Как я собрал LLM-печку на 4 GPU, и на что она способна
Всем привет, недавно я собрал и сейчас активно использую ПК для локального запуска хоть сколько-то серьезных LLM, и захотелось поделиться этим опытом …
Latest News
⚑ Report a ProblemTech news from the best sources
All topics
- игры
AI
Gear
News
Tech
agents
ai
api
architecture
automation
beginners
career
claude
devchallenge
devops
javascript
llm
machinelearning
mcp
opensource
performance
productivity
programming
python
react
security
showdev
tutorial
typescript
webdev
Qwen Is Not Yet Ready to Power Local OpenClaw Deployments
Three weeks ago I ran a model showdown — twelve tasks, five models, one RTX 5090 — and Qwen3.5-35B-A3B won. 85.3 weighted score, 206 tok/s, fits in VR…
mTLS на практике: от модели угроз до реальных атак на сервер
Это вторая часть цикла про mTLS. В первой разобрали теорию: как работает рукопожатие, от каких атак защищает и где принципиально бессилен. Здесь — пра…
I built a self-hosted Linux fleet manager with no database and zero pip dependencies
I manage a small fleet of Linux servers and got tired of the options: Ansible is great but not a dashboard. Grafana + Prometheus is powerful but heavy…
How to enhance your network with transparent Wake-on-LAN capabilities
This is the story of how I got frustrated with the energy efficiency of my Homelab network and decided to create a solution that gradually became a so…
Stop Paying for Cloud VMs — Self-Host Your Dev Stack on Refurbished Enterprise Hardware
Cloud computing is brilliant. I'm not here to argue otherwise. But there's a specific use case where it consistently costs more than it should, and mo…
Ceph on Ubuntu (Single Node): Setup Guide (OSD, iSCSI, Bonding)
I recently put together a small Ceph lab on Ubuntu using cephadm, mainly to understand how everything fits together in a real environment. Most guides…
Tailscale Subnet Routers: Accessing Your LAN Without the VPN Headache
I spent three hours trying to SSH into a legacy industrial gateway from a coffee shop, only to realize I'd forgotten to install the Tailscale agent on…
Model Showdown Round 3: Ditching Ollama in Favor of llama.cpp
In Round 1 , we ran five local models and two cloud models through a single coding task. The local models held their own. In Round 2 , we added Gemma …
Чему меня научили два месяца с легковесным локальным AI-агентом
Два месяца назад openLight был маленьким pet project’ом: Raspberry Pi, Telegram-бот, SQLite и несколько команд для управления сервисами. Мне просто хо…
GPU PCI Address Instability: When Your Card Moves Between Reboots
I spent an entire afternoon debugging a VM that refused to boot, only to find out my GPU had decided to change its PCI address. One reboot and the dev…
I Thought Portfolios Weren't for People Like Me. So I Built a Wiki Instead.
For a long time I operated under a quiet assumption: portfolios are for developers. They're for people who build things — apps, tools, open source pro…