Publicado el Deja un comentario

Amazon SageMaker AI Training Jobs announces general availability of P6-B200 instances powered by NVIDIA B200 GPUs

Today, Amazon SageMaker AI announces the general availability of Amazon EC2 P6-B200 instances in Training Jobs, powered by NVIDIA B200 GPUs. Amazon EC2 P6-B200 instances offer up to 2x performance compared to P5en instances for AI training.

P6-B200 instances feature 8 Blackwell GPUs with 1440 GB of high-bandwidth GPU memory and a 60% increase in GPU memory bandwidth compared to P5en, 5th Generation Intel Xeon processors (Emerald Rapids), and up to 3.2 terabits per second of Elastic Fabric Adapter (EFAv4) networking. P6-B200 instances are powered by the AWS Nitro System, so you can reliably and securely scale AI workloads within Amazon EC2 UltraClusters to tens of thousands of GPUs.

The instances are available through SageMaker HyperPod Flexible Training Plans in US West (Oregon) AWS Region. For on-demand reservation of B200 instances, please reach out to your account manager.

Amazon SageMaker Model Training lets you easily train machine learning models at scale using fully managed infrastructure optimized for performance and cost. To get started with Training Jobs, visit SageMaker Model Training.

 

​Today, Amazon SageMaker AI announces the general availability of Amazon EC2 P6-B200 instances in Training Jobs, powered by NVIDIA B200 GPUs. Amazon EC2 P6-B200 instances offer up to 2x performance compared to P5en instances for AI training. P6-B200 instances feature 8 Blackwell GPUs with 1440 GB of high-bandwidth GPU memory and a 60% increase in GPU memory bandwidth compared to P5en, 5th Generation Intel Xeon processors (Emerald Rapids), and up to 3.2 terabits per second of Elastic Fabric Adapter (EFAv4) networking. P6-B200 instances are powered by the AWS Nitro System, so you can reliably and securely scale AI workloads within Amazon EC2 UltraClusters to tens of thousands of GPUs. The instances are available through SageMaker HyperPod Flexible Training Plans in US West (Oregon) AWS Region. For on-demand reservation of B200 instances, please reach out to your account manager. Amazon SageMaker Model Training lets you easily train machine learning models at scale using fully managed infrastructure optimized for performance and cost. To get started with Training Jobs, visit SageMaker Model Training.  

Publicado el Deja un comentario

Ingest data from Atlassian Jira and Confluence into Amazon OpenSearch Service

Amazon OpenSearch Ingestion now allows you ingest data from Atlassian Jira and Confluence and seamlessly index it in Amazon OpenSearch managed clusters and serverless collections. With this integration, you can now create a unified searchable knowledge base of all your data in Atlassian Jira and Confluence to power your RAG applications.

This integration allows data ingestion with flexible filtering options for projects and types in Jira and spaces and pages in Confluence ensuring that only the information you need is imported. Updates to your data in Jira and Confluence is continuously monitored and automatically synchronized with indices in Amazon OpenSearch Service. To ensure secure and reliable connectivity, multiple authentication methods, including basic API key authentication and OAuth2 authentication, with the added security of managing credentials using a secret stored in AWS Secrets Manager are supported.

This feature is available in all the 16 AWS commercial regions where Amazon OpenSearch Ingestion is currently available: US East (Ohio), US East (N. Virginia), US West (Oregon), US West (N. California), Europe (Ireland), Europe (London), Europe (Frankfurt), Europe (Spain), Asia Pacific (Tokyo), Asia Pacific (Sydney), Asia Pacific (Singapore), Asia Pacific (Mumbai), Asia Pacific (Seoul), Canada (Central), South America (Sao Paulo), and Europe (Stockholm).

To get started, you can start ingesting data from Atlassian Jira and Confluence using the AWS Management Console, AWS SDK, or CLI. To learn more about this feature, see the Amazon OpenSearch Service Developer Guide.

