Amazon Aurora serverless now delivers higher initial capacity during scale-up events, reaching up to 12 ACUs within a second and continuing to scale up to 256 ACUs as your workload grows. When the workload finishes, Aurora serverless automatically scales down to zero. This makes it especially well-suited for agentic AI applications, which typically have bursts of activity, long idle windows, and unpredictable traffic patterns. Aurora serverless handles all of it automatically, scaling capacity with your agents, so you only pay for what you use.
This enhancement is enabled by default on all Aurora serverless clusters running on platform version 3 or 4, with no configuration changes required. Existing clusters on platform versions 1 and 2 can upgrade directly to the latest platform version 4 to benefit from these improvements. You can verify your cluster’s platform version in the AWS Management Console under the instance configuration section, or via the RDS API’s ServerlessV2PlatformVersion parameter.
For pricing details and Region availability, visit Amazon Aurora Pricing. To learn more, read the Aurora serverless scaling documentation, and get started by creating an Aurora serverless database in just a few steps in the AWS Management Console.
Amazon Aurora serverless now delivers higher initial capacity during scale-up events, reaching up to 12 ACUs within a second and continuing to scale up to 256 ACUs as your workload grows. When the workload finishes, Aurora serverless automatically scales down to zero. This makes it especially well-suited for agentic AI applications, which typically have bursts of activity, long idle windows, and unpredictable traffic patterns. Aurora serverless handles all of it automatically, scaling capacity with your agents, so you only pay for what you use. This enhancement is enabled by default on all Aurora serverless clusters running on platform version 3 or 4, with no configuration changes required. Existing clusters on platform versions 1 and 2 can upgrade directly to the latest platform version 4 to benefit from these improvements. You can verify your cluster’s platform version in the AWS Management Console under the instance configuration section, or via the RDS API’s ServerlessV2PlatformVersion parameter. For pricing details and Region availability, visit Amazon Aurora Pricing. To learn more, read the Aurora serverless scaling documentation, and get started by creating an Aurora serverless database in just a few steps in the AWS Management Console.
Un nuevo enfoque de los datos de IA pone a las comunidades al mando
Por: Susanna Ray, redactora de Microsoft.
Gran parte de lo que la IA sabe sobre el mundo proviene de datos que la mayoría de la gente nunca ve y de procesos técnicos que determinan qué se incluye o se oye.
A medida que la IA se vuelve cada vez más influyente en la vida diaria, una nueva plataforma pretende ayudar a las comunidades a participar en esa base que se ha comenzado a construir entre bastidores para asegurar que se representen con precisión y justicia, en especial en imágenes. Llamado Community Library Creator, el esfuerzo de investigación de Microsoft ofrece a los grupos una manera estructurada de ayudar a moldear cómo aparecen en las imágenes generadas por IA.
«No existe una verdad fundamental sobre cómo deberían representarse las personas», dice Anja Thieme, investigadora principal en Microsoft con formación en psicología social e interacción humano-ordenador. «Debe definirse y negociarse colectivamente.»
Las comunidades construyen sus propias bibliotecas en lugar de depender de datos irregulares que existen en línea, a través de la nueva herramienta para ayudarles a definir cómo es una buena representación y construir los datos alrededor de ella de manera que los modelos y sistemas de IA puedan aprender para reflejar mejor su existencia.
¿Qué es una biblioteca comunitaria?
Una biblioteca comunitaria es una colección de imágenes o vídeos creados por un grupo de personas con experiencias compartidas — como una discapacidad, identidad o perspectiva vivida — que trabajan juntos a través de organizaciones de defensa para mostrar cómo quieren ser representados en la IA. Cada imagen se acompaña de descripciones que explican qué es lo que importa sobre ella, para añadir contexto que ayuda a los sistemas de IA a entender no solo cómo es algo, sino también lo que representa desde la perspectiva de la comunidad.
