Field notes

Blog — Page 26

Engineering notes, case studies, and lessons from the field.

ByteDance Is Building Its Own CPUs — Arm and RISC-V Both on the Table
Hardware

ByteDance Is Building Its Own CPUs — Arm and RISC-V Both on the Table

ByteDance has launched an internal program to design custom CPUs for its AI data centers as x86 prices surge and delivery windows stretch to six months. The company joins Apple, Google, Amazon, and Microsoft in abandoning off-the-shelf silicon.

May 29, 20265 min read
Anthropic Releases Claude Opus 4.8 With Parallel-Subagent Workflows
AI Models

Anthropic Releases Claude Opus 4.8 With Parallel-Subagent Workflows

Anthropic shipped Claude Opus 4.8 on May 28, pushing SWE-bench Verified to 88.6% and adding Dynamic Workflows — a Claude Code preview that runs hundreds of parallel subagents in one session. Pricing is unchanged from Opus 4.7.

May 28, 20265 min read
Nextcloud Releases 33.0.4 and 32.0.10 Maintenance Updates
Open Source

Nextcloud Releases 33.0.4 and 32.0.10 Maintenance Updates

Nextcloud shipped simultaneous maintenance patches for both its current and previous stable branches on May 28. The dual release fixes a file-rename regression in 33.x and continues the project's 12-month supported-branch overlap policy.

May 28, 20265 min read
Carnival Corporation Confirms ShinyHunters Stole 6 Million Customer Records
Cybersecurity

Carnival Corporation Confirms ShinyHunters Stole 6 Million Customer Records

The world's largest cruise operator began notifying nearly 6 million customers on May 27 after a social engineering attack in April gave ShinyHunters access to loyalty program data. The company did not pay the ransom.

May 28, 20265 min read
CVE-2026-48710: 'BadHost' Auth Bypass in Starlette Exposes Millions of AI APIs
Cybersecurity

CVE-2026-48710: 'BadHost' Auth Bypass in Starlette Exposes Millions of AI APIs

A Host-header injection bug in Starlette silently bypasses path-based authentication middleware in FastAPI, vLLM, LiteLLM, and MCP servers. Patch to Starlette 1.0.1 immediately — disclosure gave operators less than 24 hours of lead time.

May 28, 20265 min read
Meta Launches Paid Subscriptions Across Facebook, Instagram, and WhatsApp
Big Tech

Meta Launches Paid Subscriptions Across Facebook, Instagram, and WhatsApp

Meta officially launched 'Meta One' tiered subscriptions on May 27, charging up to $19.99/month for premium AI compute across its apps. It's the company's first direct consumer revenue play since advertising made it a trillion-dollar platform.

May 28, 20265 min read
Pope Leo XIV Releases 'Magnifica Humanitas' — 42,300-Word Encyclical Calls on Governments to Disarm AI
Policy & Regulation

Pope Leo XIV Releases 'Magnifica Humanitas' — 42,300-Word Encyclical Calls on Governments to Disarm AI

Pope Leo XIV published his first encyclical dedicated entirely to AI ethics and governance, a 42,300-word document warning that artificial intelligence risks concentrating power, distorting truth, and enabling warfare. The encyclical was presented at the Vatican alongside Anthropic co-founder Christopher Olah.

May 27, 20265 min read
Mesa 26.1.0 Released: VirtIO-GPU Intel Support, VirGL Declared Unmaintained, Broader Vulkan Coverage
Open Source

Mesa 26.1.0 Released: VirtIO-GPU Intel Support, VirGL Declared Unmaintained, Broader Vulkan Coverage

Mesa 26.1.0 ships as the new feature release of the open-source Linux graphics stack, introducing native-context VirtIO-GPU support for Intel drivers, OpenGL ES 2.0 on PowerVR via Zink, and a slate of new Vulkan extensions. VirGL — long the main GPU acceleration option in QEMU/KVM — is officially declared unmaintained.

May 27, 20265 min read
AMD EPYC Venice (Zen 6, 256 Cores) Enters Volume Production on TSMC 2nm — First HPC CPU at This Node
Hardware

AMD EPYC Venice (Zen 6, 256 Cores) Enters Volume Production on TSMC 2nm — First HPC CPU at This Node

AMD's 6th-gen EPYC server processor is now in volume production at TSMC's 2nm fab, making it the first high-performance computing CPU at this node. Venice claims over 70% performance-per-watt improvement and 30% more thread density versus the current EPYC Turin generation.

May 27, 20265 min read

Page 26 of 61 · 549 articles