 

​Amazon OpenSearch Ingestion now allows you ingest data from Atlassian Jira and Confluence and seamlessly index it in Amazon OpenSearch managed clusters and serverless collections. With this integration, you can now create a unified searchable knowledge base of all your data in Atlassian Jira and Confluence to power your RAG applications. This integration allows data ingestion with flexible filtering options for projects and types in Jira and spaces and pages in Confluence ensuring that only the information you need is imported. Updates to your data in Jira and Confluence is continuously monitored and automatically synchronized with indices in Amazon OpenSearch Service. To ensure secure and reliable connectivity, multiple authentication methods, including basic API key authentication and OAuth2 authentication, with the added security of managing credentials using a secret stored in AWS Secrets Manager are supported. This feature is available in all the 16 AWS commercial regions where Amazon OpenSearch Ingestion is currently available: US East (Ohio), US East (N. Virginia), US West (Oregon), US West (N. California), Europe (Ireland), Europe (London), Europe (Frankfurt), Europe (Spain), Asia Pacific (Tokyo), Asia Pacific (Sydney), Asia Pacific (Singapore), Asia Pacific (Mumbai), Asia Pacific (Seoul), Canada (Central), South America (Sao Paulo), and Europe (Stockholm). To get started, you can start ingesting data from Atlassian Jira and Confluence using the AWS Management Console, AWS SDK, or CLI. To learn more about this feature, see the Amazon OpenSearch Service Developer Guide.  

Publicado el Deja un comentario

Upgrade Experience from Amazon SageMaker Studio to SageMaker Unified Studio

Amazon SageMaker now offers an upgrade experience that enables customers to transition from SageMaker Studio to SageMaker Unified Studio while preserving their existing resources and maintaining consistent access controls. This new capability allows customers to import their SageMaker AI domains, user profiles, and spaces into SageMaker Unified Studio without redeploying infrastructure. The upgrade tool ensures that identity, authentication, and authorization experiences remain consistent, with users retaining access to only the resources they were previously permitted to use.

With this upgrade experience, customers can continue to access their resources from both SageMaker Studio and SageMaker Unified Studio during the transition period, allowing teams to gradually adapt to the new experience. The tool preserves access to existing JupyterLab and CodeEditor spaces, as well as other SageMaker AI resources like training jobs, ML pipelines, models, inference endpoints etc, previously created from SageMaker Studio. Administrators maintain control over the upgrade process and can disable access to SageMaker Studio once users are comfortable with the SageMaker Unified Studio experience. The upgrade tool is available as an open-source solution that provides a guided, step-by-step process to ensure a smooth transition to SageMaker Unified Studio.

The upgrade experience is available in all AWS Commercial Regions where the next generation of Amazon SageMaker is available. See the supported regions list for more details. To learn more about upgrading from SageMaker Studio to SageMaker Unified Studio, visit the GitHub repository, and to learn more about the next generation of Amazon SageMaker, visit the product detail page.
 

 

​Amazon SageMaker now offers an upgrade experience that enables customers to transition from SageMaker Studio to SageMaker Unified Studio while preserving their existing resources and maintaining consistent access controls. This new capability allows customers to import their SageMaker AI domains, user profiles, and spaces into SageMaker Unified Studio without redeploying infrastructure. The upgrade tool ensures that identity, authentication, and authorization experiences remain consistent, with users retaining access to only the resources they were previously permitted to use. With this upgrade experience, customers can continue to access their resources from both SageMaker Studio and SageMaker Unified Studio during the transition period, allowing teams to gradually adapt to the new experience. The tool preserves access to existing JupyterLab and CodeEditor spaces, as well as other SageMaker AI resources like training jobs, ML pipelines, models, inference endpoints etc, previously created from SageMaker Studio. Administrators maintain control over the upgrade process and can disable access to SageMaker Studio once users are comfortable with the SageMaker Unified Studio experience. The upgrade tool is available as an open-source solution that provides a guided, step-by-step process to ensure a smooth transition to SageMaker Unified Studio. The upgrade experience is available in all AWS Commercial Regions where the next generation of Amazon SageMaker is available. See the supported regions list for more details. To learn more about upgrading from SageMaker Studio to SageMaker Unified Studio, visit the GitHub repository, and to learn more about the next generation of Amazon SageMaker, visit the product detail page.    