El software Community Library Creator, desarrollado por Thieme y su equipo junto con el equipo de Accesibilidad de Microsoft, proporciona las herramientas y la estructura necesarias para guiar ese proceso. Pero las decisiones sobre qué incluir, enfatizar o cambiar vienen de la propia comunidad. El resultado es un recurso que puede utilizarse para ayudar a los sistemas de IA a aprender de experiencias vividas más diversas, no solo de patrones en datos extraídos de internet.
¿Por qué necesitamos bibliotecas comunitarias?
La manera en que se retrata a las personas moldea cómo se les entiende en el mundo y qué oportunidades tienen a su alcance. Y a medida que las imágenes generadas por IA se vuelven más comunes, esas representaciones son cada vez más establecidas por estos sistemas.
Los modelos de IA aprenden patrones a partir de grandes colecciones de imágenes y texto, y pueden repetir cualquier laguna o distorsión que exista en esos datos. La gran cantidad de imágenes en internet de enanos místicos, por ejemplo, puede llevar a que los sistemas de IA generen que personas con enanismo tengan orejas puntiagudas u otras características fantásticas, dice Thieme. Y las personas con diferencias de extremidades aparecen en internet, de manera principal, en entornos médicos o deportivos, lo que significa que la IA no tiene el contexto necesario para mostrarlas en otras situaciones típicas, como trabajar o quedar para tomar un café con amigos.
En lugar de dejar la formación a los ingenieros o a lo que exista en internet, estas bibliotecas ofrecen a las comunidades una forma de moldear de manera activa tanto los datos como los estándares de evaluación que dirigen el desarrollo de la IA. Eso importa porque la representación no es solo técnica, dice Thieme, sino que debe definirse por las personas a las que afecta, basándose en experiencias vividas que otros no pueden replicar por completo.
Los miembros de grupos de defensa utilizan Community Library Creator para anotar imágenes y destacar detalles importantes para comprender sus experiencias y perspectivas de vida.Community Library Creator incluye un proceso de evaluación en el que los miembros de la comunidad revisan imágenes generadas por IA y valoran qué tan bien reflejan la representación deseada por su comunidad.En la herramienta Community Library Creator, los miembros de grupos de defensa revisan imágenes y explican por qué representan a su comunidad y qué quieren que otros entiendan sobre las personas y experiencias representadas.
Los miembros de grupos de defensa utilizan Community Library Creator para anotar imágenes y destacar detalles importantes para comprender sus experiencias y perspectivas de vida.
Community Library Creator incluye un proceso de evaluación en el que los miembros de la comunidad revisan imágenes generadas por IA y valoran qué tan bien reflejan la representación deseada por su comunidad.
En la herramienta Community Library Creator, los miembros de grupos de defensa revisan imágenes y explican por qué representan a su comunidad y qué quieren que otros entiendan sobre las personas y experiencias representadas.
¿Cómo funcionan en realidad las bibliotecas comunitarias?
Las bibliotecas comunitarias pueden construirse mediante el proceso estructurado que los investigadores de Microsoft crearon para la plataforma, basándose en principios de diseño inclusivo para ayudar a definir algo abstracto —la representación— más fácil.
Empieza de forma sencilla: los miembros de la comunidad eligen un pequeño conjunto de imágenes que resultan significativas y explican el por qué a través del método paso a paso del Community Library Creator. Esas reflexiones les ayudan a centrarse en temas clave —como la vida familiar, el trabajo o las rutinas cotidianas— que representan sus experiencias compartidas.
A partir de ahí, seleccionan una colección más amplia de imágenes o vídeos organizados en torno a esos temas, al añadir descripciones detalladas que explican qué es importante en cada escena. Una comunidad de personas negras con albinismo, por ejemplo, podría destacar elementos como llevar gorros para protegerse del sol o sentarse cerca de su trabajo debido a las discapacidades visuales comunes a su condición — detalles que ayudan a los sistemas de IA a comprender mejor su vida diaria y qué hace únicas las experiencias de su comunidad.
