Amazon FSx now offers customers the option to use Internet Protocol version 6 (IPv6) for access to Amazon FSx for OpenZFS file systems.

More and more customers are adopting IPv6 to mitigate IPv4 address exhaustion in their private networks or to satisfy government mandates such as the US Office of Management and Budget (OMB) M-21-07 memorandum. With this launch, customers can now access their file systems using IPv4, IPv6, or dual-stack clients without the need for complex infrastructure to handle IPv6 to IPv4 address translation.

IPv6 support for new FSx for OpenZFS file systems is now available in all AWS Commercial, AWS GovCloud (US), and AWS China regions where Amazon FSx is available, with IPv6 support for existing FSx for OpenZFS file systems coming in an upcoming weekly maintenance window. To learn more, visit the Amazon FSx user guide.

​Amazon FSx now offers customers the option to use Internet Protocol version 6 (IPv6) for access to Amazon FSx for OpenZFS file systems. More and more customers are adopting IPv6 to mitigate IPv4 address exhaustion in their private networks or to satisfy government mandates such as the US Office of Management and Budget (OMB) M-21-07 memorandum. With this launch, customers can now access their file systems using IPv4, IPv6, or dual-stack clients without the need for complex infrastructure to handle IPv6 to IPv4 address translation. IPv6 support for new FSx for OpenZFS file systems is now available in all AWS Commercial, AWS GovCloud (US), and AWS China regions where Amazon FSx is available, with IPv6 support for existing FSx for OpenZFS file systems coming in an upcoming weekly maintenance window. To learn more, visit the Amazon FSx user guide.  

 AWS IAM Identity Center introduces support for user background sessions, a new feature which allows Amazon SageMaker Studio users to initiate long-running jobs that continue to run in the background even after a user logs off from their computer. User background sessions leverage permissions tied to a user’s identity and their corporate directory group associations while accessing data across multiple AWS services including Amazon Athena, Amazon EMR, AWS Lake Formation, Amazon Redshift, and Amazon S3.

Building on IAM Identity Center’s trusted identity propagation (TIP) capability, this feature allows SageMaker Studio jobs to run for up to 90 days without requiring users to stay signed in. When launching jobs such as SageMaker Training, Processing, or Pipeline executions, IAM Identity Center creates a new user background session that runs independently until job completion or the background session duration limit is reached, whichever comes first. Administrators can track the creation of user background sessions via AWS CloudTrail events, update the background session duration for their IAM Identity Center instance and terminate background sessions for different users in their organization.

AWS IAM Identity Center helps you securely connect your workforce identities and manage their access centrally across AWS applications. For more information on user background sessions, see here. For the list of supported regions and compatible AWS services supported with user background sessions, refer to the Amazon SageMaker Studio page, linked here. 

​ AWS IAM Identity Center introduces support for user background sessions, a new feature which allows Amazon SageMaker Studio users to initiate long-running jobs that continue to run in the background even after a user logs off from their computer. User background sessions leverage permissions tied to a user’s identity and their corporate directory group associations while accessing data across multiple AWS services including Amazon Athena, Amazon EMR, AWS Lake Formation, Amazon Redshift, and Amazon S3.
Building on IAM Identity Center’s trusted identity propagation (TIP) capability, this feature allows SageMaker Studio jobs to run for up to 90 days without requiring users to stay signed in. When launching jobs such as SageMaker Training, Processing, or Pipeline executions, IAM Identity Center creates a new user background session that runs independently until job completion or the background session duration limit is reached, whichever comes first. Administrators can track the creation of user background sessions via AWS CloudTrail events, update the background session duration for their IAM Identity Center instance and terminate background sessions for different users in their organization.
AWS IAM Identity Center helps you securely connect your workforce identities and manage their access centrally across AWS applications. For more information on user background sessions, see here. For the list of supported regions and compatible AWS services supported with user background sessions, refer to the Amazon SageMaker Studio page, linked here.   

We are excited to announce that Amazon SageMaker Studio now supports trusted identity propagation (TIP), enabling admins to trace actions taken in SageMaker Studio back to a human user. It also enables administrators to manage permissions based on user identity to AWS Lake Formation and Amazon S3 Access Grants.

