The auto-updated technology brief
NVIDIA introduces NVLink as a high-performance scale-up network designed to meet the growing demands of AI factories, enabling faster deployment of AI compute infrastructure for larger, more complex workloads.
Knowledge workers are increasingly integrating AI agents as digital coworkers to boost productivity, but security remains a key concern when connecting large language models.
Capcom's RE ENGINE team implemented path tracing technology in two games, Resident Evil Requiem and PRAGMATA, each with distinct visual styles.
NVIDIA introduces RTX Sensor Simulation integration for developers building 3D, simulation, robotics, and digital twin applications to incorporate physical AI capabilities into existing tools and workflows.
NVIDIA CUDA 13.3 introduces native support for carryless multiplication with a new PTX instruction, enabling faster cryptographic operations on GPUs.
Apple's Q3 2026 earnings showed strong iPhone sales but declining revenue in services and concerns over memory prices.
AWS announced a one-click Lambda setup prompt for coding agents, general availability of OpenAI GPT-5.6 models on Amazon Bedrock, same-day transitions to S3 Standard-IA and One Zone-IA, self-managed code storage on Lambda, and password hash imports for Amazon Cognito users. Additional updates include Amazon SQS's 20th anniversary, open AI protocols with Strands Agents SDK, an open source Bulk Executor for DynamoDB, and Kiro CLI for AWS Support workflows.
AWS Builder Center celebrates its first anniversary with expanded features including sandbox environments and workshops. AWS Security Hub introduces Network Scanning for public internet reachability detection and extends monitoring to Microsoft Azure. Amazon SageMaker Studio integrates with Hugging Face for streamlined AI model deployment. Other updates include reduced GPU management fees for Amazon ECS and EKS, general availability of Amazon Aurora DSQL CDC, and new tools like Loom for AWS and Claude apps gateway.
ModelExpress by NVIDIA addresses the high cost of moving large model checkpoints, enabling faster distribution of model weights across clusters for tasks like cold starts, autoscaling, and rolling updates.
NVIDIA OptiX is a GPU ray tracing engine framework. The NVIDIA OptiX Toolkit (OTK) offers debugging tools to diagnose issues like invalid API arguments, black frames, or GPU-side bugs in ray tracing applications.
NVIDIA DeepStream 9.1 enables developers to build video analytics applications that track objects across multiple camera views using 3D tracking, overcoming limitations of single-camera 2D tracking.
NVIDIA's GB300 NVL72 achieved a world record in pre-training the DeepSeek-V3 671B model using mixture of experts (MoE), reaching 1,648 TFLOPs per GPU and advancing large-scale AI training efficiency.