Digital Transformation

Aligning Cloud Costs With Sustainability and Business Goals

At AWS Empower India 2024, Dr. Werner Vogels, chief technology officer of Amazon, shared the importance of aligning cloud costs with business goals, introducing the seven laws of "The Frugal Architect" to developers and technologists. Vogels underscored the need to build resilient, cost-effective systems from the outset, especially as cloud technologies evolve around innovations such as generative AI.

See Also: Endpoint Security Essentials for the C-Suite: An Executive's Digital Dilemma

"We've lost the art of architecting for cost. As builders, we need to start thinking about cost from the start," he said. The seven guiding principles of "The Frugal Architect" include:

1. Making cost a non-functional requirement;
2. Optimizing both cost and business alignment;
3. Architecting a series of trade-off measures;
4. Ensuring unobserved systems do not lead to unknown costs;
5. Implementing cost controls in cost-aware architectures;
6. Approaching cost optimization incrementally;
7. Recognizing that unchallenged success can lead to assumptions.

These principles, supported by real-world examples, provide businesses with a blueprint for maintaining a balance between cost, security and performance in today's cloud-driven environment.

Vogels said cost, alongside security and performance, is now a critical factor in system design. "Cost and sustainability should be treated as equally important as other non-functional requirements," he said.

WeTransfer's 78% reduction in carbon footprint is a case study on how cost-conscious architectures can drive sustainability, Vogels said. "The Frugal Architect turns constraints into innovation. Any costs that are ignored will lead to debts and financial burdens for the enterprises."

AWS' Approach to Economies of Scale

AWS provides a range of pricing options, giving customers the flexibility to select a plan that meets their specific workload needs. Shalini Kapoor, chief technologist for APJ public sector and director at AWS India and South Asia, discussed AWS' commitment to cost reduction through economies of scale. "It's our responsibility to lower costs for our customers. The more customers we have, the more we can reduce costs," she said.

"With our global network of regions and data centers, as of Sept. 20, 2023, we have reduced the cloud costs by 134 times since its launch in 2006. Our goal is to continue architecting solutions that bring down the total cost of ownership for our customers, allowing them to reinvest savings into innovation."

Kapoor said AWS also provides transparency by displaying service costs on the platform dashboard. "Cost is a non-functional requirement we must meet, but the biggest challenge is skills. People need to understand how to choose the right service," she said. AWS is investing heavily in skill development to democratize these technologies.

Three-Layered Gen AI Stack

AWS has developed a three-layered gen AI stack, aimed at making AI more affordable and accessible. Given that gen AI models are compute-intensive and require significant resources, customers are keen on reducing inferencing costs at the chip level. To address this need, AWS invested in Inferentia and Trainium well before the gen AI boom.

AWS is one of the largest procurers of NVIDIA GPUs, which form the foundation of its infrastructure, including GPUs, Inferentia and Trainium, providing flexible options for customers.

"Both Inferentia and Trainium are proprietary AWS chipsets, developed following its acquisition of Annapurna Labs. By being mindful of the expenses associated with running AI, we aim to help customers manage their budgets better," Kapoor said.

AI is poised for democratization, similar to the cloud. Users will have the choice and ability to use multiple models for numerous use cases. Future trends indicate a rise in culturally aware and industry-specific models that will further facilitate the democratization of AI.

Singapore's National Research Foundation launched AI Singapore - a national program to enhance the country's AI capabilities - to make its LLMs more culturally accurate, localized and tailored to Southeast Asia. AWS is working with Singapore public organizations to develop innovative, industry-first solutions powered by AI and gen AI, including AI Singapore's SEA-LION. Building on AWS' scalable compute infrastructure, SEA-LION is a family of LLMs that is specifically pre-trained and instruct-tuned for Southeast Asian languages and cultures.

AWS released the Amazon Bedrock managed service to support gen AI deployments for large enterprises. It now provides easy access to multiple large language models and foundation models from AI21 Labs, Anthropic, Cohere, Meta and Stability AI through a single API, along with a broad set of capabilities organizations need to build gen AI applications with security, privacy and responsible AI.

AWS also plans to offer a complete infrastructure for building gen AI applications. This offering will include a broad choice of vector databases, including Redis Enterprise Cloud, Pinecone, Amazon Aurora and MongoDB.