
AWS Lambda adds GPU support for serverless inference
AWS said the feature targets workloads that need GPU acceleration but are often built on self-managed infrastructure.

AWS said the feature targets workloads that need GPU acceleration but are often built on self-managed infrastructure.

Exa's neural search engine is backed by an in-house vector database, so the unit economics of embedding tens of billions of web documents drive the business. This post shares how Exa optimized embedding inference on AWS Neuron hardware (Inferentia2 and Trainium2) with custom NKI kernels, reaching a 47% gain in performance and cost per million documents.

Large training jobs can stall or compute wrong results even when every node looks healthy. Amazon SageMaker HyperPod now adds workload recovery and expanded infrastructure checks that catch these failures sooner and raise the goodput of your training workloads.

The Amazon SageMaker JumpStart Model Leaderboard compares 70+ foundation models side by side on quality, speed, and cost, directly in SageMaker Studio. Learn what the leaderboard shows, how AWS benchmarks models, and how to go from a shortlist to a deployed model.

Airbnb runs evals on real engineering tasks and uses Amazon Bedrock observability to choose models for its coding agents and control production costs. This post shows how cost per task guides model selection, why production spending can differ from eval results, and how to monitor latency, manage quotas, and attribute spend.

Learn how to make RAG chatbot answers traceable with Amazon Aurora PostgreSQL, pgvector, and Amazon Bedrock Guardrails, logging sources, scores, and guardrail traces in SQL.

Amazon shocked everyone in October when it launched its new Alexa Tablet series with Google Play support built-in, addressing the biggest shortcoming of its older Fire Tablets. For Black Friday 2026, the Alexa Tablet 8 is a mere $150 and, with Google Play on board, this feels like a Fire Tablet that’s all grown up.

Amazon Bedrock AgentCore Identity now supports private certificate authorities (CAs), so your agents can connect to self-hosted identity providers that use a private CA. This post shows how to issue an IdP certificate with AWS Private CA, configure a credential provider to trust it, rotate the CA, and protect the certificate.

Learn how Salesforce uses the cross-Region inference router in Amazon SageMaker AI to turn fragmented Regional capacity into reliable model serving. Salesforce adds capacity securely across Regions and accounts while keeping its data residency commitments, and sees per request where each call was served and why.

Social Department, an AI operating system for entertainment marketing, worked with TrackIt and AWS to build QuickSearch: single-action queries that surface shareable scenes, quotable lines, memeable frames, and trend-aligned moments across large video libraries, using TwelveLabs models on Amazon Bedrock and Amazon OpenSearch Serverless.

Many enterprises run a centralized network model: one account owns every VPC and a central firewall inspects all traffic. This post shows a tested pattern for connecting Amazon Bedrock AgentCore Gateway to private APIs and MCP servers in that model, with the VPC Lattice resource gateway in the network account so every hop crosses the firewall.

Stretto replaced its web-form lead capture with Tara, an AI voice agent built on Amazon Nova 2 Sonic that conducts bankruptcy intake interviews by phone and qualifies leads. See how Stretto, the AWS Generative AI Innovation Center, and CI&T designed the voice architecture, validated it for sensitive conversations, and moved it to production.

A recap of the inaugural AWS Quick User Group DC meetup, where data leaders explored getting enterprise data ready for AI agents. It covers what's new in AWS Quick, Financial Independence Group's AI analytics journey, and hands-on app and no-code flow building.
United Launch Alliance says preliminary findings show Vulcan's two BE-4 engines didn't get enough fuel from the rocket during Sunday's failed launch, and that the engines' own performance wasn't the problem. Blue Origin's New Glenn uses the same engines.
The 41mm Wi-Fi Pixel Watch 5 is 25% off, a nice saving on Google's newest smartwatch.

Garmin's rugged 50mm Instinct 3 Solar is 36% off, with up to 40 days of battery and unlimited smartwatch life in enough sun.

Amazon founder and chairman Jeff Bezos says he was “heavily involved” in the casting of Robert Pattinson as Amazon MGM Studios’ next James Bond, though he says there was some hesitancy in the selection. “We were slightly biased against him because he’s already so famous,” he says. “We sort of thought maybe we should have somebody a little more unknown.”
Also at The Guardian World, The Verge

A progressive think tank says nearly 1 in 3 featured holiday deals in 2024 and 2025 offered no real savings. Amazon calls the report deeply flawed.
Early findings point to core-stage underperformance, while solid boosters and Centaur appear to have worked as intended

A new video has emerged showing members of the isolated Awa tribe who live in the remote part of Brazil’s Amazon forest.

Illegal activities in the Amazon pose grave risks to uncontacted Indigenous groups like Brazil’s Awa,.

Jeff Bezos says actor Robert Pattinson won everyone over in “real old-fashioned screen tests,” even though Bezos and the filmmakers were initially wary of how famous Pattinson already is.

Around 20 Indigenous people emerged from isolation in Brazil’s Amazon rainforest, appearing thin and hungry, several organisations reported on Tuesday, raising the alarm about increased invasions into the tribe’s territory. According to the NGO Survival International, members of the Awa people, who live in voluntary isolation, arrived on Monday in a village and made contact with a government team in a “historic encounter”. The village they entered was within the Arariboia…
The administration has been seeking to ban U.S. imports of new models of Chinese optical transceivers over national security concerns.

Investors including AMD, Microsoft, Google, and Nvidia agreed to buy US$26 billion in non-voting shares through a private placement.

Amazon SageMaker HyperPod Inference now supports NVIDIA Dynamo for distributed inference, with multi-node and disaggregated serving, KV-aware routing, and automatic fault recovery. Serve large models across many nodes without building the orchestration layer yourself.

As production agents on Amazon Bedrock AgentCore process multi-megabyte JSON payloads, token costs and latency climb. This post shares three composable strategies: offloading heavy artifacts to Amazon S3, on-demand result paging, and dictionary compaction. Together they cut per-turn token use by up to 71% and latency by up to 80%.

Meta's Muse AI agent now leaves Amazon out of its shopping searches unless a user specifically asks, according to the instructions Muse provides. Amazon began blocking Muse from its store in September.

Amazon will own just under 15% of Anthropic after the IPO, with the option to go much higher. Google is a close second with holdings of nearly 14%.

Ecommerce giant Amazon has announced plans to invest $7 Bn in renewable energy projects across India by 2030. The company…