Starting today, SageMaker Studio now supports trusted identity propagation for AWS services, including AWS Lake Formation, Amazon S3, Amazon EMR, Amazon EMR-Serverless, Amazon Redshift and Amazon Athena. Customers can enable trusted identity propagation for SageMaker Studio domains, either at domain creation or while updating an existing domain. With this capability, customers can apply fine grained access controls using S3 Access Grants, Lake Formation, or Redshift Data APIs to govern user access to data in their SageMaker Studio notebooks. Customers’ Training and Processing jobs in Studio notebooks will integrate with S3 Access Grants.

With this change, administrators can trace creation of user interactive sessions as well as user background sessions across Studio applications (JupyterLab and CodeEditor) via AWS CloudTrail events. As an example, administrators can now track which user creates SageMaker Pipeline steps to run experiments using SageMaker MLFlow, trains ML models using SageMaker Training, and deploys using SageMaker Inference.

This feature is available in all AWS Regions where Amazon SageMaker Studio with IdC support is currently available, excluding China regions, and GovCloud (US) regions. To learn more, visit our documentation.

​We are excited to announce that Amazon SageMaker Studio now supports trusted identity propagation (TIP), enabling admins to trace actions taken in SageMaker Studio back to a human user. It also enables administrators to manage permissions based on user identity to AWS Lake Formation and Amazon S3 Access Grants. Starting today, SageMaker Studio now supports trusted identity propagation for AWS services, including AWS Lake Formation, Amazon S3, Amazon EMR, Amazon EMR-Serverless, Amazon Redshift and Amazon Athena. Customers can enable trusted identity propagation for SageMaker Studio domains, either at domain creation or while updating an existing domain. With this capability, customers can apply fine grained access controls using S3 Access Grants, Lake Formation, or Redshift Data APIs to govern user access to data in their SageMaker Studio notebooks. Customers’ Training and Processing jobs in Studio notebooks will integrate with S3 Access Grants. With this change, administrators can trace creation of user interactive sessions as well as user background sessions across Studio applications (JupyterLab and CodeEditor) via AWS CloudTrail events. As an example, administrators can now track which user creates SageMaker Pipeline steps to run experiments using SageMaker MLFlow, trains ML models using SageMaker Training, and deploys using SageMaker Inference. This feature is available in all AWS Regions where Amazon SageMaker Studio with IdC support is currently available, excluding China regions, and GovCloud (US) regions. To learn more, visit our documentation.  

Amazon DocumentDB (with MongoDB compatibility) offers Extended Support for version 3.6, allowing customers to maintain critical workloads on version 3.6 for up to three years beyond the standard support end date on March 30, 2026. This is designed for customers who may need more time to plan and implement version upgrades, especially when navigating application dependencies or managing enterprise-scale deployments.

Standard support for Amazon DocumentDB version 3.6 will end on March 30, 2026. After this date, clusters not upgraded to a supported version will incur Extended Support charges. During the Extended Support period, AWS will continue to provide critical security updates for Common Vulnerabilities and Exposures (CVEs). To perform an upgrade from Amazon DocumentDB version 3.6 to 5.0, customers can use an in-place major version upgrade. Upgrading to version 5.0 of Amazon DocumentDB enables customers to benefit from features such as I/O-Optimized, Graviton2, and Serverless instances, providing price per performance improvements. To get started and learn more about the Extended Support, see our blog.

​Amazon DocumentDB (with MongoDB compatibility) offers Extended Support for version 3.6, allowing customers to maintain critical workloads on version 3.6 for up to three years beyond the standard support end date on March 30, 2026. This is designed for customers who may need more time to plan and implement version upgrades, especially when navigating application dependencies or managing enterprise-scale deployments. Standard support for Amazon DocumentDB version 3.6 will end on March 30, 2026. After this date, clusters not upgraded to a supported version will incur Extended Support charges. During the Extended Support period, AWS will continue to provide critical security updates for Common Vulnerabilities and Exposures (CVEs). To perform an upgrade from Amazon DocumentDB version 3.6 to 5.0, customers can use an in-place major version upgrade. Upgrading to version 5.0 of Amazon DocumentDB enables customers to benefit from features such as I/O-Optimized, Graviton2, and Serverless instances, providing price per performance improvements. To get started and learn more about the Extended Support, see our blog.  