Cada biblioteca pretende recopilar unas 400 imágenes «del mundo real» de miembros de la comunidad, para capturar una variedad de experiencias con un enfoque en la calidad sobre la cantidad, dice Thieme.
Esas imágenes y anotaciones se convierten en material de formación. Los prompts generados desde la biblioteca se utilizan para crear imágenes con IA, que los miembros de la comunidad revisan y califican según lo bien que encajen con la representación deseada. Con el tiempo, esas valoraciones crean un bucle de retroalimentación, para brindar a los sistemas de IA una visión más clara de lo que significa «bueno», según la definición de la comunidad, dice Thieme.
¿Quién es el propietario de los datos y qué ocurre una vez que se construye una biblioteca comunitaria?
Una parte clave del modelo es que la propia comunidad, a través de la organización de defensa que construyó la biblioteca, la posee y decide si y cómo se comparten los datos, incluido si los pone a disposición de investigadores y desarrolladores en una plataforma como Hugging Face, o si impone límites a su uso.
Eso también significa que la comunidad mantiene la supervisión. Las imágenes se recopilan con consentimiento, y si alguien quiere que sus datos se eliminen más adelante, la comunidad puede hacer ese cambio. Habilitar ese tipo de control no es habitual en la IA hoy en día, dice Thieme, y requirió trabajar en contratos y procesos internos para garantizar que fuera posible.
En un ámbito donde gran parte de los datos actuales de IA se extraen de enormes cantidades de información en línea, con poca visibilidad sobre su origen y sin forma de rastrearlo, ese enfoque devuelve el control a las personas representadas — no solo en cómo se representan, sino en cómo se utilizan sus datos.
«Es poner a las personas, sus datos y sus derechos en primer lugar», dice Thieme, «y creo que necesitamos ver esto mucho más en general en el ámbito de la IA.»
¿Qué sigue con estas bibliotecas? ¿Cómo podrían moldear el futuro de la IA?
Por ahora, el Community Library Creator es una herramienta que se utiliza de manera controlada con organizaciones específicas de defensa. Thieme y Cecily Morrison co-lideran un equipo — que incluye ingenieros e investigadores en áreas como diseño, accesibilidad, aprendizaje automático e interacción humano-ordenador — que trabajan en los controles de ingeniería, seguridad y legales necesarios para expandirse a más comunidades.
Las bibliotecas son una forma de ampliar quién puede moldear la IA en primer lugar — no solo los sistemas en sí, sino los datos y criterios que definen el éxito detrás de ellos. A medida que avance el trabajo, las bibliotecas podrían involucrar a más comunidades y ayudar a entrenar y evaluar futuros modelos de IA.
«En verdad intentamos ampliar la participación en la IA y quién tendrá voz en la configuración del futuro», dice Thieme. «Abrimos caminos y brindamos herramientas para que otros tengan la oportunidad de crear el futuro de la IA. Eso es, para mí, lo más importante.»
Imagen principal: Fotos de bibliotecas comunitarias creadas por organizaciones de defensa que trabajan con la herramienta Community Library Creator de Microsoft. La plataforma ayuda a las comunidades a definir cómo quieren ser representadas y a aportar datos de entrenamiento para sistemas de IA. Fotos proporcionadas por Black Albinism, Kilimanjaro Blind Trust Africa (KBTA), Ottobock y Sociedad de Baja Estatura Kenia (SSSK).
Susanna Ray escribe sobre la IA y la tecnología, con relatos que muestran su impacto real y examinan cómo la innovación transforma el trabajo, los negocios y la sociedad. Antes, reportó para Bloomberg News y otras grandes organizaciones internacionales de noticias en EE. UU. y en el extranjero, donde cubría temas que iban desde política y gobierno hasta negocios y aviación. Sigan su trabajo en Microsoft Source.
You can now use your Network Firewall with all its existing filtering capabilities and features as an explicit forward proxy.
