Posts Tagged ‘AWS’
[AWSReInvent2025] High-Performance Storage Architectures for AI/ML, Analytics, and HPC Workloads
Lecturer
Aditi is a Senior Product Manager for Amazon FSx at Amazon Web Services (AWS). With years of experience working directly with customers on high-performance workloads, she focuses on pushing the technical boundaries of what is possible with cloud storage to meet the demands of modern compute-intensive applications.
Abstract
This article examines the critical role of high-performance storage in supporting modern AI/ML, analytics, and High-Performance Computing (HPC) workloads. As organizations scale their compute resources—incorporating hundreds or thousands of CPU and GPU cores—storage often becomes the primary bottleneck, preventing linear performance scaling. We explore the technical architectures of Amazon FSx and Amazon S3, focusing on how these services address the needs of both “lift-and-shift” file-based applications and “cloud-native” S3-based data lakes. By analyzing customer use cases in genomics, media rendering, and large language model (LLM) training, we detail the methodologies for achieving peak performance at scale.
The Storage Bottleneck in Compute-Intensive Workloads
Modern high-performance workloads are characterized by their extreme reliance on massive datasets and high-core-count compute clusters. In an ideal cloud environment, adding more compute resources should lead to a proportional increase in work completed—a concept known as linear scaling. However, traditional storage solutions often fail to keep pace with the throughput demands of these clusters, leading to a performance plateau.
When storage becomes the bottleneck, compute instances sit underutilized as they compete for access to the same data store. This is particularly detrimental given that 90% to 95% of the expenditure for these workloads is typically allocated to compute resources. Consequently, an inefficient storage layer not only extends the time to insight but also significantly increases the total cost of ownership (TCO). To avoid this, storage must be architected to scale linearly alongside compute.
Navigating the Path to the Cloud: File Systems vs. Object Storage
Organizations generally approach high-performance storage on AWS from two distinct backgrounds: those with long-standing on-premises file-based workflows and those who have built native cloud applications around object storage.
The Persistence of File-Based Architectures
Despite the rise of object storage, file systems remain the preferred interface for many researchers and developers due to three primary factors: Familiar Interface: The intuitive nature of files and directories simplifies complex data management for data scientists and developers.
* Granular Permissions: File systems provide robust POSIX permissions, allowing for fine-grained control over which users can read, write, or execute specific files.
* Consistent Data Access:* For workloads where multiple users or compute nodes access the same data simultaneously, the strong consistency of file systems ensures that all parties see the most recent data updates.
Amazon FSx for High-Performance File Access
Amazon FSx addresses these needs by providing fully managed file systems that offer the performance of local storage with the scalability of the cloud. For “lift-and-shift” scenarios, FSx allows organizations to move their existing HPC and AI/ML pipelines to AWS without refactoring their applications.
Accelerating Generative AI and ML Workloads
The emergence of generative AI has placed a renewed emphasis on data strategy. Whether an organization is building a model from scratch or fine-tuning a foundational model, the quality and accessibility of its proprietary data are the primary differentiators.
Retrieval Augmented Generation (RAG)
To move beyond generic AI responses and reduce hallucinations, many organizations are implementing Retrieval Augmented Generation (RAG). RAG allows foundational models to access evolving, large-scale data lakes without requiring the data to be manually loaded into a prompt.
The RAG methodology involves:
1. Vectorization: Converting organizational data into vectors—numeric representations that capture semantic meaning.
2. Semantic Search: Using spatial similarity to compare a query vector against the data lake’s vectors to find the most relevant information.
3. Augmentation: Feeding the retrieved context back into the model to generate a more accurate and business-specific response.
Ingestion and Data Strategy with Amazon S3
Amazon S3 serves as the foundational data lake for these AI workflows due to its cost-effectiveness and virtually unlimited scalability. Organizations typically utilize two ingestion patterns:
* Batch Ingestion: Suitable for static or infrequently changing data such as historical records and product catalogs.
* Real-Time Ingestion: Essential for agentic workflows where AI models must respond to the latest available information.
Modernizing Self-Managed Databases with Amazon FSx
While fully managed services like Amazon RDS are popular, certain business and technical requirements drive organizations toward self-managed database architectures on AWS.
Drivers for Self-Managed Databases
Organizations choose to self-manage databases like Oracle, SQL Server, or SAP HANA for several reasons:
* Granular Control: The ability to choose specific versions of the database engine and the underlying operating system.
* Custom Protection Policies: Implementing specific backup intervals and recovery procedures that may not be available in managed services.
* High Resilience: Scaling databases across multiple Availability Zones or regions with custom failover configurations.
Optimization through Storage Features
A common oversight in database deployment is the potential for the storage layer to add significant value beyond simple data persistence. Amazon FSx file systems (including FSx for NetApp ONTAP, OpenZFS, and Windows File Server) enable features like:
* Snapshots and Cloning: Facilitating rapid testing and database upgrades by creating near-instantaneous copies of production environments.
* Performance Tuning: Choosing the right FSx service can significantly optimize the TCO and performance of database environments, particularly for high-transaction workloads.
Conclusion
As compute power continues to expand, the storage layer must evolve from a passive repository into a high-performance engine. By leveraging Amazon FSx and S3, organizations can eliminate storage bottlenecks, enabling their most demanding AI, HPC, and database workloads to scale linearly and cost-effectively in the cloud.