Publicado el Deja un comentario

Amazon EFS is now available in the AWS Asia Pacific (Taipei) region

Customers can now create file systems using Amazon Elastic File System (Amazon EFS) in the AWS Asia Pacific (Taipei) Region.

Amazon EFS is designed to provide serverless, fully elastic file storage that lets you share file data without provisioning or managing storage capacity and performance. It is built to scale on demand to petabytes without disrupting applications, growing and shrinking automatically as you add and remove files. Because Amazon EFS has a simple web services interface, you can create and configure file systems quickly and easily. The service is designed to manage file storage infrastructure for you, meaning that you can avoid the complexity of deploying, patching, and maintaining complex file system configurations.

For more information, visit the Amazon EFS product page, and see the AWS Region Table for complete regional availability information.
 

 

​Customers can now create file systems using Amazon Elastic File System (Amazon EFS) in the AWS Asia Pacific (Taipei) Region.
Amazon EFS is designed to provide serverless, fully elastic file storage that lets you share file data without provisioning or managing storage capacity and performance. It is built to scale on demand to petabytes without disrupting applications, growing and shrinking automatically as you add and remove files. Because Amazon EFS has a simple web services interface, you can create and configure file systems quickly and easily. The service is designed to manage file storage infrastructure for you, meaning that you can avoid the complexity of deploying, patching, and maintaining complex file system configurations. For more information, visit the Amazon EFS product page, and see the AWS Region Table for complete regional availability information.    

Publicado el Deja un comentario

Copilot Research Insights: 4 lecciones para líderes

junio 6, 2025

Copilot Research Insights: 4 lecciones para líderes

Ilustración abstracta con elementos gráficos que representan la investigación sobre inteligencia artificial en el entorno laboral, con figuras humanas y vegetación en el fondo.

Ilustración de Emanuela Carnevale

Esta historia aparece en el boletín de WorkLab. Inscríbanse aquí.

Durante el último año, los investigadores de Microsoft han estudiado los efectos de la IA en el trabajo a través de una serie de experimentos del mundo real en empresas que utilizan Copilot. La configuración era simple: un grupo usaba Copilot, el otro no.

En general, las personas que utilizaban Copilot trabajaban más rápido y producían mejores resultados. Pero nuestros hallazgos también revelaron nuevos patrones en la forma en que las personas trabajan, aprenden y se adaptan. He aquí cuatro conclusiones que los líderes pueden aplicar a su propia adopción de la IA a medida que nos adentramos en una era de inteligencia disponible y de creciente colaboración entre humanos y agentes. 

1. La IA ya está integrada en nuestra forma de trabajar

En muchos de nuestros estudios, los miembros del grupo de control, que se suponía que no  debían usar IA, la usaron de todos modos. Esto es coherente con nuestros datos que muestran que los empleados a los que no se les proporciona IA en el trabajo la buscarán ellos mismos (BYOAI). Ya no existe tal cosa como un «no usuario»: la IA ya está integrada en la forma en que las personas piensan, trabajan y hacen las cosas. 

Esto presenta una oportunidad para todos los líderes y empresas. Imaginen que podrían remontarse a los primeros días de Internet, cuando las personas ya lo usaban de manera amplia, pero las empresas aún buscaban descubrir cómo aplicarlo a los negocios. ¿Qué harían diferente? Las Empresas Frontera, estructuradas en torno a la inteligencia bajo demanda e impulsadas por equipos «híbridos» de humanos + agentes, contienen lecciones clave sobre cómo aprovechar esta oportunidad única en una generación y están preparadas para obtener un valor sin precedentes de la IA. Para aprovechar el momento, empiecen por contratar a tus primeros empleados digitales, para establecer su proporción única de agentes humanos e impulsar una adopción amplia y decidida. 