On Nov 25, 2025, AWS introduced Network Firewall proxy in public preview to help customers exert centralized security controls against data exfiltration and malware injection. At the time, the Network Firewall proxy was introduced as a standalone product, separate from Network Firewall transparent firewall and used its own separate proxy security policy. Customers who tested it in preview shared that they want the Network Firewall proxy to maintain parity with Network Firewall’s existing set of capabilities and use the same security policy across the two functionalities. In keeping with customer feedback, we are reintroducing explicit proxy as a functionality of Network Firewall. With this launch, you can configure Network Firewall with your existing Firewall policy in a new no-source-preservation deployment where it can be used as an explicit proxy with all its existing features including managed rule groups, active threat defense, Geo-IP filtering, URL and domain category filtering, container attribute-based rules for Amazon EKS and Amazon ECS, etc. You can create a single security policy and use it for both explicit proxy and transparent firewall functionalities.
Try out AWS Network Firewall in no-source-preservation deployment with proxy functionality in your test environment today in US East (Ohio) region. no-source-preservation Network Firewall is available for free during public preview. For more information, check no-source-preservation Network Firewall documentation.
You can now use your Network Firewall with all its existing filtering capabilities and features as an explicit forward proxy. On Nov 25, 2025, AWS introduced Network Firewall proxy in public preview to help customers exert centralized security controls against data exfiltration and malware injection. At the time, the Network Firewall proxy was introduced as a standalone product, separate from Network Firewall transparent firewall and used its own separate proxy security policy. Customers who tested it in preview shared that they want the Network Firewall proxy to maintain parity with Network Firewall’s existing set of capabilities and use the same security policy across the two functionalities. In keeping with customer feedback, we are reintroducing explicit proxy as a functionality of Network Firewall. With this launch, you can configure Network Firewall with your existing Firewall policy in a new no-source-preservation deployment where it can be used as an explicit proxy with all its existing features including managed rule groups, active threat defense, Geo-IP filtering, URL and domain category filtering, container attribute-based rules for Amazon EKS and Amazon ECS, etc. You can create a single security policy and use it for both explicit proxy and transparent firewall functionalities. Try out AWS Network Firewall in no-source-preservation deployment with proxy functionality in your test environment today in US East (Ohio) region. no-source-preservation Network Firewall is available for free during public preview. For more information, check no-source-preservation Network Firewall documentation.
Amazon Connect Customer now supports exporting cases to a CSV file directly from the agent workspace, making it easier to share case data with internal teams and external stakeholders such as vendors, legal teams, or business partners. Agents can filter and select cases and choose which case fields to include in the export. Admins can control access using a security profile permission.
Cases is available in the following AWS regions: US East (N. Virginia), US West (Oregon), Canada (Central), Europe (Frankfurt), Europe (London), Asia Pacific (Seoul), Asia Pacific (Singapore), Asia Pacific (Sydney), Asia Pacific (Tokyo), and Africa (Cape Town). To learn more and get started, visit the Cases webpage and documentation.
Amazon Connect Customer now supports exporting cases to a CSV file directly from the agent workspace, making it easier to share case data with internal teams and external stakeholders such as vendors, legal teams, or business partners. Agents can filter and select cases and choose which case fields to include in the export. Admins can control access using a security profile permission. Cases is available in the following AWS regions: US East (N. Virginia), US West (Oregon), Canada (Central), Europe (Frankfurt), Europe (London), Asia Pacific (Seoul), Asia Pacific (Singapore), Asia Pacific (Sydney), Asia Pacific (Tokyo), and Africa (Cape Town). To learn more and get started, visit the Cases webpage and documentation.
Amazon EMR on EC2 now supports interactive Apache Spark sessions with Spark Connect. Data engineers and data scientists can develop and debug Apache Spark applications interactively from managed notebooks in Amazon SageMaker Unified Studio and their own IDEs, such as Jupyter and Visual Studio Code, with each session running on dedicated EMR on EC2 clusters. You can also monitor and debug active and completed sessions in the EMR console.
