All news with #nvidia tag

🚀 Amazon EC2 P6-B300 instances are now available in the Asia Pacific (Seoul) Region. These instances offer 8x NVIDIA Blackwell Ultra GPUs with 2.1 TB high-bandwidth GPU memory, 6.4 Tbps EFA networking, 300 Gbps dedicated ENA throughput, and 4 TB of system memory. The P6-B300 delivers 2x networking bandwidth, 1.5x GPU memory, and 1.5x GPU TFLOPS (FP4) versus P6-B200, targeting large trillion-parameter FMs and LLM training and deployment. The p6-b300.48xlarge size is available in Seoul, US West (Oregon), AWS GovCloud (US-East), and US East (N. Virginia).

🚀 SageMaker AI Studio now offers Generative AI Inference Recommendations, providing a guided low-code/no-code workflow to identify optimal inference configurations for generative workloads. The feature builds on an April 2026 API launch and benchmarks candidate setups on real GPU infrastructure using NVIDIA AIPerf, applying techniques like speculative decoding and kernel tuning. Users pick a use-case profile, optimization goal, and model source, then receive ranked, production-ready recommendations that can be deployed directly to SageMaker endpoints, with only standard compute costs for benchmarking.

🆕 Amazon SageMaker JumpStart now offers three foundation models: Z.ai’s GLM-5.2 FP8, NVIDIA’s Nemotron-Nano-12B-v2, and Z.ai’s GLM-OCR. These models cover long-horizon agentic engineering, efficient hybrid reasoning, and advanced document understanding, enabling customers to deploy high-performance AI on AWS with minimal setup. Each model is optimized for specific enterprise workflows and can be deployed via the SageMaker console or SDK.

🚀 Amazon Elastic Container Service (Amazon ECS) now supports fractional GPU scheduling on Amazon EC2 G6f instances, allowing containers to request GPU partitions as small as one-eighth of an NVIDIA L4 Tensor Core GPU (3 GB). This enables cost-efficient runs for small-model AI inference, model experimentation, and graphics rendering by right-sizing GPU resources. Configure fractional GPUs by setting GPU=0.125, GPU=0.25, or GPU=0.5 in your ECS task definition; support is available on ECS Managed Instances and ECS on EC2 with integrated monitoring and automated instance lifecycle handling.

🚀 Amazon EC2 G7 instances powered by NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs are now available in the Europe (Spain) Region. G7 delivers up to 4.6x AI inference and up to 2.1x graphics performance versus G6, with faster GPU-accelerated analytics. Instances offer up to 8 GPUs (32 GB each), custom Intel Xeon 6 CPUs, up to 192 vCPUs, 768 GiB memory, 700 Gbps EFA, and 7.6 TB NVMe local storage. G7s are available as On-Demand, Spot, and Savings Plans in four regions today.

🔒 Check Point has joined the Open Secure AI Alliance, an initiative introduced by NVIDIA to advance open, measurable, and enterprise-ready AI security. The company will contribute open research, objective benchmarks, datasets and runtime protection experience to support collaborative AI safety and security efforts. This participation aims to help organizations identify, remediate and responsibly disclose vulnerabilities while preserving control over data and infrastructure.

🚀 Amazon SageMaker Studio notebooks now support EC2 G7 instances powered by NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs. G7 offers up to 4.6x AI inference performance versus G6, with up to 8 GPUs and 700 Gbps EFA-enabled bandwidth to accelerate inference, graphics, and analytics workloads. These instances are available in AWS US East (N. Virginia and Ohio) and US West (Oregon). Refer to developer guides for JupyterLab and CodeEditor setup and the pricing page for cost details.

🛡️ NVIDIA has convened nearly 40 technology firms to form the Open Secure AI Alliance, a coalition aimed at building open source security tools for AI, announced on July 27. Members include Adobe, Cisco, Microsoft, CloudStrike, SpaceX, SAP and the Linux Foundation, while notable frontier model developers such as Google, Anthropic and OpenAI are absent. The alliance will focus on finding, fixing and disclosing vulnerabilities, and aims to create an open defense stack for agents, covering identity, isolation, secure model formats and secure coding workflows.

🔒 NVIDIA and 36 organizations have launched the Open Secure AI Alliance to develop and share open technologies, techniques, and tools for securing software and AI agents. The group spans cloud, security, enterprise software, and AI companies including Microsoft, Cisco, CrowdStrike, Hugging Face, IBM, and the Linux Foundation. The alliance’s scope covers identity, permissions, isolation, guardrails, logs, model formats, scanning, and secure coding workflows. Its first technical contribution is NVIDIA-labs OO Agents (NOOA), an Apache 2.0 research framework to test, trace, audit, and govern agent behavior.

🔒 The Open Secure AI Alliance, spearheaded by Nvidia and backed by more than 30 major AI vendors and users, aims to promote open-source defensive AI tools after an incident revealed limitations of closed commercial models. Hugging Face’s forensic work was blocked by safety guardrails on hosted models, forcing it to use an open-weight model on its own infrastructure. The alliance emphasizes that open models and harnesses democratize defense, increase transparency, and allow localized control. OpenAI has not commented on whether it will join the initiative.