You can now update your Amazon DynamoDB table throughput mode from provisioned capacity to on-demand mode up to four times in a rolling 24-hour period. This enhancement makes it simpler for customers who have use cases which require loading large volumes of data into their DynamoDB tables multiple times per day, or want greater flexibility to manage their workload requirements and optimize costs. The increased quota also simplifies CloudFormation deployments and rollbacks, allowing templates to execute more readily without being blocked by previous changes. Previously, you could only set or update your table throughput mode to on-demand capacity once in a 24-hour period.

DynamoDB on-demand mode offers a truly serverless database experience with pay-per-request pricing and automatic scaling to accommodate the most demanding workloads without capacity planning. Many customers prefer the simplicity of on-demand mode to build modern, serverless applications that can start small and scale to millions of requests per second.

There is no charge for switching capacity modes, and the increased quota automatically applies to all DynamoDB tables. On-demand throughput is available in all AWS Regions. For more information about DynamoDB on-demand throughput, see the Developer Guide.

​You can now update your Amazon DynamoDB table throughput mode from provisioned capacity to on-demand mode up to four times in a rolling 24-hour period. This enhancement makes it simpler for customers who have use cases which require loading large volumes of data into their DynamoDB tables multiple times per day, or want greater flexibility to manage their workload requirements and optimize costs. The increased quota also simplifies CloudFormation deployments and rollbacks, allowing templates to execute more readily without being blocked by previous changes. Previously, you could only set or update your table throughput mode to on-demand capacity once in a 24-hour period. DynamoDB on-demand mode offers a truly serverless database experience with pay-per-request pricing and automatic scaling to accommodate the most demanding workloads without capacity planning. Many customers prefer the simplicity of on-demand mode to build modern, serverless applications that can start small and scale to millions of requests per second. There is no charge for switching capacity modes, and the increased quota automatically applies to all DynamoDB tables. On-demand throughput is available in all AWS Regions. For more information about DynamoDB on-demand throughput, see the Developer Guide.  

• Categoría: Noticias de Microsoft

agosto 13, 2025

Compartimos instrucciones prácticas: lanzamiento de patrones y prácticas de Microsoft Secure Future Initiative (SFI)

Por: Joy Chik y Ann Johnson.

Nos complace lanzar los patrones y prácticas de Microsoft Secure Future Initiative (SFI): una nueva biblioteca de instrucciones prácticas diseñadas para ayudar a las organizaciones a implementar medidas de seguridad a escala.

Este lanzamiento marca el siguiente paso en nuestro recorrido para hacer que nuestros aprendizajes de SFI sean prácticos para nuestros clientes, socios y un ecosistema de seguridad más amplio. Estos patrones y prácticas se basan en una variedad de arquitecturas de seguridad probadas y procedimientos recomendados, incluidos, entre otros, Confianza cero, operacionalizados para proteger la infraestructura de Microsoft y ahora compartidos para ayudarlos a hacer lo mismo.

Descubran más sobre Microsoft Secure Future Initiative

Por qué son importantes los patrones y prácticas de SFI

Desde el lanzamiento de Secure Future Initiative (SFI) en noviembre de 2023, hemos movilizado el equivalente a más de 34 mil ingenieros para mitigar el riesgo y mejorar la seguridad de Microsoft y nuestros clientes.¹ Guiados por tres principios de seguridad: seguridad por diseño, por defecto y en operaciones, hemos logrado un progreso medible en las áreas de cultura, gobernanza y nuestros seis pilares de ingeniería. Aún hay más por hacer y los equipos de toda la empresa trabajan para mejorar la seguridad de cada producto, abordar los aprendizajes de cada incidente y mejorar de manera continua nuestros métodos y prácticas.

Lean el último informe de SFI de abril de 2025

Además, hemos escuchado comentarios de clientes y socios que quieren que compartamos cómo mejoramos la seguridad en Microsoft, no solo a nivel de arquitectura estratégica, sino también a nivel de implementación y práctica. Ahí es donde entra en juego la biblioteca de patrones y prácticas de SFI.

¿Qué hay en la primera ola de patrones y prácticas de SFI?