An interactive session provides a persistent Spark context that spans across cells and scripts, letting you blend local Python code execution with remote Spark operations. Spark Connect’s client-server architecture decouples your application client from the Spark driver and allows you to maintain your preferred development environment and tooling while Spark infrastructure runs on the cluster. This architecture supports workflows including ad hoc data exploration, iterative step-by-step debugging, and incremental PySpark job development before deploying to production. For observability, you get real-time session monitoring via the Spark UI, history tracking through the Spark History Server, and session management from the EMR console or API/CLI/SDK.
Interactive Sessions is available on Amazon EMR on EC2 with AWS runtime for Apache Spark (emr-spark-8.0) and later, in all AWS Regions where Amazon EMR is available, except the AWS GovCloud Regions and the China Regions. The Amazon SageMaker Unified Studio experience is available in supported regions. To get started, visit the Interactive sessions with Spark Connect guide or the Amazon SageMaker Unified Studio Getting Started guide.
Amazon EMR on EC2 now supports interactive Apache Spark sessions with Spark Connect. Data engineers and data scientists can develop and debug Apache Spark applications interactively from managed notebooks in Amazon SageMaker Unified Studio and their own IDEs, such as Jupyter and Visual Studio Code, with each session running on dedicated EMR on EC2 clusters. You can also monitor and debug active and completed sessions in the EMR console.
An interactive session provides a persistent Spark context that spans across cells and scripts, letting you blend local Python code execution with remote Spark operations. Spark Connect’s client-server architecture decouples your application client from the Spark driver and allows you to maintain your preferred development environment and tooling while Spark infrastructure runs on the cluster. This architecture supports workflows including ad hoc data exploration, iterative step-by-step debugging, and incremental PySpark job development before deploying to production. For observability, you get real-time session monitoring via the Spark UI, history tracking through the Spark History Server, and session management from the EMR console or API/CLI/SDK.
Interactive Sessions is available on Amazon EMR on EC2 with AWS runtime for Apache Spark (emr-spark-8.0) and later, in all AWS Regions where Amazon EMR is available, except the AWS GovCloud Regions and the China Regions. The Amazon SageMaker Unified Studio experience is available in supported regions. To get started, visit the Interactive sessions with Spark Connect guide or the Amazon SageMaker Unified Studio Getting Started guide.
Today, we are announcing the general availability of Web Search on Amazon Bedrock, a built-in server side tool that performs web search entirely within AWS, enabling OpenAI models (GPT-5.4, GPT-5.5, and GPT-5.6 Sol/Terra/Luna) to ground responses with current web knowledge while maintaining data residency within your secured AWS environment with zero data egress. Previously, adding web grounding required onboarding a third-party search provider, managing separate API keys and billing, building custom orchestration, and conducting additional compliance reviews for each external vendor. Web Search removes this heavy lifting by enable grounding with a single parameter in an existing API call, with no vendor onboarding, no external APIs to orchestrate, and no additional vendor security reviews to conduct.
Web Search is built by Amazon, informed by years of experience across Alexa+, Amazon Quick and Kiro. It combines a web index operated by Amazon, spanning tens of billions of documents refreshed continually, with a built-in knowledge graph that provides verified facts. Rather than returning raw pages, Web Search performs semantic snippet extraction, delivering context-efficient results optimized for the model’s context window with low latency. Web Search integrates through a standardized tool-use interface, compatible with the OpenAI Responses API. Simply add the web search tool to your API call, and Bedrock handles the entire search lifecycle server-side; a single API call returns a grounded response with citations.