Links:
- Amazon FSx Product Page
- Amazon S3 Product Page
- AWS re:Invent 2025 – High-performance storage for AI/ML, analytics, and HPC workloads (STG336)
- AWS re:Invent 2025 – Accelerate gen AI and ML workloads with AWS storage (STG201)
- AWS re:Invent 2025 – Improve self-managed database performance and agility with Amazon FSx (STG337)
[AWSReInvent2025] Advancements in AWS Infrastructure as Code: A Comprehensive Year-in-Review of CloudFormation and CDK Innovations
Lecturer
The session is delivered by product managers from Amazon Web Services who oversee the development and roadmap of AWS CloudFormation and the AWS Cloud Development Kit.
Abstract
This article provides an exhaustive and detailed retrospective on the notable progress achieved throughout the past year in AWS infrastructure as code services, with particular emphasis on both AWS CloudFormation and the AWS Cloud Development Kit (CDK). It meticulously examines a range of enhancements, including improved validation mechanisms, clearer error diagnostics, expanded construct libraries, seamless integration with artificial intelligence assistance through Model Context Protocol servers, and advanced troubleshooting utilities. The discussion analyzes how these collective innovations substantially elevate deployment reliability, enhance developer productivity, and introduce greater intelligence into infrastructure management practices for organizations of all scales.
The Critical and Enduring Role of Infrastructure as Code in Modern Cloud Architectures
Infrastructure as code has firmly established itself as an indispensable discipline for enterprises striving to achieve consistency, traceability, and accelerated iteration in their cloud operations. AWS CloudFormation offers a robust declarative approach, allowing practitioners to define resources through structured templates in JSON or YAML formats, thereby guaranteeing identical provisioning outcomes across development, staging, and production environments.
Complementing this, the AWS Cloud Development Kit empowers developers with programmatic flexibility, enabling infrastructure definition in familiar programming languages while automatically generating underlying CloudFormation templates. This duality accommodates diverse team preferences and skill sets.
The advancements introduced over the year have strategically bridged these paradigms, delivering unified capabilities that address contemporary challenges related to scale, complexity, and the evolving demands of developer experience in dynamic cloud ecosystems.
Significant Refinements Enhancing AWS CloudFormation Reliability and Practitioner Usability
AWS CloudFormation has benefited from meaningful improvements in change set validation processes, enhanced clarity in error messaging, and more intuitive management of deployment workflows. These refinements work collectively to substantially reduce the frequency of failed deployments by surfacing potential conflicts, resource constraints, or configuration incompatibilities earlier in the provisioning lifecycle.
Furthermore, the introduction of server-side APIs now enables programmatic pre-validation of proposed changes, allowing integration into continuous integration pipelines for automated safeguards that prevent runtime disruptions and promote greater confidence in infrastructure updates.
Substantial Growth and Maturation Within the AWS Cloud Development Kit Ecosystem
The AWS Cloud Development Kit has experienced considerable expansion in supported programming languages and the availability of high-level constructs. Numerous libraries, both community-contributed and AWS-maintained, have progressed from experimental developer preview stages to full general availability, covering an extensive array of common architectural patterns across networking, security, serverless computing, and data processing domains.
This maturation process provides developers with higher-level abstractions that encapsulate established best practices, thereby significantly reducing the amount of boilerplate code required and promoting greater architectural consistency across distributed teams.
Transformative Integration of Artificial Intelligence Assistance Through Model Context Protocol Servers
One of the most pivotal innovations involves the creation of specialized Model Context Protocol servers tailored specifically for CDK and CloudFormation contexts. These servers curate and expose AWS-specific expertise—including recommended practices, construct libraries at various maturity levels, and detailed cloud context information—directly to artificial intelligence-powered coding assistants.
As a result, developers receive highly contextually relevant suggestions that align precisely with AWS service conventions and idioms, dramatically accelerating the creation of secure, efficient, and idiomatic implementations while substantially lowering the cognitive burden associated with recalling intricate service details.
Strengthening Troubleshooting and Validation Tooling for Proactive Issue Resolution
New diagnostic capabilities encompass server-side APIs designed for interrogating deployment states and identifying root causes of issues, complemented by local static analysis utilities that perform early detection of syntax errors within CDK source code.
These tools operate across both programmatic CDK definitions and the generated CloudFormation templates, enabling practitioners to identify and resolve configuration problems well before they manifest during actual deployments.
Community-Driven Construct Libraries and Enhanced Cloud Context Integration
The ecosystem continues to benefit from active contributions spanning AWS internal teams and external community participants, with constructs systematically progressing through alpha evaluation and eventual general availability phases.
Additional cloud context features further enrich artificial intelligence interactions by providing service-specific insights and recommendations.
Practitioners are strongly encouraged to explore dedicated workshops that offer guided paths for understanding and implementing MCP server integration in real-world scenarios.
Measurable Organizational Benefits and Strategic Adoption Considerations
These multifaceted improvements collectively lower entry barriers for effective infrastructure management while delivering tangible advantages. Development teams realize enhanced confidence in deployment outcomes, accelerated onboarding for new members, and improved adherence to evolving architectural standards across projects.
The incorporation of artificial intelligence guidance represents a fundamental paradigm shift toward more intelligent, assisted development experiences that amplify human expertise rather than seeking to replace it.
Looking Toward the Future of Intelligent Infrastructure Orchestration
Continued investment in these areas clearly signals an ongoing commitment to deepening the convergence between programmatic expressiveness and declarative safety, increasingly augmented by artificial intelligence capabilities that guide practitioners toward optimal architectural outcomes.
