Sumit Mukhija, CEO, ST Telemedia Global Data Centers India on the power of AI and its impact on the data center industry
Artificial Intelligence (AI) stands as a multifaceted enigma, sparking diverse perspectives and definitions.
Many have a distorted view of AI that does not match its current usage. The idea of existential threats, also known as X-risks among AI experts, evokes scenes from popular sci-fi movies, feeding to the fear that people have.
Despite AI captivating global imagination and driving a wave of digital innovation, a crucial question arises: are the foundational technologies within data centers evolving at a pace that matches this transformative wave?
Sumit Mukhija, CEO, ST Telemedia Global Data Centers India (STT GDC), the Indian arm of Singapore-based STT GDC, provides valuable insights into the profound implications of AI and its impact on the data center landscape.
Interpreting AI’s essence
Sumit opens the dialogue by unraveling the core of AI.
To him, AI represents an area where systems, technologies, and algorithms converge to execute tasks that traditionally required human intelligence. The breadth of AI’s influence extends far beyond the mere replication of human functions. He envisions AI as a key that unlocks possibilities previously deemed unattainable.
“From streamlining repetitive tasks to catalyzing revolutions in healthcare and education, the potential of AI is boundless,” Sumit says.
AI remaking data centers
Transitioning seamlessly from the theoretical to the practical, Sumit expounds on AI’s tangible impact on data centers. The accelerated pace of innovation in servers, storage, and networks has ushered in a radical shift in the data center landscape.
Even from a deployment perspective these components have undergone a change, right from centralized main frames to de-centralized client server environments, back to decentralization to hybrid deployments, where some servers and applications were placed on-prem and some went to the cloud environments.
“All this has led to some design implication on way we design, build and operate data centers,” Sumit says.
With a focus on the surge in processing power, particularly in Graphics Processing Units (GPUs) essential for AI, Sumit sheds light on the challenges faced by traditional deployment models. The shift towards higher power densities per rack, soaring to unprecedented levels like 70-80 kilowatts for AI, necessitates a meticulous reevaluation of technology choices, deployment strategies, and operational methodologies.
“These changes are making us rethink, change and reselect the technology choices that we make in the data centers, the way they are deployed, and the way SLAs are delivered,” Sumit says.
The conversation pivots to the formidable task of redesigning data centers in the face of escalating power densities. Sumit candidly discusses the substantial challenges posed by this shift. From the replacement of conventional air handling units with fan wall units to the embracing liquid-cooled setups tailored for GPU deployments, each solution presents its unique set of challenges. Liquid cooling, while efficient, introduces complexities such as mechanical systems within data halls, potential service impacts, and security risks.
“The ongoing challenge lies in striking a delicate balance between innovation and practicality,” Sumit says.
AI’s nuanced impact on jobs
The discourse gracefully segues into the hot-button issue of AI’s impact on employment. Sumit navigates the landscape by acknowledging the potential for job losses in certain areas easily automated by AI.
However, he emphasizes that countless jobs reliant on human qualities like empathy, intuition, and real-time decision-making remain irreplaceable.
“The crux lies in reskilling and adapting. AI, when viewed as a tool complementing human capabilities, paves the way for a harmonious coexistence of humans and machines,” Sumit says.
Envisioning AI’s broader impact on society
As the conversation draws to a close, Sumit widens the lens to consider AI’s broader societal impact.
Beyond economic and job-related aspects, he underscores the need for a careful examination of the social, ethical, cultural, and religious dimensions of AI. Issues such as misinformation, data privacy, and ethical concerns demand responsible navigation from the entire ecosystem.
“It’s a delicate balance between embracing innovation and safeguarding values, requiring a balanced approach to ensure AI’s positive impact on society while mitigating potential risks,” Sumit says.
