Cloud & Infra

Infrastructure that ships product. Cloud platforms, serverless, containers, databases, and deployment — the releases and architecture shifts that change how and where your code runs.

Article 3d ago 3

HCP Gets SCIM Provisioning, With Some Sharp Edges

SAML-only, disabled default roles, and group-name collisions: what to check before you flip the switch.

Ji-ho Choi

$Hetzner Kubernetes Math Still Works After the Price Hikes$

Hetzner Kubernetes Math Still Works After the Price Hikes

A six-node HA k3s cluster now runs about EUR 57 a month, still less than an empty EKS control plane.

Article · 1mo ago5

Linux Outsources the $ORIGIN Problem to BPF

A 21-patch binfmt_misc series makes interpreter selection programmable, finally giving Nix-style relocatable binaries a real path.

Article · 1mo ago0

Eminent Domain Is the AI Buildout's Next Bottleneck

Landowner fights over what counts as 'public use' now shape where, and how fast, new cloud capacity gets built.

Article · 1mo ago1

Database branches live in the storage layer

Copy-on-write under the block device turns full clones into cheap point-in-time forks.

Article · 1mo ago7

Making hundreds of Postgres nodes look like one

Sharding solves capacity. The proxy is what keeps application code from drowning in it.

Article · 1mo ago7

Telegram's Sticky Home-DC Architecture

A multi-DC messaging platform that prioritizes legal distribution and user affinity over pure geo-latency.

Article · 1mo ago1

Telegram Serverless Cuts the Bot Hosting Tax

Co-located V8 isolates and a built-in SQLite store rewrite how Telegram bots get backends.

Article · 1mo ago1

New York's Data Center Freeze Is a Bigger Bark Than Bite

The first statewide moratorium on hyperscale data centers matters politically more than it matters to your GPU capacity plans, at least for now.

News · 1mo ago1

x86 ACE Specs Matrix AI, Silicon Still Far Off

A joint AMD-Intel outer-product extension freezes the ISA contract, but real cores are years away.

Article · 1mo ago1

CUDA Without NVIDIA: What Actually Works

Drop-in translators, HIP, and SYCL solve different problems. Choose by binaries versus long-term portability.

Article · 1mo ago0

VRAM Beats TOPS for 2026 Local AI GPUs

Memory capacity and bandwidth decide what models you can run; software maturity still decides how painful it feels.

Article · 1mo ago2

Discarded Teslas Still Deliver Local AI VRAM

Enterprise cards at $60–200 remain usable for intermittent inference and training if you tolerate older stacks and power draw.

Article · 1mo ago0

AWS MCP Plugin Puts Agents Under CloudTrail

Official skills, sandboxed Python, and IAM-scoped calls turn coding agents into auditable cloud operators.

Article · 1mo ago0

ZeroFS and S3 Files: same mount, opposite bets

POSIX over object storage is not one idea. The bucket layout decides what you can do next.

Article · 1mo ago0

Scaling PgBouncer Beyond the Single-Threaded Limit

How to run a multi-process PgBouncer fleet using SO_REUSEPORT and peering without breaking query cancellations.

Article · 1mo ago0