Organizations that fully leverage these evolving tools position themselves advantageously for sustained operational excellence amid the accelerating complexity of modern cloud environments.
Links:
[AWSReInvent2025] Basketball’s AI Revolution: How AWS and the NBA Are Changing the Game
Lecturer
Chris Benyarko is Executive Vice President of Direct-to-Consumer at the NBA, overseeing fan engagement and digital strategies. Andy Oh serves as Principal of Live Sports Events at Prime Video, leading NBA broadcasting partnerships. Kristen Schaff is Global Director of Sports Partnerships at AWS, managing collaborations across major leagues. Relevant links include Chris Benyarko’s LinkedIn profile (https://www.linkedin.com/in/chris-benyarko-/) and Kristen Schaff’s LinkedIn profile (https://www.linkedin.com/in/kristen-schaff/).
Abstract
This article investigates the NBA’s digital transformation via AWS, focusing on AI-driven analytics, fan personalization, and broadcasting innovations. It analyzes partnerships enhancing game strategies, viewer experiences, and global engagement, with implications for sports technology scalability.
The NBA-AWS Partnership: Shared Vision and Technological Foundations
The NBA’s strategic alliance with AWS, formally unveiled on October 1st, is rooted in a mutual commitment to innovation and an unwavering focus on fan experiences. Chris Benyarko emphasizes that this partnership transcends mere technology provision, positioning AWS as a true collaborator in advancing the league’s goals. At its foundation lies a shared philosophy: while the NBA prioritizes fan and future fan obsession, AWS brings its renowned customer-centric approach, creating a synergy that amplifies their joint efforts. This alignment enables the league to harness AWS’s robust infrastructure for seamless integration across various operations, ultimately accelerating the pace of technological advancements.
In the broader context of basketball’s ongoing evolution, the need for sophisticated, data-driven solutions has never been more pressing. AWS offers a scalable cloud platform that excels in handling complex analytics, artificial intelligence, and machine learning tasks, converting vast amounts of raw data into meaningful insights that inform decision-making at every level. Kristen Schaff highlights what drew AWS to the NBA, pointing out the league’s dynamic, fast-paced nature and its abundance of data as ideal attributes that align perfectly with AWS’s technological strengths. From player performance tracking to predictive modeling, this collaboration leverages AWS’s tools to address the unique demands of professional sports.
The methodology underpinning this partnership involves a comprehensive migration of workflows to AWS services, ensuring low-latency streaming and personalized content delivery that reaches audiences worldwide. By combining the NBA’s deep domain knowledge with AWS’s technical prowess, the alliance not only enhances current offerings but also paves the way for future innovations that could redefine the sport.
AI and Analytics Transforming Gameplay and Strategy
Artificial intelligence is at the forefront of reshaping basketball analytics, influencing everything from individual player development to collective team strategies during games. Chris Benyarko delves into the capabilities of Second Spectrum’s optical tracking system, which deploys 29 cameras in each arena to capture an astonishing 100 million data points per night. These metrics encompass detailed aspects such as player speed, defensive positioning, and shot quality, providing coaches and analysts with granular information that was previously unattainable.
AWS plays a pivotal role in this transformation by powering machine learning models that forecast game outcomes and simulate various scenarios, thereby assisting coaches in refining their tactics. The implications are significant, as teams can now gain substantial competitive advantages through data-informed decisions, while fans benefit from enriched content on platforms like NBA League Pass, including automated highlight reels that capture the most thrilling moments. Andy Oh complements this by describing how Prime Video integrates AWS for real-time statistical overlays, which add layers of depth to the viewing experience and foster greater immersion.
Nevertheless, challenges such as data latency persist, and the partnership addresses these through continuous infrastructure optimizations, ensuring that the flow of information remains timely and reliable.
Enhancing Fan Engagement Through Personalization
Personalization has emerged as a key driver in elevating fan engagement, utilizing AI to deliver content that resonates on an individual level. Chris Benyarko explains the progression of NBA League Pass, which now employs AI to generate highlights in multiple languages, offer alternate viewing streams focused on specific players, and provide predictive elements like real-time win probabilities. These features not only cater to diverse global audiences but also deepen the connection between fans and the game.
AWS’s extensive global network facilitates this by guaranteeing low-latency delivery to over 200 countries, making high-quality experiences accessible regardless of location. Kristen Schaff underscores the importance of data privacy within these personalization efforts, ensuring that the NBA’s fan-first principles are upheld through secure, unified data management practices.
An analysis of this approach reveals its potential to shift traditional passive spectatorship toward more interactive and tailored interactions, which in turn boosts viewer retention and opens new avenues for monetization through precisely targeted advertising.
Broadcasting Innovations and Latency Challenges
Prime Video’s integration of NBA content exemplifies how AWS enables groundbreaking broadcasting innovations. Andy Oh outlines the process of capturing feeds directly from arenas and minimizing transmission hops to achieve near-real-time delivery, a critical factor especially for integrations involving live betting.
Among the notable advancements is AI-generated commentary available in various languages, powered by AWS Bedrock for natural and accurate translations. The broader implications extend to democratizing access to premium content, thereby expanding the NBA’s global footprint and attracting new demographics. However, the persistent challenge of avoiding spoilers drives an ongoing emphasis on latency reduction, with AWS tools providing the means for continuous monitoring and swift adjustments to maintain optimal performance.
