groundy

infrastructure & runtime

107 articles·rss

Top in infrastructure & runtime

infra

V8 Isolates vs MicroVMs vs Wasm: Where Spectre Still Draws the Line

Cloudflare's 2026 Spectre revisit is unverified. This guide maps V8 isolates, Firecracker microVMs, and Wasm sandboxes to define a data-classification rule for edge secrets.

aug 21
infra

Fine-Tuning DeepSeek Without NVIDIA: What the Ascend SuperPOD Run Shows

SLAI T-Rex reports 34.22% MFU for full-parameter DeepSeek-V4 post-training on Ascend. The single-cluster result narrows the CUDA moat for fine-tuning, but cost data remains.

aug 20
infra

Cloudflare WebMCP: The Security Baseline for Agent-Ready Sites

Cloudflare's WebMCP proposal is unverified, but the MCP ecosystem is not. With 91.8% of servers unauthenticated, treat agent endpoints as strict APIs, not crawler extensions.

infra

Cloudflare AI Search vs Self-Hosted RAG: Where the Build-vs-Buy Line Lands

Cloudflare's AI Search lacks verifiable pricing and access control docs. For Postgres 13+ teams, pgvector offers a documented, self-hosted alternative that avoids vendor.

infra

Cloudflare H1 2026 DDoS Report: DNS Floods and Sizing Past 1 Tbps

Cloudflare's H1 2026 DDoS report claims DNS floods exceeded 1 Tbps. This article verifies the structural risks of UDP amplification and anycast absorption, while noting the.

infra

daVinci-kernel shifts the RL kernel bottleneck from reward shaping to skill libraries

The daVinci-kernel preprint argues RL-tuned CUDA kernels plateau because skill libraries are mis-architected, not because rewards are mis-shaped. It co-evolves skill.

infra

Calibrated LLM Monitoring: Conformal Prediction with Drift Detection

A cloud-native architecture packages conformal prediction as a sub-2ms microservice to replace raw logprob confidence with finite-sample coverage guarantees, gated by.

infra

Cloudflare AI Crawler Controls: Block, Charge, or Allow Bots Per Route

Cloudflare ships per-route AI crawler controls for Search, Agent, and Training bots. Site owners must model pay-per-crawl yield against ad inventory and measure Agent.