Lanzamos la primera ola de ocho artículos de patrones y prácticas que ayudan a resolver los desafíos más solicitados, urgentes y complejos que enfrentan los profesionales de la seguridad en la actualidad:

Nombre del patrón	Pilar SFI	Lo que les ayuda a hacer
Autenticación multifactor (MFA, por sus siglas en inglés) resistente al phishing	Protección de identidades y secretos	La MFA tradicional ya no es suficiente. Este patrón ayuda a las organizaciones a cambiar a la autenticación criptográfica resistente al phishing mediante FIDO2, claves de acceso y métodos basados en certificados, lo que reduce la exposición a ciberataques basados en credenciales. Reemplacen MFA vulnerable con métodos criptográficos resistentes a la suplantación de identidad.
Eliminación del movimiento lateral de identidad	Aislamiento de inquilinos y sistemas de producción	Los ciberatacantes a menudo explotan las rutas de pivote de identidad para escalar privilegios. Este patrón describe cómo segmentar el acceso, aplicar el acceso condicional y bloquear la autenticación de invitado de riesgo para evitar intrusiones silenciosas. Eviten que los ciberatacantes pivoten entre inquilinos y roles.
Eliminación de los sistemas heredados que ponen en riesgo la seguridad	Aislamiento de inquilinos y sistemas de producción	Los inquilinos no administrados y la infraestructura heredada introducen desviaciones de configuración y superficie de ataque. Microsoft eliminó más de 5,75 millones de inquilinos inactivos: este patrón muestra cómo ustedes pueden hacer lo mismo.   Retirar inquilinos no administrados e infraestructura heredada.
Estandarización de las canalizaciones de desarrollo seguras	Protección de sistemas de ingeniería	Las canalizaciones de CI/CD suelen estar fragmentadas e incoherentes. Este patrón le ayuda a implementar plantillas gobernadas que aplican puertas de seguridad, fomenta la creación de listas de materiales de software (SBOM, por sus siglas en inglés) y agiliza el cumplimiento.   Utilicen plantillas de CI/CD gobernadas para aplicar la seguridad y el cumplimiento.
Inventario completo de la infraestructura de producción	Monitoreo y detección de amenazas	No pueden proteger lo que no pueden ver. Este patrón guía a las organizaciones en la creación de inventarios de activos en tiempo real, la centralización de la telemetría y la eliminación de aplicaciones no utilizadas para reducir el riesgo.   Mantengan la visibilidad en tiempo real de todos los activos y la telemetría.
Detección y respuesta rápidas de anomalías	Monitoreo y detección de amenazas	Los ciberatacantes modernos se mueven rápido. Este patrón muestra cómo usar IA, análisis de comportamiento y entidades de usuario (UEBA, por sus siglas en inglés) y registro centralizado para detectar comportamientos sospechosos y automatizar la respuesta, lo que reduce el tiempo de permanencia y mejora la eficiencia del centro de operaciones de seguridad (SOC, por sus siglas en inglés).   Utilicen la IA y el análisis del comportamiento para detectar y responder a las ciberamenazas más rápido.
Estándares de retención de registros de seguridad	Monitoreo y detección de amenazas	Los registros son la columna vertebral de la detección y el análisis forense. Este patrón les ayuda a estandarizar formatos, centralizar el acceso y ampliar la retención para admitir investigaciones y cumplimiento a largo plazo.   Estandaricen, centralicen y amplíen la retención de registros para una mejor detección y análisis forense.
Aceleración de la mitigación de vulnerabilidades	Acelerar la respuesta y la corrección	La gestión eficaz de vulnerabilidades es fundamental para reducir el riesgo en entornos digitales complejos al permitir respuestas más rápidas y sistemáticas a las amenazas de seguridad. La automatización, los flujos de trabajo integrados y las comunicaciones enriquecidas pueden acelerar de manera significativa los plazos de mitigación y mejorar la resiliencia de la organización.   Automaticen la detección, la clasificación y la aplicación de parches para reducir el tiempo de mitigación.

Introducción a la taxonomía de patrones y prácticas de SFI

Al igual que los patrones de diseño de software brindan soluciones reutilizables a problemas comunes de ingeniería, los patrones y prácticas de SFI ofrecen enfoques repetibles y probados para resolver desafíos complejos de ciberseguridad. Cada patrón está diseñado para abordar un riesgo de seguridad específico, ya sea el movimiento lateral de identidad, la infraestructura heredada o las canalizaciones de integración continua y entrega continua (CI/CD, por sus siglas en inglés) incoherentes, y se basa en la propia experiencia de Microsoft. Al igual que los patrones de diseño en la arquitectura de software, estos patrones de seguridad son modulares, extensibles y están diseñados para su reutilización en diversos entornos.