Implications for Sports and Broader Industries
The NBA-AWS partnership offers valuable insights that transcend the realm of sports, demonstrating the power of real-time data platforms, personalized content delivery, and AI in production environments. Chris Benyarko envisions extending these technologies to non-professional leagues, potentially increasing participation by making advanced analytics more widely available.
Looking ahead, AI could further innovate by predicting injuries or optimizing training regimens, fundamentally altering athletic preparation and performance. These developments not only enhance the sport but also provide scalable models applicable to other industries seeking to leverage data for competitive advantage.
Conclusion
The synergy between AWS and the NBA vividly illustrates the transformative potential of AI in sports. By enhancing analytics, personalization, and broadcasting through advanced cloud technologies, this collaboration redefines fan engagement and sets a precedent for innovation across various sectors.
Links:
- https://www.youtube.com/watch?v=pZczwGVzWxo
- https://www.linkedin.com/in/chris-benyarko-/
- https://www.linkedin.com/in/kristen-schaff/
[AWSReInventPartnerSessions2024] Usage
spec = “Sort a list of numbers”
code = generate_code(spec)
tests = [([3, 1, 2], [1, 2, 3]), ([5, 4], [4, 5])]
if test_code(code, tests):
print(“Code passes tests”)
“`
This exemplifies the iterative process of generation and validation central to the platform.
Analytical Implications for Efficiency and Innovation
The deployment of GenWizard reveals profound implications for operational efficiency. By automating repetitive tasks, it allows teams to focus on high-value activities, reducing project timelines by up to seventy percent in some cases. This efficiency stems from the platform’s ability to handle complex correlations and predictions, as seen in incident management where noise reduction leads to faster resolutions.
Innovation is fostered through enhanced decision-making. The system’s knowledge base, enriched with historical data and AI insights, supports proactive strategies like predictive maintenance and application rationalization. For instance, analyzing application portfolios identifies redundancies, enabling cost savings and streamlined operations.
Collaboration with technology partners like AWS amplifies these benefits. Amazon Q’s integration ensures seamless natural language interactions, democratizing access to advanced tools and promoting a culture of continuous improvement.
Consequences for Enterprise Adoption and Future Directions
Enterprise adoption of such platforms mitigates risks associated with legacy systems, facilitating smoother migrations and modernizations. However, challenges include ensuring data privacy and model accuracy, addressed through robust governance frameworks.
Future directions involve expanding agentic capabilities to encompass more lifecycle stages, potentially incorporating multimodal AI for broader applications. This could revolutionize industries by enabling autonomous operations, where systems self-optimize based on real-time data.
In conclusion, the fusion of generative AI with service delivery platforms like GenWizard, powered by AWS, represents a paradigm shift toward intelligent, efficient technology management, promising sustained competitive advantages.
Links:
[KotlinConf2025] Blueprints for Scale: What AWS Learned Building a Massive Multiplatform Project (2nd version)
In a talk on building large-scale, multiplatform projects, Ian Botsford and Matis Lazdins of Amazon Web Services (AWS) shared their experiences creating the AWS SDK for Kotlin. This project is colossal, spanning over 300 services and targeting eight different platforms, with its code distributed across four repositories and nearly 500 Gradle modules. The talk provided a blueprint for managing a codebase with over 8.6 million lines of code, 98% of which is auto-generated. The key to their success, they claimed, was a set of five core principles that kept their maintainers sane and productive.
A Principled Approach to Development
Botsford and Lazdins detailed five tenets for managing a project of this scale: owning your dependencies, structuring the project for growth, designing for Kotlin Multiplatform (KMP) from the beginning, maintaining backward compatibility, and optimizing the maintainer experience. They provided a practical example of owning dependencies by discussing their choice of HTTP clients. Instead of exposing third-party library types directly, which could lead to inconsistent configurations and vulnerability to unexpected API changes, they created a common, abstract interface to maintain consistency and shield users from underlying implementation details.
Automating for Maintainer Sanity
A significant part of their strategy focused on the maintainer experience. Lazdins explained the importance of automating repetitive and mundane tasks to free up time for more complex work. They developed broad checks to catch issues before they are merged, which helps prevent regressions and enforce project standards. The speakers stressed that these checks should be highly informative but also overridable, giving developers autonomy while providing valuable feedback. This focus on a positive maintainer experience is crucial for the health of any large open-source project and is a key factor in the daily releases that happen sometimes multiple times a day.
Links:
[AWSReInvent2025] Introducing Nitro Isolation Engine: Transparency through Mathematics
Lecturer
JD Bean is a principal architect in AWS’s compute and ML services organization, specializing in virtualization and security innovations. Kareem Raslan serves as a senior principal engineer in AWS’s Nitro hypervisor team, focusing on hardware-software integration for cloud security. Nathan Chong is a principal applied scientist in AWS’s automated reasoning group, with expertise in formal verification and mathematical proofs. Relevant links include JD Bean’s LinkedIn profile (https://www.linkedin.com/in/jdbean/) and Nathan Chong’s LinkedIn profile (https://www.linkedin.com/in/nathan-chong-aws/).
Abstract
This article explores the AWS Nitro Isolation Engine, an advancement in the Nitro System that employs formal verification to ensure mathematical certainty in workload isolation. It examines the evolution of Nitro’s design, the application of automated reasoning for proofs, and the implications for cloud security, emphasizing compartmentalization and transparency.