2. Escalar lo que funciona

Para evaluar el rendimiento de los empleados en tareas complejas y específicas del trabajo, nuestros investigadores necesitaban una forma de evaluar la calidad, aunque no fueran expertos en, por ejemplo, el análisis de la causa raíz. Así que recurrieron a Copilot, para construir un calificador de IA para ayudar a comprender la calidad de las respuestas que obtuvieron. En primer lugar, un líder de la empresa en la que se llevó a cabo el estudio validó cómo era una versión de alta calidad del producto final. Con esa vara de medir, Copilot pudo evaluar con objetividad el trabajo de todos los participantes del estudio y asignar a cada uno una puntuación cualitativa. 

Este es un poderoso caso de uso de IA para los líderes empresariales. Según nuestro último Índice de Tendencias Laborales, el 55% de los empleados de Empresas Frontera dicen que pueden asumir más trabajo, en parte porque utilizan Copilot para probar los entregables, comparar la calidad y sacar a la luz los puntos ciegos sin esperar a la revisión del gerente. Esto puede ser tan simple como capacitar a un agente sobre cómo ustedes y otros líderes han criticado el trabajo anterior de su equipo para que los empleados puedan realizar una revisión de primer paso. 

La conclusión general: la verdadera transformación de la IA se produce cuando las organizaciones capturan lo que funciona y lo escalan a todos los equipos. Al formalizar las mejores prácticas, como en el ejemplo del agente anterior, y codificarlas con IA, las empresas pueden garantizar una calidad constante y una toma de decisiones más rápida. Esta capacidad de poner en práctica los aprendizajes de la IA es lo que diferencia a las Empresas Frontera de las que aún están en proceso de experimentarla.

3. La capacitación no puede ser una ocurrencia tardía

Un tema que apareció en toda la investigación de Copilot: las mayores ganancias de rendimiento en realidad se produjeron cuando los sujetos del estudio recibieron orientación sobre cómo aplicar la IA a tareas laborales específicas.

En un estudio, nuestros investigadores desarrollaron prompts óptimos que los sujetos podían usar. Los empleados que recibieron estas indicaciones y consejos sobre la mejor manera de aprovechar la IA vieron el mayor impacto positivo en el rendimiento. La clave es establecer la conexión entre lo que alguien necesita hacer y cómo la IA puede ayudarlos a hacerlo mejor.

Para ver resultados significativos, los líderes deben ir más allá del acceso y centrarse en la habilitación: orientación clara, casos de uso del mundo real y formación que conecte la IA directo con el trabajo que realizan los empleados. Cuando las personas entiendan cómo usar la IA con un propósito, verán mejores resultados.

4. Poner en orden sus procesos

Al igual que la orientación específica ayuda a las personas, los procesos estructurados ayudan a los equipos a sacar el máximo partido de la IA. En todos nuestros experimentos, las mayores mejoras se produjeron cuando los participantes tenían objetivos claros, tareas bien definidas y acceso a recursos organizados, como bibliotecas estructuradas de SharePoint o plantillas compartidas. En esos casos, Copilot actuaba como un potente amplificador. Pero cuando los procesos estaban desorganizados (roles poco claros, documentos desordenados, falta de comprensión compartida), las mejoras en el rendimiento eran limitadas.  

La IA puede acelerar el trabajo, pero no puede desenredar la disfunción. Antes de aplicar la IA, aclaren los objetivos, alineen los equipos y limpien los sistemas en los que las personas confían todos los días.

Cuando se trata de la transformación de la IA, todos aprendemos sobre la marcha. Convertirse en una Empresa Frontera se trata de experimentar, capturar lo que aprenden y aplicar de manera rápida esas lecciones para escalar con velocidad, y empoderar a todos para un nuevo mundo laboral.

The post Copilot Research Insights: 4 lecciones para líderes appeared first on Source LATAM.

 

​The post Copilot Research Insights: 4 lecciones para líderes appeared first on Source LATAM.  

Publicado el Deja un comentario

AWS announces new AWS Direct Connect location and expansion in Taipei, Republic Of China (ROC)

Today, AWS announced the opening of a new AWS Direct Connect location within the Chief Telecom HD data center near Taipei, ROC. By connecting your network to AWS at the new location, you gain private, direct access to all public AWS Regions (except those in China), AWS GovCloud Regions, and AWS Local Zones. This site is the third AWS Direct Connect location within ROC. This Direct Connect location offers dedicated 10 Gbps and 100 Gbps connections with MACsec encryption available.

