engblogs

Introducing A4 VMs powered by NVIDIA B200 GPUs for modern AI workloads in a Google Cloud preview

Enhancing PyTorch model performance with PyTorch/XLA 2.6 through experimental scan operator, host offloading, and C++11 ABI improvements

Optimize digital pathology workflows with H-optimus-0 on AWS for faster, accurate diagnoses and AI-driven insights

Developing an FM-based email classifier using prompt engineering with Amazon Bedrock for Travelers Insurance

Deploy DeepSeek R1 with Mosaic AI Model Serving on Databricks and boost efficiency with distilled reasoning models.

Deploy DeepSeek-R1 model through Amazon Bedrock Marketplace and Amazon SageMaker JumpStart to enhance reasoning capabilities and scale generative AI ideas on AWS

Maximize ROI by optimizing your Databricks costs with pre-built AI/BI tools and best practices

Exploring the impact of U.S. National Laboratories on enhancing America's AI leadership

Addressing hallucinated translations in large language models through hallucination-focused preference optimization during model training phase.

DeepSeek's efficiency breakthrough revolutionizes AI compute demands, challenging industry giants and democratizing access to advanced hardware.

Simplify Kubernetes developer experience with KRO, a Kubernetes-native framework for defining groupings of resources for end users.

Developing a high-quality compact neural TTS system with low latency and low disk footprint for accessibility applications.