Search @thenewstack | teapawt

@thenewstack12m

Moving AI pilots into production? Watch Satya Nadella and Judson Althoff open Microsoft Ignite on November 17, then learn from 100-plus organizations already doing it. Thanks to Microsoft thenewstack.io/microsoft-ign…

See what AI looks like in production at Microsoft Ignite

See what AI looks like in production at Microsoft Ignite

Moving AI pilots into production? Watch Satya Nadella and Judson Althoff open Microsoft Ignite on November 17, then learn from 100-plus organizations already doing it.

105

Gain device-level GPU visibility on Amazon EKS, eliminate blind spots, and cut compute waste with CloudWatch accelerated compute insights. Thanks to AWS Marketplace thenewstack.io/cloudwatch-ac…

Expensive accelerators, no workload-level visibility: Introducing Amazon CloudWatch accelerated...

Expensive accelerators, no workload-level visibility: Introducing Amazon CloudWatch accelerated...

Gain device-level GPU visibility on Amazon EKS, eliminate blind spots, and cut compute waste with CloudWatch accelerated compute insights.

178

We explore how Kubernetes is getting new guidance for running AI workloads, including a new book from TNS, a dedicated Kubernetes working group, and advice on running LLMs across multiple GPUs. Thanks to @HPE thenewstack.io/kubernetes-wo…

Kubernetes just formed a workload conformance group. Here’s why it matters for AI teams.

Kubernetes just formed a workload conformance group. Here’s why it matters for AI teams.

Kubernetes can already certify clusters for AI. It still can’t vouch for the workloads.

212

Render just removed its AI sandbox time limit — and Vercel still has a 24-hour cap thenewstack.io/render-vercel…

Render just removed its AI sandbox time limit — and Vercel still has a 24-hour cap

Render just removed its AI sandbox time limit — and Vercel still has a 24-hour cap

Render is challenging Vercel with AI sandboxes that run indefinitely, plus durable workflows and shared storage for long-running agent jobs.

1

1

249

A $60 billion fintech hired a 14-year-old. Its CEO explains why. thenewstack.io/ramp-hires-te…

A $60 billion fintech hired a 14-year-old. Its CEO explains why.

A $60 billion fintech hired a 14-year-old. Its CEO explains why.

Ramp co-CEO Eric Glyman on why the fintech hired a 14-year-old intern — and how it finds engineering talent through side projects, not résumés.

271

Finding and Fixing Excessive Cloud Costs with the New Metering API… and an AI Sherlock Holmes via @ScyllaDB, by @TzachL scylladb.com/2026/11/03/find…

Finding and Fixing Excessive Cloud Costs with the New Metering API… and an AI Sherlock Holmes

Finding and Fixing Excessive Cloud Costs with the New Metering API… and an AI Sherlock Holmes

Let's use a new API and UI to investigate ScyllaDB Cloud usage and discover some cost reductions

1

283

Why Nvidia gave OpenShell to the CNCF thenewstack.io/nvidia-donate…

Why Nvidia gave OpenShell to the CNCF

Why Nvidia gave OpenShell to the CNCF

Nvidia’s Ali Golshan says securing AI agents is an infrastructure job, which is why OpenShell went to the CNCF and its Kubernetes platform teams.

1

291

Meta launches Muse Ember to take on Claude Haiku and GPT-6 Luna thenewstack.io/meta-muse-emb…

Meta launches Muse Ember to take on Claude Haiku and GPT-6 Luna

Meta launches Muse Ember to take on Claude Haiku and GPT-6 Luna

Meta’s Muse Ember is a low-cost multimodal AI model for coding and computer use, with benchmark scores close to Claude Haiku 5.5 and Meta’s Muse Spark 1.3.

351

Google claims its new AI software cuts inference latency by over 90% — but there's a catch thenewstack.io/google-tpu-sy…

Google claims its new AI software cuts inference latency by over 90% — but there's a catch

Google claims its new AI software cuts inference latency by over 90% — but there's a catch

Google’s new open-source TPU Sync aims to speed AI inference by moving KV-cache data between memory tiers, cutting time to first token by up to 90%.