Además, cada patrón en la biblioteca de patrones y prácticas de SFI sigue una estructura consistente y con un propósito. Cada artículo comienza con un nombre de patrón, un identificador conciso que captura la esencia del desafío de ciberseguridad. La  sección de problemas describe el riesgo de seguridad y su contexto en el mundo real, lo que ayuda a los lectores a comprender por qué es importante. La solución describe cómo Microsoft abordó el problema a nivel interno. La sección de orientación proporciona recomendaciones prácticas que los clientes pueden considerar aplicar en sus propios entornos. Por último, la  sección de implicaciones describe los resultados y las compensaciones de implementar el patrón, lo que ayuda a las organizaciones a anticipar tanto los beneficios como las consideraciones operativas.

Esta estructura ofrece un marco para comprender, aplicar y evolucionar las prácticas de seguridad.

Unirse al recorrido de patrones y prácticas de SFI

Los patrones y prácticas de SFI son su guía para convertir la arquitectura en acción. Al adoptar estos patrones, las organizaciones pueden acelerar su madurez de seguridad, reducir la fricción de implementación y crear sistemas que sean más seguros por diseño, valor predeterminado y funcionamiento.

¿Qué viene después?

Esto es solo el comienzo. En los próximos meses, publicaremos patrones adicionales para compartir más orientación alineada con los pilares de SFI. Cada nuevo patrón se publicará en el blog de seguridad de Microsoft y en la página principal de la Iniciativa para el Futuro Seguro de Microsoft.

Manténganse al día con todas las actualizaciones de SFI

Comenzar

Exploren el primer conjunto de patrones:

• Página temática del blog de seguridad SFI
• Seguro por diseño: un kit de herramientas de UX
• Blog de la comunidad tecnológica de seguridad de Microsoft
• Cuenta de Microsoft Security X

Construyamos juntos un futuro seguro

Hablen con su equipo de cuentas de Microsoft para integrar estas prácticas en su hoja de ruta.

Para obtener más información sobre las soluciones de seguridad de Microsoft, visiten nuestro sitio web. Agreguen a Favoritos el blog de Seguridad para mantenerse al día con nuestra cobertura experta en asuntos de seguridad. Además, síganos en LinkedIn (Microsoft Security) y X (@MSFTSecurity) para conocer las últimas noticias y actualizaciones sobre ciberseguridad.

¹Informe de la iniciativa Microsoft Secure Future, noviembre de 2024

Etiquetas:

• Seguridad
• Tecnología

The post Compartimos instrucciones prácticas: lanzamiento de patrones y prácticas de Microsoft Secure Future Initiative (SFI) appeared first on Source LATAM.

​The post Compartimos instrucciones prácticas: lanzamiento de patrones y prácticas de Microsoft Secure Future Initiative (SFI) appeared first on Source LATAM.  

Today, AWS announces new Amazon Elastic Compute Cloud (Amazon EC2) P5 instance size with one NVIDIA H100 GPU that allows businesses to right-size their machine learning (ML) and high-performance computing (HPC) resources with cost-effectiveness.

The new instance size enables customers to start small and scale in granular increments, providing more flexible control over infrastructure costs. Customers developing small to medium Large Language Models (LLMs) such as chatbots or specialized language translation tools can now run inference tasks more economically. Customers can also use these instances to deploy HPC applications for pharmaceutical discovery, fluid flow analysis, and financial modeling without committing to expensive, large-scale GPU deployments.

P5.4xlarge instances are now available through Amazon EC2 Capacity Blocks for ML in the following AWS Regions: US East (North Virginia, Ohio), US West (Oregon), Europe (London), Asia Pacific (Mumbai, Sydney, Tokyo) and South America (Sao Paulo) regions. These instances can be purchased On-Demand, Spot or through Savings Plans in Europe (London), Asia Pacific (Mumbai, Jakarta, Tokyo), and South America (Sao Paulo) regions.

To learn more about P5.4xlarge instances, visit Amazon EC2 P5 instances.