Today, we are announcing the general availability of Web Search on Amazon Bedrock, a built-in server side tool that performs web search entirely within AWS, enabling OpenAI models (GPT-5.4, GPT-5.5, and GPT-5.6 Sol/Terra/Luna) to ground responses with current web knowledge while maintaining data residency within your secured AWS environment with zero data egress. Previously, adding web grounding required onboarding a third-party search provider, managing separate API keys and billing, building custom orchestration, and conducting additional compliance reviews for each external vendor. Web Search removes this heavy lifting by enable grounding with a single parameter in an existing API call, with no vendor onboarding, no external APIs to orchestrate, and no additional vendor security reviews to conduct. Web Search is built by Amazon, informed by years of experience across Alexa+, Amazon Quick and Kiro. It combines a web index operated by Amazon, spanning tens of billions of documents refreshed continually, with a built-in knowledge graph that provides verified facts. Rather than returning raw pages, Web Search performs semantic snippet extraction, delivering context-efficient results optimized for the model’s context window with low latency. Web Search integrates through a standardized tool-use interface, compatible with the OpenAI Responses API. Simply add the web search tool to your API call, and Bedrock handles the entire search lifecycle server-side; a single API call returns a grounded response with citations. Web Search on Amazon Bedrock is generally available today in US East (N. Virginia), US East (Ohio), and US West (Oregon). To get started, read our blog post Introducing Web Search on Amazon Bedrock for foundation model grounding, review the Web Search section in the Amazon Bedrock User Guide for technical documentation, and visit the Amazon Bedrock pricing page for cost details.
The AWS Security Hub Extended plan now includes Supply Chain Security as its 10th security category, with Chainguard and Socket as the curated partners. As developers adopt open-source libraries at scale, security teams need confidence that the packages entering their environments are trustworthy and free from malicious code. With this addition, you can detect and block malicious dependencies before they are built into your applications, with the same streamlined activation and pay-as-you-go pricing as every other Extended category. This brings the Extended plan to 23 curated partner solutions. All solutions are on a single AWS bill and no required long-term commitments.
Security Hub Extended is a plan within AWS Security Hub that helps simplify how you procure, deploy, and integrate a full-stack enterprise security solution across endpoint, identity, email, network, data, browser, cloud, AI, security operations, and supply chain. Security findings from all participating solutions are emitted in the Open Cybersecurity Schema Framework (OCSF) and automatically aggregated in AWS Security Hub. With the Extended plan, you can combine AWS and curated partner solutions to quickly identify and respond to risks that span boundaries.
We will continue to expand the Extended plan based on customer feedback. The two new curated partner solutions are available today in all AWS commercial Regions where Security Hub is available. For a list of supported Regions, see the AWS Region table. For more information about pricing, visit the AWS Security Hub pricing page. To get started, visit the AWS Security Hub console or product page.
The AWS Security Hub Extended plan now includes Supply Chain Security as its 10th security category, with Chainguard and Socket as the curated partners. As developers adopt open-source libraries at scale, security teams need confidence that the packages entering their environments are trustworthy and free from malicious code. With this addition, you can detect and block malicious dependencies before they are built into your applications, with the same streamlined activation and pay-as-you-go pricing as every other Extended category. This brings the Extended plan to 23 curated partner solutions. All solutions are on a single AWS bill and no required long-term commitments.
Security Hub Extended is a plan within AWS Security Hub that helps simplify how you procure, deploy, and integrate a full-stack enterprise security solution across endpoint, identity, email, network, data, browser, cloud, AI, security operations, and supply chain. Security findings from all participating solutions are emitted in the Open Cybersecurity Schema Framework (OCSF) and automatically aggregated in AWS Security Hub. With the Extended plan, you can combine AWS and curated partner solutions to quickly identify and respond to risks that span boundaries.
We will continue to expand the Extended plan based on customer feedback. The two new curated partner solutions are available today in all AWS commercial Regions where Security Hub is available. For a list of supported Regions, see the AWS Region table. For more information about pricing, visit the AWS Security Hub pricing page. To get started, visit the AWS Security Hub console or product page.