🚀 Amazon SageMaker AI now supports EC2 G7e instances in Asia Pacific (Seoul), Europe (London), and Asia Pacific (Tokyo). These instances include up to 8 NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs (96 GB each), 5th Gen Intel Xeon processors, and up to 1,600 Gbps EFA networking, offering up to 2.3x inference performance versus G6e. G7e delivers up to 768 GB total GPU memory enabling single-node serving of medium-to-large LLMs up to 70B with FP8 precision and is suited for LLM inference, image/video generation, spatial and scientific computing.

🚀 Amazon SageMaker AI Inference now supports G6 instances in AWS GovCloud (US‑East). These instances feature up to 8 NVIDIA L4 Tensor Core GPUs with 24 GB each and third‑generation AMD EPYC processors, offering up to 2x inference performance versus G4dn. The expansion enables government and regulated customers to deploy generative AI and computer vision endpoints while meeting compliance and data residency requirements. G6 offers competitive price‑performance for production workloads that fit within 24 GB GPU memory.

🚀 Amazon SageMaker AI Inference now supports G7 instances powered by NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs, delivering up to 4.6x inference performance versus G6. G7 offers 32 GB per GPU, 5th Gen Tensor Cores, up to 700 Gbps EFA-enabled networking, and up to 7.6 TB NVMe local storage, enabling efficient serving of 7B–30B models, image/video generation, and multi-model endpoints. Deploy via the SageMaker console, API, or SDK by selecting ml.g7.* instance types; availability currently includes US East (N. Virginia, Ohio) and US West (Oregon).

🚀 Amazon EC2 G7e instances, powered by NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, are now available in AWS Regions Europe (Frankfurt, Stockholm) and Asia Pacific (Mumbai). These instances deliver up to 2.3x inference performance versus G6e and are optimized for LLMs, multimodal generative AI, agentic AI, and spatial computing. G7e supports up to 8 GPUs with 96 GB each, 192 vCPUs, 1600 Gbps networking, and advanced NVIDIA GPUDirect capabilities for multi-GPU and multi-node workloads.

🚀 Amazon EC2 G7 instances powered by NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs are available in US East (N. Virginia). G7 instances deliver up to 4.6x AI inference and up to 2.1x graphics performance versus G6, and improve GPU-accelerated data analytics. Configurations support up to 8 GPUs (32 GB each), 192 vCPUs, 768 GiB memory, 7.6 TB NVMe, and 700 Gbps networking. G7 is available in US East (N. Virginia, Ohio) and US West (Oregon) and can be purchased On-Demand, Spot, or via Savings Plans.

🔒 Two new models are now available in the Kiro IDE and CLI for the AWS GovCloud (US-West) Region. OpenAI GPT-5.4 supports complex reasoning, coding, document analysis, and multi-step agentic workflows, running on Amazon Bedrock with a 272K context window and 1.2x credit multiplier. NVIDIA Nemotron 3 Super 120B is offered as an open weight, hybrid MoE option with a 256K context window, 32K max output, and 0.25x credit multiplier. Update your IDE or CLI and restart to access the new models.

🚀 Amazon EC2 G6e instances are now generally available for SageMaker notebook instances, offering up to 8 NVIDIA L40s GPUs and third-generation AMD EPYC processors. G6e delivers up to 2.5x better performance versus G5 and supports interactive model testing and training, including generative AI fine-tuning and LLMs up to 13B parameters. G6e is available in multiple US, Europe, Asia-Pacific and Middle East regions.

🛡️ A security firm, AIR, created a benign but deceptive AI agent skill named brand-landingpage, pushed it through a major skill marketplace and promoted it with an Instagram ad, and reports it reached roughly 26,000 agents including corporate accounts. Scanners from vendors like Cisco and NVIDIA marked the package safe because the skill pointed to external setup documentation rather than embedding malicious code. AIR later swapped the external page to deliver a harmless payload that collected email addresses, demonstrating how scanners miss links that can be rewritten after review. The experiment highlights structural trust problems with skills and common mitigations such as pinning versions and vetting external references.

🚀 Amazon EC2 G7e instances deliver up to 8 NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, 96 GB per GPU, 5th Gen Intel Xeon CPUs, up to 192 vCPUs and 1600 Gbps EFA networking. They support NVIDIA GPUDirect P2P and GPUDirect RDMA with EFAv4 for reduced latency in multi-node and multi-GPU workloads. G7e instances target LLMs, agentic and multimodal generative AI, spatial computing, and workloads needing combined graphics and AI acceleration. G7e is now available for SageMaker Studio notebooks in US East (N. Virginia, Ohio) and US West (Oregon).

🚀 Today AWS announces the general availability of Amazon EC2 G7 instances, powered by NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs. G7 delivers up to 4.6x AI inference and 2.1x graphics performance versus G6 and supports AI inference, real-time cinematic graphics, game streaming, and large-scale data analytics. Instances offer up to 8 GPUs with 32 GB each, custom Intel Xeon 6 CPUs, and up to 700 Gbps EFA; available now in US East (Ohio) and US West (Oregon) as On-Demand, Savings Plans, or Spot.