​Today, AWS announces new Amazon Elastic Compute Cloud (Amazon EC2) P5 instance size with one NVIDIA H100 GPU that allows businesses to right-size their machine learning (ML) and high-performance computing (HPC) resources with cost-effectiveness.
The new instance size enables customers to start small and scale in granular increments, providing more flexible control over infrastructure costs. Customers developing small to medium Large Language Models (LLMs) such as chatbots or specialized language translation tools can now run inference tasks more economically. Customers can also use these instances to deploy HPC applications for pharmaceutical discovery, fluid flow analysis, and financial modeling without committing to expensive, large-scale GPU deployments.
P5.4xlarge instances are now available through Amazon EC2 Capacity Blocks for ML in the following AWS Regions: US East (North Virginia, Ohio), US West (Oregon), Europe (London), Asia Pacific (Mumbai, Sydney, Tokyo) and South America (Sao Paulo) regions. These instances can be purchased On-Demand, Spot or through Savings Plans in Europe (London), Asia Pacific (Mumbai, Jakarta, Tokyo), and South America (Sao Paulo) regions.
To learn more about P5.4xlarge instances, visit Amazon EC2 P5 instances.  

Starting today, the Amazon Elastic Compute Cloud (Amazon EC2) G6 instances powered by NVIDIA L4 GPUs are now available in AWS GovCloud (US-East). G6 instances can be used for a wide range of graphics-intensive and machine learning use cases.

Customers can use G6 instances for deploying ML models for natural language processing, language translation, video and image analysis, speech recognition, and personalization as well as graphics workloads, such as creating and rendering real-time, cinematic-quality graphics and game streaming. G6 instances feature up to 8 NVIDIA L4 Tensor Core GPUs with 24 GB of memory per GPU and third generation AMD EPYC processors. They also support up to 192 vCPUs, up to 100 Gbps of network bandwidth, and up to 7.52 TB of local NVMe SSD storage.

Amazon EC2 G6 instances are already available today in the AWS US East (N. Virginia and Ohio) , US West (Oregon), Europe (Frankfurt, London, Paris, Spain, Stockholm and Zurich), Asia Pacific (Mumbai, Tokyo, Malaysia, Seoul and Sydney), South America (Sao Paulo) and Canada (Central) regions. Customers can purchase G6 instances as On-Demand Instances, Reserved Instances, Spot Instances, or as part of Savings Plans.

To get started, visit the AWS Management Console, AWS Command Line Interface (CLI), and AWS SDKs. To learn more, visit the G6 instance page.

​Starting today, the Amazon Elastic Compute Cloud (Amazon EC2) G6 instances powered by NVIDIA L4 GPUs are now available in AWS GovCloud (US-East). G6 instances can be used for a wide range of graphics-intensive and machine learning use cases.
Customers can use G6 instances for deploying ML models for natural language processing, language translation, video and image analysis, speech recognition, and personalization as well as graphics workloads, such as creating and rendering real-time, cinematic-quality graphics and game streaming. G6 instances feature up to 8 NVIDIA L4 Tensor Core GPUs with 24 GB of memory per GPU and third generation AMD EPYC processors. They also support up to 192 vCPUs, up to 100 Gbps of network bandwidth, and up to 7.52 TB of local NVMe SSD storage.
Amazon EC2 G6 instances are already available today in the AWS US East (N. Virginia and Ohio) , US West (Oregon), Europe (Frankfurt, London, Paris, Spain, Stockholm and Zurich), Asia Pacific (Mumbai, Tokyo, Malaysia, Seoul and Sydney), South America (Sao Paulo) and Canada (Central) regions. Customers can purchase G6 instances as On-Demand Instances, Reserved Instances, Spot Instances, or as part of Savings Plans.
To get started, visit the AWS Management Console, AWS Command Line Interface (CLI), and AWS SDKs. To learn more, visit the G6 instance page.  

AWS Deadline Cloud is a fully managed service that simplifies render management for teams creating computer-generated graphics and visual effects for films, television, broadcasting, web content, and design. Today, we’re excited to announce a new wait and save feature for Deadline Cloud service-managed fleets that can reduce rendering costs with prices starting as low as $0.006 per vCPU-hour.

This new feature is ideal for non time-sensitive rendering workloads with flexible completion times. Submitting jobs using this wait and save approach allows you to achieve significant cost savings so you can do more creative iteration and exploration on your next project. This feature complements existing AWS Deadline Cloud compute options in its service-managed fleets, giving you more flexibility to optimize your resource utilization across different priorities and budgets.