AWS also announced the addition of 10Gbps and 100Gbps MACsec services in the existing Chunghwa Telecom data center near Taipei, ROC.

The Direct Connect service enables you to establish a private, physical network connection between AWS and your data center, office, or colocation environment. These private connections can provide a more consistent network experience than those made over the public internet.

For more information on the Direct Connect locations worldwide, visit the locations section of the Direct Connect product detail pages. Or, visit our getting started page to learn more about how to purchase and deploy Direct Connect.

 

​Today, AWS announced the opening of a new AWS Direct Connect location within the Chief Telecom HD data center near Taipei, ROC. By connecting your network to AWS at the new location, you gain private, direct access to all public AWS Regions (except those in China), AWS GovCloud Regions, and AWS Local Zones. This site is the third AWS Direct Connect location within ROC. This Direct Connect location offers dedicated 10 Gbps and 100 Gbps connections with MACsec encryption available.
AWS also announced the addition of 10Gbps and 100Gbps MACsec services in the existing Chunghwa Telecom data center near Taipei, ROC. The Direct Connect service enables you to establish a private, physical network connection between AWS and your data center, office, or colocation environment. These private connections can provide a more consistent network experience than those made over the public internet. For more information on the Direct Connect locations worldwide, visit the locations section of the Direct Connect product detail pages. Or, visit our getting started page to learn more about how to purchase and deploy Direct Connect.  

Publicado el Deja un comentario

Announcing ASN match support for AWS WAF

AWS WAF now supports matching incoming request against Autonomous System Numbers (ASNs). By monitoring and restricting traffic from specific ASNs, you can mitigate risks associated with malicious actors, comply with regulatory requirements, and optimize the performance and availability of your web applications. This new ASN Match Statement integrates seamlessly with existing WAF rules, making it easy for you to incorporate ASN based security controls into your overall web application defense strategy.

You can specify a list of ASNs to match against incoming request and take appropriate action such as block or allow the request. You can also use ASN in your rate-based rule statements. These rules aggregate requests according to your criteria, counts and rate limits the requests based on the rule’s evaluation window, request limit, and action settings.

ASN Match statement is available in all regions where AWS WAF is available. The rate-based rule support with ASN is available in regions where the enhanced rate-based rules are currently supported. There is no additional cost for using ASN in Match statement and rate-based rules, however standard AWS WAF charges still apply. For more information about the service, visit the AWS WAF page. For more information about pricing, visit the AWS WAF Pricing page

 

​AWS WAF now supports matching incoming request against Autonomous System Numbers (ASNs). By monitoring and restricting traffic from specific ASNs, you can mitigate risks associated with malicious actors, comply with regulatory requirements, and optimize the performance and availability of your web applications. This new ASN Match Statement integrates seamlessly with existing WAF rules, making it easy for you to incorporate ASN based security controls into your overall web application defense strategy. You can specify a list of ASNs to match against incoming request and take appropriate action such as block or allow the request. You can also use ASN in your rate-based rule statements. These rules aggregate requests according to your criteria, counts and rate limits the requests based on the rule’s evaluation window, request limit, and action settings. ASN Match statement is available in all regions where AWS WAF is available. The rate-based rule support with ASN is available in regions where the enhanced rate-based rules are currently supported. There is no additional cost for using ASN in Match statement and rate-based rules, however standard AWS WAF charges still apply. For more information about the service, visit the AWS WAF page. For more information about pricing, visit the AWS WAF Pricing page  

Publicado el Deja un comentario

AWS Invoice Summary API is now generally available

Today, AWS announces the general availability of the Invoice Summary API. This allows you to retrieve your AWS invoice summary details programmatically via SDK. You can retrieve multiple invoice summary details by making a single API call that accepts input parameters like AWS Account ID, AWS Invoice ID, billing period, or a date range as input.

The output of the Invoice Summary API will include data elements like Invoice Amount in base currency and tax currency, purchase order number and other meta data that can be found in this link. You can integrate the API to your accounts payable systems to automate invoice processing and improve efficiency. 