319

Anthropic launches Fable 5.2 with a price cut and stronger coding scores thenewstack.io/anthropic-fab…

Anthropic launches Fable 5.2 with a price cut and stronger coding scores

Anthropic launches Fable 5.2 with a price cut and stronger coding scores

Claude Fable 5.2 cuts token prices by 20% and leads Anthropic’s coding comparisons, while a new Fast mode promises up to 2.5 times faster token generation.

2

383

Anthropic built a model that beats Opus. It recommends Opus anyway. thenewstack.io/anthropic-fab…

Anthropic built a model that beats Opus. It recommends Opus anyway.

Anthropic built a model that beats Opus. It recommends Opus anyway.

Anthropic’s new Fable 5.2 nearly matches Opus 5.5 on key coding benchmarks at half the task cost, yet it still names Opus its pick for most developer work.

1

1

413

“A quick first token can mask slow subsequent generation; a fast complete response can still be inaccurate or unhelpful.” Thanks to @HPE thenewstack.io/ai-inference-…

Busy GPUs can hide slow AI inference on Kubernetes

Busy GPUs can hide slow AI inference on Kubernetes

Why a GPU running at 100% might be nothing to celebrate

1

1

394

Developers have been hacking around MCP’s file-upload problem for two years. A fix is taking shape. thenewstack.io/mcp-file-uplo…

Developers have been hacking around MCP’s file-upload problem for two years. A fix is taking shape.

Developers have been hacking around MCP’s file-upload problem for two years. A fix is taking shape.

MCP began as a way to help developers connect models with tools. But it never had a standard way to move files - a new draft extension aims to close that gap.

2

450

@thenewstack10h

AWS had its agents build a free and open emulator for AWS. They now maintain it, too. thenewstack.io/aws-emulator-…

AWS had its agents build a free and open emulator for AWS. They now maintain it, too.

AWS had its agents build a free and open emulator for AWS. They now maintain it, too.

AWS Emulator lets developers and coding agents test against AWS locally. No human wrote the code, and AWS won’t take pull requests for it.

2

407

@thenewstack11h

Anthropic’s Claude Fable 5.2 built a shooter in 1 prompt — and it only took 8 hours and $550 thenewstack.io/claude-fable-…

Anthropic’s Claude Fable 5.2 built a shooter in 1 prompt — and it only took 8 hours and $550

Anthropic’s Claude Fable 5.2 built a shooter in 1 prompt — and it only took 8 hours and $550

Claude Fable 5.2’s viral demos build games from a single prompt — but behind them sit hours of compute, steep token bills, and real work left undone.

1

2

550

@thenewstack11h

Advanced Java Deployments with Octopus via @OctopusDeploy octopus.com/blog/java-deploy…

Advanced Java Deployments with Octopus

Advanced Java Deployments with Octopus

Learn how to deploy Java applications using Octopus Deploy

385

@thenewstack13h

Vector Database Use Cases: Search, RAG, and AI Apps from @couchbase, by @hannahblaurel couchbase.com/blog/vector-da…

Vector Database Use Cases: Search, RAG, and AI Apps

Vector Database Use Cases: Search, RAG, and AI Apps

The most common vector database use cases are semantic search, RAG, recommendations, and multimodal search. Learn how each works and when to use them.

2

390

@thenewstack14h

Who tests the code your AI agents write? Catch the on-demand session from The New Stack and @datadoghq on goal based testing, testing non deterministic AI features, and the metrics that matter when coverage is no longer the bottleneck. Watch now: bit.ly/4yM3tGt

Image from The New Stack (@thenewstack)

349

@thenewstack14h

It's almost time! Join The New Stack and @Dynatrace at the Museum of Illusions in SLC for a #KubeCon party from 6-8pm. Food, drinks, good times, and of course, brain bending ILLUSIONS! eventbrite.com/e/night-at-th…

Image from The New Stack (@thenewstack)

389

@thenewstack15h

OpenShell on Kubernetes: A guide for platform engineers from @CloudNativeFdn cncf.io/blog/2026/11/04/open…

1

422