AWS Deadline Cloud wait and save is available in all AWS Regions where AWS Deadline Cloud is offered. To learn more about this new cost-saving feature and how it can help optimize your rendering workloads, visit the AWS Deadline Cloud product page or review the AWS Deadline Cloud documentation.

​AWS Deadline Cloud is a fully managed service that simplifies render management for teams creating computer-generated graphics and visual effects for films, television, broadcasting, web content, and design. Today, we’re excited to announce a new wait and save feature for Deadline Cloud service-managed fleets that can reduce rendering costs with prices starting as low as $0.006 per vCPU-hour. This new feature is ideal for non time-sensitive rendering workloads with flexible completion times. Submitting jobs using this wait and save approach allows you to achieve significant cost savings so you can do more creative iteration and exploration on your next project. This feature complements existing AWS Deadline Cloud compute options in its service-managed fleets, giving you more flexibility to optimize your resource utilization across different priorities and budgets. AWS Deadline Cloud wait and save is available in all AWS Regions where AWS Deadline Cloud is offered. To learn more about this new cost-saving feature and how it can help optimize your rendering workloads, visit the AWS Deadline Cloud product page or review the AWS Deadline Cloud documentation.  

Today, Amazon SageMaker AI announces support for P6e-GB200 UltraServers in SageMaker HyperPod and Training Jobs. With P6e-GB200 UltraServers, you can leverage up to 72 NVIDIA Blackwell GPUs under one NVLink domain to accelerate training and deployment of foundational models at trillion-parameter scale. P6e-GB200 UltraServers are available in two sizes: ml.u-p6e-gb200x72 (72 GPUs within NVLink) and ml.u-p6e-gb200x36 (36 GPUs within NVLink).

P6e-GB200 UltraServers deliver over 20x compute and over 11x memory under one NVIDIA NVLink compared to P5en instances. Within each NVLink domain you can leverage 360 petaflops of FP8 compute (without sparsity) and 13.4 TB of total high bandwidth memory (HBM3e). When you use P6e-GB200 UltraServers on SageMaker AI, you get the GB200’s superior performance combined with SageMaker’s managed infrastructure such as security, built-in fault tolerance, topology aware scheduling (SageMaker HyperPod EKS & Slurm), integrated monitoring capabilities, and native integration with other SageMaker AI and AWS services.

The UltraServers are available through SageMaker Flexible Training Plans in the Dallas Local Zone («us-east-1-dfw-2a»), an extension of the US East (N. Virginia) AWS Region. For on-demand reservation of GB200 UltraServers, please reach out to your account manager. Amazon SageMaker AI lets you easily train and deploy machine learning models at scale using fully managed infrastructure optimized for performance and cost. To get started with UltraServers on SageMaker AI, visit the documentation.

​Today, Amazon SageMaker AI announces support for P6e-GB200 UltraServers in SageMaker HyperPod and Training Jobs. With P6e-GB200 UltraServers, you can leverage up to 72 NVIDIA Blackwell GPUs under one NVLink domain to accelerate training and deployment of foundational models at trillion-parameter scale. P6e-GB200 UltraServers are available in two sizes: ml.u-p6e-gb200x72 (72 GPUs within NVLink) and ml.u-p6e-gb200x36 (36 GPUs within NVLink). P6e-GB200 UltraServers deliver over 20x compute and over 11x memory under one NVIDIA NVLink compared to P5en instances. Within each NVLink domain you can leverage 360 petaflops of FP8 compute (without sparsity) and 13.4 TB of total high bandwidth memory (HBM3e). When you use P6e-GB200 UltraServers on SageMaker AI, you get the GB200’s superior performance combined with SageMaker’s managed infrastructure such as security, built-in fault tolerance, topology aware scheduling (SageMaker HyperPod EKS & Slurm), integrated monitoring capabilities, and native integration with other SageMaker AI and AWS services. The UltraServers are available through SageMaker Flexible Training Plans in the Dallas Local Zone («us-east-1-dfw-2a»), an extension of the US East (N. Virginia) AWS Region. For on-demand reservation of GB200 UltraServers, please reach out to your account manager. Amazon SageMaker AI lets you easily train and deploy machine learning models at scale using fully managed infrastructure optimized for performance and cost. To get started with UltraServers on SageMaker AI, visit the documentation.