Starting today, Amazon Elastic Compute Cloud (Amazon EC2) C8g instances are available in AWS Europe (Paris), AWS Africa (Cape Town), AWS Israel (Tel Aviv), and AWS Canada West (Calgary) regions. These instances are powered by AWS Graviton4 processors and deliver up to 30% better performance compared to AWS Graviton3-based instances. Amazon EC2 C8g instances are built for compute-intensive workloads, such as high performance computing (HPC), batch processing, gaming, video encoding, scientific modeling, distributed analytics, CPU-based machine learning (ML) inference, and ad serving. These instances are built on the AWS Nitro System, which offloads CPU virtualization, storage, and networking functions to dedicated hardware and software to enhance the performance and security of your workloads.
AWS Graviton4-based Amazon EC2 instances deliver the best performance and energy efficiency for a broad range of workloads running on Amazon EC2. These instances offer larger instance sizes with up to 3x more vCPUs and memory compared to Graviton3-based Amazon C7g instances. AWS Graviton4 processors are up to 40% faster for databases, 30% faster for web applications, and 45% faster for large Java applications than AWS Graviton3 processors. C8g instances are available in 12 different instance sizes, including two bare metal sizes. They offer up to 50 Gbps enhanced networking bandwidth and up to 40 Gbps of bandwidth to the Amazon Elastic Block Store (Amazon EBS).
Starting today, Amazon Elastic Compute Cloud (Amazon EC2) C8g instances are available in AWS Europe (Paris), AWS Africa (Cape Town), AWS Israel (Tel Aviv), and AWS Canada West (Calgary) regions. These instances are powered by AWS Graviton4 processors and deliver up to 30% better performance compared to AWS Graviton3-based instances. Amazon EC2 C8g instances are built for compute-intensive workloads, such as high performance computing (HPC), batch processing, gaming, video encoding, scientific modeling, distributed analytics, CPU-based machine learning (ML) inference, and ad serving. These instances are built on the AWS Nitro System, which offloads CPU virtualization, storage, and networking functions to dedicated hardware and software to enhance the performance and security of your workloads.
AWS Graviton4-based Amazon EC2 instances deliver the best performance and energy efficiency for a broad range of workloads running on Amazon EC2. These instances offer larger instance sizes with up to 3x more vCPUs and memory compared to Graviton3-based Amazon C7g instances. AWS Graviton4 processors are up to 40% faster for databases, 30% faster for web applications, and 45% faster for large Java applications than AWS Graviton3 processors. C8g instances are available in 12 different instance sizes, including two bare metal sizes. They offer up to 50 Gbps enhanced networking bandwidth and up to 40 Gbps of bandwidth to the Amazon Elastic Block Store (Amazon EBS).
To learn more, see Amazon EC2 C8g Instances. To get started, see the AWS Management Console.
AWS Application Load Balancer (ALB) and Network Load Balancer (NLB) now support new TLS-based security policies that comply with RFC 9151 TLS server requirements for Commercial National Security Algorithm (CNSA) 1.0 suite requirements. These policies implement the cryptographic requirements defined by the US National Security Agency (NSA) for secure communications using TLS 1.2 and TLS 1.3 protocols.
Customers who are required to meet CNSA 1.0 TLS security requirements can now use ALB and NLB with RFC 9151 compliant security policies. Broader interoperability policies are also supported, allowing you to implement CNSA by default while maintaining compatibility with non-CNSA clients during their transition to RFC 9151 compliance, minimizing service disruption.
This feature is available for ALB and NLB in all AWS Commercial Regions, the AWS GovCloud (US) Regions, and the China region at no additional cost. To use this capability, update your existing ALB HTTPS listeners or NLB TLS listeners to a RFC 9151 compliant security policy, or select a compliant policy when creating new listeners through the AWS Management Console, CLI, API, or SDK.