The Evolution of the AWS Nitro System
The AWS Nitro System has fundamentally transformed the landscape of cloud virtualization by prioritizing enhanced security, superior performance, and accelerated innovation. JD Bean traces its development back to 2012, explaining how it culminated in a public launch in 2017 that marked a departure from conventional hypervisors such as Xen. At its core, the system relies on a customized version of the KVM hypervisor tailored specifically for cloud environments, complemented by the sixth generation of proprietary Nitro Silicon. This infrastructure underpins all EC2 instances introduced since 2018, demonstrating AWS’s commitment to reimagining virtualization.
In earlier iterations, systems like Xen depended on a component known as Dom0, which essentially functioned as a general-purpose operating system to handle essential tasks such as input/output operations, orchestration, and monitoring. However, as AWS expanded its services and built deeper relationships with customers, the limitations of Xen became increasingly apparent. The team recognized the need to push beyond these constraints, leading to a comprehensive reinvention that eliminated superfluous elements and relocated AWS-specific functions to dedicated hardware. Consequently, the Nitro System features a streamlined host operating system reduced to a minimal kernel, which not only minimizes potential attack surfaces but also enforces a policy of zero operator access, thereby isolating customer data from AWS personnel.
Within this broader context, the rise of cloud adoption has amplified the demand for confidential computing, where sensitive workloads require robust protections against unauthorized access. The Nitro architecture addresses these needs by compartmentalizing only the most critical isolation functions, which in turn optimizes efficiency and reduces vulnerabilities. This design philosophy ensures that customers can leverage the cloud’s scalability without compromising on security, setting the stage for subsequent advancements like the Nitro Isolation Engine.
Design and Implementation of the Nitro Isolation Engine
Building upon the foundational principles of the Nitro System, the Nitro Isolation Engine introduces a compact and formally verified module that significantly bolsters isolation assurances. Kareem Raslan elaborates on its compartmentalization strategy, noting how non-essential operations are shifted to user space, leaving behind a concise kernel comprising fewer than 100,000 lines of code dedicated solely to vital activities such as memory allocation and interrupt handling.
This engine is currently implemented on the Graviton 5 processor, available in preview mode, and utilizes specialized hardware extensions to facilitate secure transitions across compartments. The implementation methodology centers on rigorous specification, where the engine’s expected behaviors—such as maintaining strict workload separation—are articulated through precise mathematical models. Subsequently, the team employs tools like Isabelle to prove that the actual code aligns perfectly with these specifications, thereby guaranteeing that no deviations occur.
Nathan Chong further illuminates the process of automated reasoning, beginning with intuitive examples like the formula for the sum of the first n natural numbers and progressing to sophisticated machine-checked proofs. For the engine, this approach extends to verifying properties over potentially infinite states, which ensures that unauthorized access paths are entirely eliminated. The result is a system that not only performs efficiently but also withstands rigorous scrutiny, providing customers with unparalleled confidence in their data’s protection.
The implications of this design are profound, as it substantially diminishes the risk of exploitation by confining the trusted computing base to a minimal footprint. By verifying a smaller codebase through automated means, the engine mitigates issues stemming from legacy components, paving the way for a more secure cloud ecosystem.
Automated Reasoning and Mathematical Proofs
Automated reasoning stands as a cornerstone of the Nitro Isolation Engine, offering what the presenters describe as “transparency through mathematics” by delivering incontrovertible assurances of isolation. Nathan Chong contrasts informal proofs and specifications with their machine-checked counterparts in the Isabelle theorem prover, where each logical step is mechanically validated to prevent errors.
At the heart of this process lie core concepts such as specifications, which define the precise behaviors a system must exhibit, and proofs, which consist of finite chains of reasoning that irrefutably establish desired properties. For domains involving infinite possibilities, such as the natural numbers, techniques like mathematical induction are employed: a base case confirms the property for the initial value, while the inductive step demonstrates its preservation across subsequent values, much like a cascade of falling dominoes.
Scaling these methods to the complexities of the Nitro Isolation Engine requires advanced mathematical frameworks, including separation logic for managing memory resources, refinement techniques for bridging abstraction levels, and theorem provers to automate verification. Drawing on decades of research in formal methods, this approach ensures comprehensive coverage of real-world scenarios, including concurrent operations that could otherwise introduce subtle vulnerabilities.
An analysis of this methodology reveals its inherent value: unlike traditional testing, which is confined to finite scenarios, mathematical proofs provide exhaustive guarantees, fostering a level of trust that is essential for confidential computing environments. This not only elevates security standards but also enables organizations to innovate with greater assurance.
Implications for Cloud Security and Future Innovations
The introduction of the Nitro Isolation Engine heralds a new era in cloud security, where mathematical proofs become the benchmark for verifying system integrity. By emphasizing compartmentalization, the engine effectively minimizes the trusted computing base, thereby reducing the potential for exploits and enhancing overall resilience. Currently available as an always-on feature on Graviton 5 processors in preview, it invites users to request access through designated AWS channels, signaling AWS’s proactive stance in deploying cutting-edge security measures.
On a broader scale, the consequences extend to industries with stringent privacy requirements, such as finance and healthcare, where verifiable isolation can mitigate compliance risks and build customer confidence. AWS’s ongoing commitment to elevating security standards—evident throughout the Nitro System’s history—suggests that future innovations will continue to prioritize robust protections, allowing for rapid advancements without sacrificing safety.
