CHANNEL

Infra

"Infra" is a topic channel on AGI Hunt, an AI news site updated around the clock in real time. Coverage: Compute, GPUs and chips, data centers, inference and serving stacks, training costs, supply chain and compute economics.

Daily roundup: the latest AI News Daily — the past 24 hours across the whole site, per channel and per company · browse the archive

• Burned $5,300 of AI tokens in 30 days on a $300 subscription plan — theteknosaur · 2026-10-26
• Jane Street, Hudson River Trading build data centers and back chip startups for AI — The Decoder · 2026-10-26
• Hamster wheel 'powers' local AI in a hilariously polished livestream-style setup — Sad_Measurement_7502 · 2026-10-26
• $10-12K build for local agentic coding: 1x 5090 vs 4x 3090 vs Mac Studio M5 Ultra — Shadmelor · 2026-10-26
• $1.5M local inference cluster running 240+ kWh/day is the new 'lottery winner' flex — TheZachMueller · 2026-10-26
• Gartner: 80% of neoclouds to cease independent ops by 2030, $4T at write-down risk — rohanpaul_ai · 2026-10-26
• Dev Builds Telco Anomaly Detection Stack, Jokes ASCII Art Is the Last Human Job — l4rz · 2026-10-26(2 related)
• SoftBank plans $50B AI and data center outlay in Malaysia — VectorWireAI · 2026-10-26
• Chunking latency cut from 37.1s to 585ms in another weekly perf improvement — DanielLockyer · 2026-10-26
• Smart TVs spy on you: Samsung TV found with always-on mic and ACR data shared with ~400 partners — FinanceYF5 · 2026-10-26
• AirLLM runs 70B models on a single 4GB GPU with no quantization or pruning — techNmak · 2026-10-26
• AI factories need frontier models plus robotics plus on-site energy to bypass grid queues — ingliguori · 2026-10-26
• Strix Halo benchmark: Swift 1.5 vs Unsloth quants of Qwen3.8-Flash-Next — deepu105 · 2026-10-26
• Reddit user builds $3360 Threadripper box with 256GB DDR4 to run large LLMs locally — sitytitan · 2026-10-26
• Claude API pricing visualized: Batch saves 50%, cache reads down to 0.1x — dthrdr · 2026-10-26
• Top AI spenders put over a third of budgets into raw GPU compute — rohanpaul_ai · 2026-10-26(2 related)
• Running Qwen3.8B across 3 GPUs and 130GB RAM only hits 18 T/s — MageLD · 2026-10-26
• Used RTX 3090s Now Top $1500 on eBay as Owner Hunts a Replacement — like_a_tensor · 2026-10-26
• Ternary 1.58-bit Nemotron 30B-A3B GGUF Runs 368K Context on 12GB GPU — pmttyji · 2026-10-26
• DeepSeek V4.1-Flash hits 49 tok/s decode on dual Radeon R9700 in local test — Longjumping-Elk-7756 · 2026-10-26
• Anthropic built Scion to replace git as AI-driven merges jumped from 2,500 to 30,000/day — bibryam · 2026-10-26
• NVIDIA shows RTX Spark powering sim-to-real robotics: Microduck robot learns somersault in simulation — Aiden_Tech_Ai · 2026-10-26
• MiniMax H3 community adds 8-step reference control, 4-step distill, Apple Silicon support — Hailuo_AI · 2026-10-26
• SPCX acquires Modal, per unconfirmed one-line report — arthurcolle · 2026-10-26
• New PCIe-optimized all-reduce kernels boost DeepSeek decode 21% (32% with spec decode) — teortaxesTex · 2026-10-26
• Guess the era: multimode graded-index glass waveguide optical backplane — jwt0625 · 2026-10-26
• Stanford and UC Berkeley's TokenGate speeds long-context prefill by up to 11.3x — MundanePercentage674 · 2026-10-26
• Three papers chart advanced packaging scaling: CoWoS-R, EMIB-T, UCIe bandwidth density — jwt0625 · 2026-10-26
• RTX 4090 workstation vs AI subscriptions: an editor's break-even math — iamthecarguy1234 · 2026-10-26
• Cerebras CS-4 teardown: 2x bandwidth and power, 50% fewer parts vs CS-3 — jwt0625 · 2026-10-26
• theo doubts Windows ARM: porting 20 years of tech debt vs Intel improving — wateriscoding · 2026-10-26
• TokenSpeed Open-Source Inference Engine Explains Its Polyglot Design — zhyncs42 · 2026-10-26(2 related)
• AWS Explains Why Speculative decoding Never Changes LLM Output — anbarth · 2026-10-26(2 related)
• You Can Train a 1B Model in 3 Weeks: Dev Shares 15.9-Day Training Run Log — arthurcolle · 2026-10-26
• TWIG: weights as shared discrete decision paths unify pruning, quantization, and adder-based circuits — ITMO · 2026-10-26
• BofA sees 21M incremental CPUs by 2030 from agentic AI — analyst calls the math nonsense — AccBalanced · 2026-10-26
• Resin Author Explains Design: Multi-language Split to Boost Coding Agents — ekzhang1 · 2026-10-26(2 related)
• Why NVL72 scale-up bandwidth sits mostly idle — cutting it won't hurt inference — bookwormengr · 2026-10-26
• Custom OpenAI-branded gigabit network switch PCB shown off in photos — mckbrando · 2026-10-26
• No coding agent vendor has shipped DeepSeek's open-sourced DeepInfer yet — pingzawa162 · 2026-10-26
• Open-source recipe: abliterated DeepSeek-V4.1-Flash at ~200 tok/s on 4x RTX PRO 6000 — burny_tech · 2026-10-26
• Micron opens second advanced packaging plant in Houli, a concrete HBM step — tengyanAI · 2026-10-26
• What do you do with the spare 8 GPUs in a 64-card NVL72 setup? Keep them hot as spares — _xjdr · 2026-10-26
• Jalapeño chip is just a systolic array with HBM: why custom inference silicon is easier than you think — ai · 2026-10-26
• PyTorch nested_graph_breaks switch delivers 4.53x speedup on NVIDIA B200 — cfregly · 2026-10-26(2 related)
• Dev gets Qwen3.8-Flash-Next W4A16 tuned and running on a 128GB RTX Spark N1X, open-sources the setup — mycall · 2026-10-26
• Jassy: $350M Annapurna deal is Amazon's best acquisition; chip business at $25B run rate — SumitGup · 2026-10-26
• Anthropic engineer: Claude Code cache hit rates are now high, savings live — teortaxesTex · 2026-10-26
• Six-month bet: hybrid AI deployments, not a wholesale shift to local models — yi111 · 2026-10-26
• Deco decouples 'thinking' from 'spelling' in generative recommenders for 3.2-8.6x throughput — _reachsumit · 2026-10-26