancy in storage solutions for enterprise clients?

We ensure exceptional availability and data redundancy in our enterprise-focused storage solutions through a comprehensive blend of hardware and software strategies. These measures work cohesively to minimise downtime and protect data integrity. On the hardware front, we leverage advanced RAID (Redundant Array of Independent Disks) configurations that distribute data across multiple drives, establishing redundancy in the event of drive failures. This architecture not only enhances data availability but also accelerates data recovery. Additionally, our enterprise storage solutions feature hot-swappable components, enabling hassle-free replacement of faulty drives without disrupting ongoing operations.

We also leverage data mirroring and snapshot-based backups. Data mirroring duplicates data across distinct drives or systems, ensuring continuous access even if one copy becomes inaccessible. Snapshot-based backups capture precise data replicas at specific time points, enabling swift restoration to desired instances, invaluable for addressing data corruption or user-induced errors. Furthermore, we also take advantage of predictive analytics and monitoring tools to proactively identify potential hardware issues, pre-empting service disruptions.

How do Seagate’s data migration and seamless integration of storage solutions into existing enterprise IT infrastructure?

Our expertise in data migration and seamless integration of storage solutions into existing enterprise IT infrastructure is facilitated through a highly focused approach. We use innovative data migration tools to enable the secure and effective transfer of data from outdated systems to new storage environments. This entails meticulous planning, minimising downtime, and ensuring data integrity throughout the process.
Our storage solutions are designed to integrate seamlessly with diverse IT configurations, whether on-premises or cloud-based. Our solutions effortlessly integrate with the current infrastructure of our clients by concentrating on open designs and interoperability, boosting storage capacity without disturbing operations. This strategy allows them to maximise their IT investments while adopting cutting-edge storage technologies.

What kind of storage management and monitoring tools or features do you provide to enterprise clients for efficient and effective data management?

While data creation is growing exponentially, IT teams and budgets are not. Our best example of how we keep this in mind when we design our products is Seagate Exos CORVAULT. Exos CORVAULT is an intelligent storage that utilises Autonomous Drive Regeneration (ADR) to minimise downtime, human intervention, and e-waste by renewing errant drives “in situ.” That means fewer opportunities for human error and more cost savings.

Can you explain how your storage solutions accommodate different types of data, such as structured, unstructured, and semi-structured data, for enterprise clients?

Our storage solutions seamlessly accommodate a range of data types, including structured, unstructured, and semi-structured data, effectively meeting diverse enterprise requirements. Structured data, characterised by its orderly arrangement, is efficiently managed within our systems, ensuring swift retrieval and processing. Unstructured data, devoid of a predefined structure, is stored, managed, and accessed while maintaining its integrity. Our mass-capacity storage systems centered around hard drives, will continue to provide significant TCO advantages for unstructured data as HDD technology breakthroughs make ever-higher capacities possible.

We are living in the early days of a generative AI boom. As companies look to train their AI tools and expedite their AI offerings, mass-data storage will become the focal point. Businesses will need robust data storage strategies that encompasses unstructured and structured data. Businesses boost their commercial value when they save both the raw data that is fed into AI systems and the resulting insights for future use.

They should look to the cloud for some of their AI workloads, and they will also want to store some of their data on their premises. There’s a good reason why public cloud is made up of roughly 90% hard drives and 10% flash storage. Hard drives are a cost-effective, high-capacity, durable, and reliable solution built for massive data sets. The revolution happening now in hard drives’ areal density advancements means that the economy of scale that had been available to hyperscale data centres is also an option for companies’ private data centres. Hard drive storage can be used to store the vast data fed back into AI models for continuous training. Anchored on the areal density advancements, our solutions remain adaptable, empowering enterprises to extract value from all data types.


Please enter your comment!
Please enter your name here