more in this beat1–50 of 99 · page 1 / 2
  1. jul 25infraPostgres LISTEN/NOTIFY Scales: When to Drop Redis for Job Fan-Out
  2. jul 24infraTailscale on Azure: Measure Direct vs DERP Routing to Control Latency and Egress
  3. jul 24infraAccelerate vs Megatron Core: The Model Size Curve for Distributed Training
  4. jul 23infraWhy cgroups, not permission prompts, bound AI agent CPU and memory
  5. jul 21infravLLM Configs Shift Energy, Latency, and Accuracy: A 9,000-Run Study
  6. jul 19infraAWS Estimated Billing Was Off by $1.7B: Reconciling Actual Cloud Spend
  7. jul 19infraCloudflare Attribution vs Custom Logs: The Per-Path AI Crawler Decision
  8. jul 18infraSpectral Compute CUDA Translation: vLLM Procurement vs Porting Cost
  9. jul 18infraRunning MiniCPM-V-4.6 on Fermi: What 6 GB of VRAM Forces
  10. jul 17infraLLM Inference Without a GPU: Pure CPU vs Hybrid CPU-GPU Scheduling
  11. jul 17infraRL-Found CUDA Kernels Beat cuBLAS: Kernel Tuning Shifts to Reward Design
  12. jul 15infrapgvector vs Pinecone vs Qdrant: Picking a Vector Database in 2026
  13. jul 14infraOllama vs LM Studio: Picking a Local LLM Runtime in 2026
  14. jul 14infraBeyond Quantization: LLM Efficiency Is Now a Memory-Bandwidth Problem
  15. jul 13infraHow Sparse LLM Weights Cut GPU Inference Cost Without Quantization
  16. jul 11infraGLM-5.2: vLLM Int4 Drops MTP Without Patches, SGLang FP8/NVFP4 Keeps It
  17. jul 11infraServing DeepSeek on Azure: Compliance Without Owning the GPU Fleet
  18. jul 10infraVercel Makes WAF Mitigated Traffic Free: Recompute Your Edge Cost Model
  19. jul 10infraGLM 5.2 Hosting Compared: Vercel AI Gateway vs Self-Hosted vLLM
  20. jul 10infraCloudflare DMARC Management GA: What to Configure Before p=reject
  21. jul 09infraClaude Code Permissions vs OS Privilege Isolation: What the Gap Costs
  22. jul 09infraLLM Memory Without the RAM: What SSD-Backed Paging Actually Costs
  23. jul 09infraTriton Kernels Pass Tests but Run Slow: The GPU Kernel Eval Gap
  24. jul 09infraVercel Edge Config: What Global Feature Flags Actually Cost at the Edge
  25. jul 09infraServerless GPU Inference on GCP: What the Cold Starts Actually Cost
  26. jul 09infraVercel In-Function Concurrency: What It Changes for Stateful Node.js
  27. jul 09infraRunning LLMs on AMD GPUs With ROCm: What Actually Works
  28. jul 08infraCloudflare Meerkat: What Globally Distributed Consensus Costs at the Edge
  29. jul 08infraVercel CDN Now Honors External Origin Cache-Control: Audit Your Headers
  30. jul 08infraAI Found Real Bugs in Cloudflare's Circl Crypto Library
  31. jul 08infraPruning RAG Context: What to Cut Before the LLM Sees It
  32. jul 08infraCloudflare's x402 Gateway: What Per-Request API Billing Actually Needs
  33. jun 30infraDoubao 2.1 Pro: What 180 Trillion Daily Tokens Means for Inference Infrastructure
  34. jun 30infraEvery CUDA Kernel Pays a Launch Tax: The Host-to-Device Walkthrough
  35. jun 29infraVercel Montreal Region: Audit Residency Before You Migrate
  36. jun 29infraGLM-5.2 on vLLM and Ascend: Open Weights Beyond NVIDIA
  37. jun 28infraHow Vercel Runs Its Own CDN in Front of Discourse: A Self-Dogfooding Case Study
  38. jun 28infraVercel Runtime Logs Surface CDN Cache Hits, Not the Eviction Cause
  39. jun 28infraMultimodal Knowledge Graph RAG vs Vector RAG: What MKG-RAG-Bench Shows
  40. jun 28infraVercel Observability Now Tracks Redirects and Rewrites Beside Function Errors
  41. jun 28infraCloudflare Workflows Saga Rollbacks: Compensating Actions in Serverless Orchestration
  42. jun 27infraStatic Corpus RAG: The Bible Case for Separating Churn from Algorithm Complexity
  43. jun 27infraVercel's KIKO Milano Black Friday Case Study: What the Scaling Claims Skip
  44. jun 27infraVercel Postgres vs Neon vs Supabase: When the Bundled DB Wins
  45. jun 27infraFine-Tuning a 20B LLM With RLHF on a 24GB GPU: What Fits
  46. jun 27infraVercel Flat Rate CDN Beta: Break-Even Math for Spiky Workloads, Tax for the Rest
  47. jun 26infraWhere DeepSeek Weights Actually Run on Vercel's AI Gateway
  48. jun 26infraVercel's Anti-Lock-In Pitch: What the Open-Source Bet Still Locks In
  49. jun 26infraVercel Adds Tag-Based CDN Cache Invalidation: Surrogate Keys at the Edge
  50. jun 26infraGLM 5.2 Fast on Vercel AI Gateway: What Routing Through Wafer Actually Buys
about this beateditorial framing

Production AI runs on infrastructure that was never designed for it. Inference serving is a moving target as prefill and decode pull apart onto different hardware, KV caches spill into tiered storage, and collective communication libraries get rewritten to claw back bandwidth. Every benchmark win on synthetic workloads has to survive long-context synthesis, multi-tenant interference, and the unglamorous math of tokens-per-dollar before it counts.

The fabric underneath is just as contested. Vector databases are converging with the OLTP stack, serverless runtimes are quietly absorbing what connection poolers used to own, and overlay networks keep colliding with cloud-provider NAT and egress policy in ways that turn architecture diagrams into invoices. Storage density is outrunning rebuild windows, forcing erasure-coding choices that used to be theoretical. Cheaper-inference research keeps threatening the assumption that scale must mean GPU farms, while denser GPU farms keep proving it.

This beat covers that tension on the merits. We track serving architectures, networking and peering economics, retrieval and caching layers, GPU and storage hardware, and the cloud-account dependencies that quietly underwrite the whole stack. We compare vendor claims against published numbers, flag when a throughput headline hides a quality regression, and pay attention to the boring failure modes that take down platforms more often than the exciting ones do.