AWS Application Load Balancer (ALB) and Network Load Balancer (NLB) now support new TLS-based security policies that comply with RFC 9151 TLS server requirements for Commercial National Security Algorithm (CNSA) 1.0 suite requirements. These policies implement the cryptographic requirements defined by the US National Security Agency (NSA) for secure communications using TLS 1.2 and TLS 1.3 protocols. Customers who are required to meet CNSA 1.0 TLS security requirements can now use ALB and NLB with RFC 9151 compliant security policies. Broader interoperability policies are also supported, allowing you to implement CNSA by default while maintaining compatibility with non-CNSA clients during their transition to RFC 9151 compliance, minimizing service disruption. This feature is available for ALB and NLB in all AWS Commercial Regions, the AWS GovCloud (US) Regions, and the China region at no additional cost. To use this capability, update your existing ALB HTTPS listeners or NLB TLS listeners to a RFC 9151 compliant security policy, or select a compliant policy when creating new listeners through the AWS Management Console, CLI, API, or SDK. To learn more, visit the ALB User Guide and NLB User Guide documentation. Get started with Elastic Load Balancing.
AWS announces the general availability of Amazon EC2 Storage Optimized I8g instances in AWS Europe (Paris) and Asia Pacific (Jakarta) regions. I8g instances are powered by AWS Graviton4 processors and offer the best compute performance in Amazon EC2 for storage-intensive workloads. I8g instances use the third generation AWS Nitro SSDs, local NVMe storage that deliver up to 65% better real-time storage performance per TB while offering up to 50% lower storage I/O latency and up to 60% lower storage I/O latency variability compared to I4g instances. These instances are built on the AWS Nitro System, which offloads CPU virtualization, storage, and networking functions to dedicated hardware and software enhancing the performance and security for your workloads.
Amazon EC2 I8g instances are designed for I/O intensive workloads that require rapid data access and real-time latency from storage. These instances excel at handling transactional, real-time, distributed databases, including MySQL, PostgreSQL, Hbase and NoSQL solutions like Aerospike, MongoDB, ClickHouse, and Apache Druid. They’re also optimized for real-time analytics platforms such as Apache Spark, data lakehouse and AI LLM pre-processing for training. I8g instances are available in eleven different sizes including two metal sizes, 1.5 TiB of memory, and 45 TB local instance storage. They deliver up to 100 Gbps of network performance bandwidth, and 60 Gbps of dedicated bandwidth for Amazon Elastic Block Store (EBS).
AWS announces the general availability of Amazon EC2 Storage Optimized I8g instances in AWS Europe (Paris) and Asia Pacific (Jakarta) regions. I8g instances are powered by AWS Graviton4 processors and offer the best compute performance in Amazon EC2 for storage-intensive workloads. I8g instances use the third generation AWS Nitro SSDs, local NVMe storage that deliver up to 65% better real-time storage performance per TB while offering up to 50% lower storage I/O latency and up to 60% lower storage I/O latency variability compared to I4g instances. These instances are built on the AWS Nitro System, which offloads CPU virtualization, storage, and networking functions to dedicated hardware and software enhancing the performance and security for your workloads.
Amazon EC2 I8g instances are designed for I/O intensive workloads that require rapid data access and real-time latency from storage. These instances excel at handling transactional, real-time, distributed databases, including MySQL, PostgreSQL, Hbase and NoSQL solutions like Aerospike, MongoDB, ClickHouse, and Apache Druid. They’re also optimized for real-time analytics platforms such as Apache Spark, data lakehouse and AI LLM pre-processing for training. I8g instances are available in eleven different sizes including two metal sizes, 1.5 TiB of memory, and 45 TB local instance storage. They deliver up to 100 Gbps of network performance bandwidth, and 60 Gbps of dedicated bandwidth for Amazon Elastic Block Store (EBS).
To learn more, visit Amazon EC2 I8g instances. To begin your Graviton journey, visit the Level up your compute with AWS Graviton page. To get started, see AWS Management Console, AWS Command Line Interface (AWS CLI), and AWS SDKs.