This transparency through mathematics not only demystifies complex systems but also empowers users to make informed decisions about their cloud strategies, ultimately contributing to a more secure digital landscape.
Conclusion
The Nitro Isolation Engine exemplifies AWS’s unwavering dedication to pioneering secure and innovative cloud infrastructure. Through the rigorous application of formal verification, it achieves mathematical certainty in workload isolation, thereby redefining transparency and trust in the realm of virtualization.
Links:
- https://www.youtube.com/watch?v=hqqKi3E-oG8
- https://www.linkedin.com/in/jdbean/
- https://www.linkedin.com/in/nathan-chong-aws/
[AWSReInvent2025] Revolutionizing DevSecOps: How Cathay Pacific Achieved 75% Faster Security with Agentic AI
Lecturer
Mike Markell is a Practice Manager for AWS Professional Services in Hong Kong, where he leads digital transformation and security initiatives for major enterprises across Asia. Naresh Sharma is a senior technology leader at Cathay Pacific Airways, overseeing the airline’s global application security and DevSecOps strategy. Tony Leong is a Senior Security Architect at Cathay, specialized in building AI-powered security tooling and integrating AppSec-as-Code into high-velocity deployment pipelines.
Abstract
In the highly regulated and high-stakes environment of global aviation, managing security across more than 4,000 annual deployments presents a massive operational challenge. This article details how Cathay Pacific Airways revolutionized its “security-first” culture by moving beyond traditional security scanning to a comprehensive DevSecOps model. The core methodology centers on the implementation of Agentic AI and a RAG-based (Retrieval-Augmented Generation) assistant to solve the industry’s “false positive crisis.” By deploying “AI-powered security champions” and customized scanning rules, Cathay achieved a 75% reduction in vulnerability remediation time and a 50% reduction in security operations costs. The analysis explores the technical and cultural shifts required to empower over 1,000 developers to become proactive security practitioners while maintaining the airline’s rapid pace of innovation.
Context: The Bottleneck of Manual Security Reviews
For a global leader like Cathay Pacific, the pace of digital innovation is essential for maintaining a competitive edge in the aviation industry. However, this speed was being severely hindered by the limitations of traditional security scanning tools. The primary conflict centered on a high noise-to-signal ratio, where approximately 78% of the vulnerabilities identified by standard tools were determined to be false positives. This created a crisis where security teams were overwhelmed by alerts, leading to significant delays in the deployment of features for the airline’s fleet.
Furthermore, the manual review process required to validate these alerts created significant friction between the security and development teams. Developers often viewed security requirements as a hurdle that slowed down their ability to deliver value, while security professionals struggled to keep up with the volume of code being produced. To overcome these challenges, Cathay needed a solution that could scale with their deployment frequency—which covers everything from customer-facing apps to critical flight operation systems—without compromising on the rigorous safety standards that define the brand.
Methodology: Implementing Shift-Left Security with AI
The solution implemented by Cathay Pacific and AWS Professional Services involved a comprehensive “shift-left” strategy, which integrates security at the very beginning of the software development lifecycle. The cornerstone of this methodology is the use of Agentic AI. Unlike traditional static scanners, these AI agents act as “security champions” that provide real-time, context-aware guidance to developers as they write code. This allows for the identification of security anti-patterns and the suggestion of defensive coding practices before the code is even committed to a repository.
Another critical component of the methodology is the AppSec-as-Code library. This centralized knowledge base translates complex security policies into programmatic requirements that can be automatically enforced within CI/CD pipelines. To make this information accessible to developers, the team developed a RAG-based (Retrieval-Augmented Generation) assistant. This tool allows developers to query internal security standards using natural language, receiving accurate and context-specific advice instantly. Finally, the team moved away from “out of the box” tool configurations in favor of highly customized scanning rules. This technical fine-tuning was essential for drastically reducing the false-positive rate and ensuring that the security team only focused on legitimate threats.
Technical Analysis of Operational Gains
The implementation of AI-driven DevSecOps has yielded remarkable quantitative results for Cathay Pacific. The most significant outcome is a 75% reduction in the time required to remediate vulnerabilities. Because the AI agents filter out the vast majority of false positives and provide developers with clear, actionable fix suggestions, the entire security lifecycle has been compressed. Qualitatively, this has led to a 70% improvement in developer security capability, as the tools effectively serve as an automated, on-the-job training system that reinforces secure coding habits.
From a financial perspective, the automation of manual reviews and the reduction in wasted engineering time have led to a 50% cost reduction in security operations. The airline is now able to manage over 4,000 deployments annually with a higher level of confidence and lower overhead than was previously possible. A critical technical lesson learned during the journey was that “by default, no tool is perfect.” Success required a commitment to continuous customization and a willingness to collaborate with product vendors to tune their tools to the specific needs of the aviation industry. This iterative feedback loop was the key to moving from “human-in-the-loop” automation to a more efficient “AI-informed” model.
Consequences: A Cultural and Technical Transformation
The transformation at Cathay Pacific extended far beyond the technical architecture; it required a fundamental shift in the organization’s culture. The success of the project was predicated on a “can-do” spirit and the setting of ambitious targets that challenged the status quo. By providing developers with the tools to take ownership of security, the organization has fostered a culture where security is seen as a shared responsibility rather than an external constraint.