Invoice Summary API is available in all AWS Regions, except the AWS GovCloud (US) Regions and the China Regions.

 

​Today, AWS announces the general availability of the Invoice Summary API. This allows you to retrieve your AWS invoice summary details programmatically via SDK. You can retrieve multiple invoice summary details by making a single API call that accepts input parameters like AWS Account ID, AWS Invoice ID, billing period, or a date range as input. The output of the Invoice Summary API will include data elements like Invoice Amount in base currency and tax currency, purchase order number and other meta data that can be found in this link. You can integrate the API to your accounts payable systems to automate invoice processing and improve efficiency.  Invoice Summary API is available in all AWS Regions, except the AWS GovCloud (US) Regions and the China Regions.  

Publicado el Deja un comentario

Amazon MWAA now available in additional Region

Amazon Managed Workflows for Apache Airflow (MWAA) is now available in AWS Region Asia Pacific (Malaysia).

Amazon MWAA is a managed service for Apache Airflow that lets you use the same familiar Apache Airflow platform as you do today to orchestrate your workflows and enjoy improved scalability, availability, and security without the operational burden of having to manage the underlying infrastructure. Learn more about using Amazon MWAA on the product page.
Please visit the AWS region table for more information on AWS regions and services. To learn more about Amazon MWAA visit the Amazon MWAA documentation.

Apache, Apache Airflow, and Airflow are either registered trademarks or trademarks of the Apache Software Foundation in the United States and/or other countries.
 

 

​Amazon Managed Workflows for Apache Airflow (MWAA) is now available in AWS Region Asia Pacific (Malaysia). Amazon MWAA is a managed service for Apache Airflow that lets you use the same familiar Apache Airflow platform as you do today to orchestrate your workflows and enjoy improved scalability, availability, and security without the operational burden of having to manage the underlying infrastructure. Learn more about using Amazon MWAA on the product page. Please visit the AWS region table for more information on AWS regions and services. To learn more about Amazon MWAA visit the Amazon MWAA documentation. Apache, Apache Airflow, and Airflow are either registered trademarks or trademarks of the Apache Software Foundation in the United States and/or other countries.    

Publicado el Deja un comentario

Amazon Q Developer Eclipse IDE plugin is now generally available

Amazon Q Developer plugin for the Eclipse IDE is now generally available. With this launch, developers can leverage the power of Amazon Q Developer, the most capable generative AI-powered assistant for software development, within the Eclipse IDE.

Within the Eclipse IDE, you’ll now be able to utilize Amazon Q Developer’s agentic coding experience to seamlessly execute complex workflows. With this coding experience, Q Developer can intelligently take actions on your behalf. It can read your project files to intelligently build the context it needs, suggest code diffs, and run shell commands. As Q Developer works through your tasks, it provides continuous status updates, instantly applying your changes and feedback along the way. This helps Q Developer create code, generate unit tests, and perform code reviews significantly faster, streamlining development workflows across the entire software development lifecycle.

The Amazon Q Developer plugin for the Eclipse IDE is available in all AWS regions where Q Developer is supported. Learn more and download the free Amazon Q Developer plugin for Eclipse to get started.

 

​Amazon Q Developer plugin for the Eclipse IDE is now generally available. With this launch, developers can leverage the power of Amazon Q Developer, the most capable generative AI-powered assistant for software development, within the Eclipse IDE. Within the Eclipse IDE, you’ll now be able to utilize Amazon Q Developer’s agentic coding experience to seamlessly execute complex workflows. With this coding experience, Q Developer can intelligently take actions on your behalf. It can read your project files to intelligently build the context it needs, suggest code diffs, and run shell commands. As Q Developer works through your tasks, it provides continuous status updates, instantly applying your changes and feedback along the way. This helps Q Developer create code, generate unit tests, and perform code reviews significantly faster, streamlining development workflows across the entire software development lifecycle. The Amazon Q Developer plugin for the Eclipse IDE is available in all AWS regions where Q Developer is supported. Learn more and download the free Amazon Q Developer plugin for Eclipse to get started.