With the rise of AI projects, corporate sustainability is increasingly under strain. As AI and ML transform industries, the demand for data storage and computational power is growing at an unprecedented pace.
However, Pure Storage claims to be an exception and has a solution. Sudharsan Aravamuthan, head of systems engineering at Pure Storage, emphasised the company’s commitment to attaining sustainable innovation.
“As we scale from 48TB, we moved to 75TB, and now 150TB, our power envelope has remained the same, almost like one watt per terabyte,” he said. This suggests that the company has managed to improve its efficiency in scaling storage without increasing the power consumption proportionally.
According to enterprise-grade data flash storage provider Pure Storage, data centres alone consume approximately 4% of global power, with storage accounting for 20% of that consumption.
Organisations need to expand their infrastructure to keep up with demand while keeping sustainability in mind. This makes finding smart and eco-friendly storage solutions more important than ever.
Scaling Without More Consumption
The company faces direct competition in the all-flash storage market, with giants like IBM, Hewlett Packard Enterprise (HPE), NetApp, and Dell EMC. However, the company aims to ship 300TB DirectFlash Modules (DFMs) by the end of 2025.
“In 2025, when the industry may be shipping a 61.44TB SSD in volume, we’ll be shipping a 300TB DFM (which, by the way, will enable 12PB in 5U),” says Eric Burgener, director of technical strategy at Pure Storage.
Pure Storage claims its approach to sustainability places it ahead of competitors with its well-established benchmarks in efficiency and density.
Innovations, such as zero move tiering, direct flash technology, and a new entry-level performance model, S100, aim to balance performance with sustainability. This gives customers the flexibility and performance they need while staying budget-friendly.
Zero move tiering enables data tiering without physically moving the data. This reduces operational delays and energy costs. It is yet to be seen if this unique service provided by the company truly sets a new standard or merely addresses a niche problem.
In response to growing demands for smaller and performance-oriented solutions, Pure Storage is set to launch the S100 model, which will cater to smaller datasets while maintaining high performance.
This aligns with the company’s commitment to provide tailored solutions across diverse customer requirements.
Expanding Innovation in India
India is emerging as a critical player in the global data ecosystem and is generating nearly 20% of the world’s data. This was also earlier stated by Bhavish Aggarwal, CEO of Ola, in a conversation about how India is exporting most of its digitised data to companies abroad.
With the arrival of Industry 4.0 and increasing connectivity across the nation, the need for seamless, scalable, and secure data infrastructure is now greater than ever.
Initiatives like Open Network for Digital Commerce (ONDC) and Ayushman Bharat are reshaping industries, calling for strong back-end systems capable of delivering consistent customer experiences.
Pure Storage claims to lead this transformation with its cutting-edge solutions to address these challenges. Through its ‘Enterprise Data Cloud’ concept, the company aims to ensure that data remains accessible and secure. This approach aligns with the increasing need for self-secure infrastructures to tackle rising cyber threats.
India’s predominantly wireless infrastructure allows for the swift adoption of these innovations without the burden of legacy systems.
Pure Storage’s advancements, including its purity operating system and direct flash technology, aim to reduce costs, improve power efficiency, and scale to meet data demands.
The company’s expansion in India reflects its commitment to make use of top-tier talent for global innovation. The R&D centre, established in Bengaluru in 2022, is already marking its third anniversary with significant growth, as the company reported.
Bengaluru was chosen for its unparalleled concentration of systems talent, and the centre has embraced a “global leadership” approach, where local teams lead critical projects like FlashBlade and Portworx.
The Infrastructure Crunch
The rapid adoption of AI has exposed the inadequacies of existing infrastructure. “Customers told us, ‘I need to double my computing power,’ but 85% said their existing infrastructure would not support it,” Aravamuthan noted.
This is because users would need to connect to a faster network. The company claims to address this challenge by “solving through densification without compromising on performance”.
The increasing reliance on inference, training, and test-time computing for AI initiatives highlights the infrastructure challenges many organisations face. AI initiatives ranging from training LLMs to deploying real-time inference systems, demand robust systems that can handle vast datasets while maintaining efficiency.
As AI-driven initiatives grow, companies are shifting from traditional data storage to ecosystems supporting inference, training, and test-time compute. GPUs, now central to AI deployments, demand costly infrastructure upgrades for optimal performance.
This shift highlights the need for innovative solutions. Traditional approaches often lead to increased downtime, higher costs, and greater energy consumption – challenges that are difficult to reconcile with sustainability goals.
The healthcare, finance, and telecommunication industries rely on AI for use for fraud detection and customer analytics. Organisations must unify datasets, streamline pipelines, and ensure efficient data cleansing to extract actionable insights.
Sustainability Through Localisation in India
When AIM asked the company about upcoming initiatives in sustainability, the company said, “Our vision is to actually reduce carbon footprint, and we know that data centres contribute to 1-1.5% of the world’s carbon footprint. If we can impact that, it would be a great milestone.”
As India emerges as a hub for data centres, creating sustainable infrastructures that align with global environmental goals is very crucial. Pure Storage aims to minimise the need for cross-continental data transfers by localising data storage solutions in India. This further lowers the environmental impact.
Traditional SSDs integrate with application stacks by relying on multiple layers of power-intensive components. Pure’s storage solutions are built directly on NAND technology, eliminating inefficiencies, reducing power consumption, and lessening the need for additional hardware layers.
What’s next? Organisations aiming to survive in the AI landscape must focus on creating systems that are not only powerful and efficient but also aligned with environmental goals. Sustainable innovation is no longer an option; it is imperative.
The post Pure Storage to Ship 300TB DirectFlash Modules by the End of 2025 appeared first on Analytics India Magazine.