The implications for the global aviation and enterprise sectors are significant. Cathay has proven that it is possible to maintain a high-velocity deployment schedule in a safety-critical environment by leveraging the power of generative AI. Looking forward, the organization plans to develop even more insightful dashboards to provide security leaders with real-time visibility into the health of the application portfolio. The journey serves as a powerful testament to how Agentic AI can bridge the gap between agility and security, turning a potential bottleneck into a powerful competitive advantage.
Links:
[AWSReInvent2025] Amazon S3 Performance: Architecture, Design, and Optimization for Data-Intensive Systems
Lecturer
Ian Heritage is a Senior Solutions Architect at Amazon Web Services, specializing in Amazon S3 and large-scale data storage architectures. With deep expertise in performance engineering and distributed systems, Ian Heritage helps organizations design and optimize their storage layers for high-throughput and low-latency applications, including machine learning training and real-time analytics. He is a prominent figure in the AWS storage community, known for his technical deep-dives into S3’s internal mechanics and best practices for performance at scale.
Abstract
This article explores the internal architecture and performance optimization strategies of Amazon S3, the industry-leading object storage service. It provides a detailed analysis of the differences between S3 General Purpose and the newly introduced S3 Express One Zone storage class, highlighting the architectural trade-offs between regional durability and sub-millisecond latency. The discussion covers advanced request management techniques, including prefix partitioning, request routing, and the role of the AWS Common Runtime (CRT) in maximizing throughput. By examining these technical foundations, the article offers practical guidance for architecting storage solutions that can handle millions of requests per second and petabytes of data for modern AI and analytics workloads.
S3 Storage Class Selection for High Performance
The performance of an S3-based application is fundamentally determined by the selection of the storage class. For over a decade, S3 General Purpose (Standard) has been the default choice, offering 99.999999999% (11 9s) of durability by replicating data across at least three Availability Zones. While this provides extreme reliability, the regional replication introduces a baseline latency that may be too high for certain “request-intensive” applications, such as machine learning model checkpoints or high-frequency trading logs.
To address these needs, AWS introduced S3 Express One Zone. This storage class is designed for workloads that require consistent, single-digit millisecond latency. By storing data within a single Availability Zone and utilizing a new, purpose-built architecture, Express One Zone can deliver up to 10x the performance of S3 Standard at a 50% lower request cost. This class is ideal for applications that perform frequent, small I/O operations where the overhead of regional replication would be the primary bottleneck. The choice between Standard and Express One Zone is thus a strategic decision between geographic durability and extreme performance.
Request Routing, Partitioning, and the Scale-Out Architecture
At its core, Amazon S3 is a massively distributed system that scales out to handle virtually unlimited throughput. The key to this scaling is “partitioning.” S3 automatically partitions buckets based on the object keys (names). Each partition can support a specific number of requests: 3,500 PUT/COPY/POST/DELETE requests and 5,500 GET/HEAD requests per second per prefix. For many years, users were advised to use randomized prefixes to ensure even distribution across partitions.
Modern S3 architecture has evolved to handle this automatically, but understanding prefix design remains crucial for performance. When an application’s request rate increases, S3 detects the hot spot and splits the partition to handle the load. However, this process takes time. For workloads that burst from zero to millions of requests instantly, pre-partitioning or using a wide range of prefixes is still a best practice. By spreading data across multiple prefixes (e.g., bucket/prefix1/, bucket/prefix2/), an application can linearly scale its throughput to accommodate massive concurrency, limited only by the client’s network bandwidth and CPU.
Client-Side Optimization with AWS CRT and SDKs
While the S3 service is designed for scale, the performance experienced by the end-user is often limited by the client-side implementation. To bridge this gap, AWS developed the Common Runtime (CRT) library. The CRT is a set of open-source, C-based libraries that implement high-performance networking best practices, such as automatic request retries, congestion control, and most importantly, multipart transfers.
'''
Conceptual example of enabling CRT in the AWS SDK for Python (Boto3)
'''
import boto3
from s3transfer.manager import TransferConfig
'''
The CRT allows for automatic parallelization of large object transfers
'''
config = TransferConfig(use_threads=True, max_concurrency=10)
s3 = boto3.client('s3')
s3.upload_file('large_data.zip', 'my-bucket', 'data.zip', Config=config)
The CRT automatically breaks large objects into smaller parts and uploads or downloads them in parallel. This utilizes the full network capacity of the EC2 instance and mitigates the impact of single-path network congestion. For applications using the AWS CLI or SDKs for Java, Python, and C++, opting into the CRT-based clients can result in a significant throughput increase—often double or triple the speed of standard clients for large files. Additionally, the CRT handles the complexities of DNS load balancing and connection pooling, ensuring that requests are distributed efficiently across the S3 frontend fleet.
Case Study: Optimization for Machine Learning and Analytics
Machine learning training is a premier use case for S3 performance optimization. During the training of large language models (LLMs), hundreds or thousands of GPUs must simultaneously read training data and write model “checkpoints.” These checkpoints are multi-gigabyte files that must be saved quickly to avoid idling expensive compute resources. By combining S3 Express One Zone with the CRT-based client, researchers can achieve the throughput necessary to saturate the high-speed networking of P4 and P5 instances.
In analytics, the use of “Range Gets” is a critical optimization. Instead of downloading an entire 1GB Parquet file to read a few columns, an application can request specific byte ranges. This reduces the amount of data transferred and speeds up query execution. S3 is optimized to handle these range requests efficiently, and when combined with a partitioned data layout (e.g., partitioning by date or region), it enables sub-second query responses over petabytes of data. This architectural synergy between storage class, partitioning, and client-side logic is what allows S3 to serve as the foundation for the world’s largest data lakes.
