News - IPv4, Telecom, Hosting, AI - Page 6
GLM-5.2 - The New Leading Open Weights LLM
Z.ai's GLM-5.2 is the new leading open weights model on the Artificial Analysis Intelligence Index, scoring 51 under an MIT license with a 1M token context.
Snapshot Before Risky Changes, Roll Back Instantly
Create an on-demand snapshot of your DCXV cloud server before any risky change, then roll back in seconds. Add a snapshot in the control panel with one click.
Install Any OS - Boot Your Cloud VM From Your Own ISO
Upload a bootable ISO from any HTTPS URL and boot your DCXV cloud VM from it - install any operating system or run a rescue disk, straight from the control panel.
Cloud Server for Ollama in Europe: Self-Host AI EU Guide
Run Ollama on a GDPR-compliant EU cloud server. Covers model selection, GPU setup, API configuration, and performance benchmarks for self-hosted AI in Europe.
Cloud Server for AI Inference in Europe: GPU & CPU Guide
Run AI inference on a GDPR-compliant EU cloud server. Covers GPU vs CPU tradeoffs, hardware specs, model serving setup, and throughput benchmarks for Europe.
Cloud Server for Elasticsearch in Europe: EU Search Hosting
Run Elasticsearch on a GDPR-compliant EU cloud server. Covers heap sizing, shard strategy, index tuning, and search benchmarks for European deployments.
Cloud Server for Redis in Europe: Low-Latency EU Setup
Run Redis on a GDPR-compliant EU cloud server. Covers memory sizing, persistence modes, Sentinel setup, and latency benchmarks for European deployments.
Cloud Server for MongoDB in Europe: Replica Set Guide
Run MongoDB on a GDPR-compliant EU cloud server. Learn WiredTiger tuning, replica set setup, recommended specs, and performance benchmarks for Europe.
Cloud Server for MySQL in Europe: InnoDB Tuning Guide
Host MySQL on a GDPR-compliant EU cloud server. Covers InnoDB tuning, replication setup, recommended specs, and performance benchmarks for European deployments.
DeepSeek V4: 1.6T MoE Model with 1M Context on EU Server
DeepSeek V4 launches Pro (1.6T) and Flash (284B) MoE models with 1M token context, hybrid attention architecture, and three reasoning modes for EU self-hosting.