Links:
[NDCOslo2024] Choosing The Best AWS Service For Your Website + API – Brandon Minnick
In the sprawling spectrum of cloud solutions, where a plethora of platforms perplex even the seasoned, Brandon Minnick, an AWS architect and mobile maestro, navigates the nebulous nebula of Amazon’s offerings. As a developer advocate with a penchant for demystifying deployment, Brandon dissects the dizzying array of AWS services—Lambda, Elastic Beanstalk, Lightsail, Amplify, S3—distilling their distinct domains to guide builders toward bespoke backends. His exploration, enriched with empirical evaluations, empowers enterprises to align ambition with architecture, balancing cost, celerity, and scalability.
Brandon begins with a confession: his own odyssey, as a mobile maestro thrust into AWS’s vast vault, was overwhelmed by options—acronyms and aliases abounding. His mission: map the maze, matching motives to mechanisms, ensuring websites and APIs ascend with alacrity.
Decoding the Domain: AWS’s Hosting Horizons
AWS’s arsenal is abundant: S3 stores static simplicities, buckets brimming with bits; Amplify augments apps, knitting frontends to functions. Brandon breaks down the basics: Elastic Beanstalk builds bridges, automating infrastructure; Lightsail lightens loads, offering preconfigured planes; Lambda launches lean, serverless scripts scaling seamlessly.
Each excels in its enclave: S3’s simplicity suits static sites, Amplify’s agility aids authenticated apps, Lambda’s litheness loves lightweight logic. Brandon’s benchmark: cost—S3’s cents versus Lambda’s low levies; speed—CloudFront’s celerity; scale—Fargate’s fluidity.
Cost and Celerity: Calculating the Calculus
Price predicates priority: S3’s storage starts at sub-dollar sums, Lambda’s invocations linger at $0.20 per million, Amplify’s adaptability aligns at $0.023 per GB. Brandon’s breakdown: static sites savor S3’s thrift, dynamic domains demand Amplify’s depth—authentication via Cognito, APIs via API Gateway.
Performance pulses: CloudFront’s CDN cuts latency to 300ms, Lambda’s cold starts cede to containers’ constancy. Brandon advises: weigh user whims—300ms matters for markets, less for leisurely loads.
Scalability and Simplicity: Structuring for Surge
Scalability shapes success: Lambda’s limitless leaps, Fargate’s fleet-footed fleets, Beanstalk’s balanced ballast. Brandon illustrates: API Gateway guards gates, throttling torrents; Amplify’s auto-scaling absolves administrative aches.
Simplicity seals the deal: Lightsail’s one-click launches lure lone developers; Amplify’s abstractions attract architects. Brandon’s beacon: start small—S3 for static, scale to Amplify for ambition.
Strategic Selection: Synthesizing Solutions
Brandon’s synthesis: match mission to mechanism—S3 for static starters, Amplify for authenticated ascents, Lambda for lean logic. His counsel: consult AWS’s compendium—getting-started guides, web app wisdom—curated for clarity.
His clarion: choose consciously, calibrating cost, celerity, scalability—AWS’s arsenal awaits.
Links:
[KotlinConf2025] Blueprints for Scale: What AWS Learned Building a Massive Multiplatform Project
Ian Botsford and Matis Lazdins from Amazon Web Services (AWS) shared their experiences and insights from developing the AWS SDK for Kotlin, a truly massive multiplatform project. This session provided a practical blueprint for managing the complexities of a large-scale Kotlin Multiplatform (KMP) project, offering firsthand lessons on design, development, and scaling. The speakers detailed the strategies they adopted to maintain sanity while dealing with a codebase that spans over 300 services and targets eight distinct platforms.
Architectural and Development Strategies
Botsford and Lazdins began by breaking down the project’s immense scale, explaining that it is distributed across four different repositories and consists of nearly 500 Gradle projects. They emphasized the importance of a well-defined project structure and the strategic use of Gradle to manage dependencies and build processes. A key lesson they shared was the necessity of designing for Kotlin Multiplatform from the very beginning, rather than attempting to retrofit it later. They also highlighted the critical role of maintaining backward compatibility, a practice that is essential for a project with such a large user base. The speakers explained the various design trade-offs they had to make and how these decisions ultimately shaped the project’s architecture and long-term sustainability.
The Maintainer Experience
The discussion moved beyond technical architecture to focus on the human element of maintaining such a vast project. Lazdins spoke about the importance of automating repetitive and mundane processes to free up maintainers’ time for more complex tasks. He detailed the implementation of broad checks to catch issues before they are merged, a proactive approach that prevents regressions and ensures code quality. These checks are designed to be highly informative while remaining overridable, giving developers the autonomy to make informed decisions. The presenters stressed that a positive maintainer experience is crucial for the health of any large open-source project, as it encourages contributions and fosters a collaborative environment.
Lessons for the Community
In their concluding remarks, Botsford and Lazdins offered a summary of the most valuable lessons they learned. They reiterated the importance of owning your own dependencies, structuring projects for scale, and designing for KMP from the outset. By sharing their experiences with a real-world, large-scale project, they provided the Kotlin community with actionable insights that can be applied to projects of any size. The session served as a powerful testament to the capabilities of Kotlin Multiplatform and the importance of a thoughtful, strategic approach to software development at scale.