@Hostkey profile picture
HOSTKEY @Hostkey

Dedicated server leasing and private clouds in the Netherlands, USA, Finland, Iceland and Turkey - premium WEB services provider. GPU servers with NVIDIA cards

Joined November 2010
  • Tweets 1,135
  • Following 69
  • Followers 311
1,011 Photos and videos
HOSTKEY documentation is now available in Dutch (NL). Better accessibility for international teams — faster onboarding, clearer infrastructure workflows. 👉 hostkey.com/documentation/nl…
22
Intel is pushing rack-scale AI infrastructure for inference and agentic workloads. Xeon SambaNova RDUs Foxconn integration → system-level AI design instead of single-server compute. AI infra is becoming rack-native.
69
We launched Managed Apps in HOSTKEY. No VMs. No setup. No infra management. Just pick an app → get a ready service with a URL. Faster deployment, zero infrastructure overhead. 👉 hostkey.com/apps/managed-app…
9
We analyzed real HOSTKEY order data across hundreds of servers. Result: hosting panel choices in production look very different from online reviews. Actual usage > opinions. 👉 youtu.be/7jOqUXXUrqs
17
We built an LLM-based translation agent for technical documentation. It preserves structure, code and terminology while scaling multilingual docs across infrastructure teams. Translation is now an infrastructure problem, not just language. 👉 hostkey.com/blog/161-how-our…
23
I Squared is buying 10 Cogent data centers for $225M to build an AI inference & edge colocation platform. Trend is clear: AI compute is moving closer to users. Latency and geography now matter as much as GPUs.
16
S3-compatible storage is now a standard layer in modern infrastructure. In this guide: how to connect to S3 step by step — credentials, endpoints, tools and real-world usage in AI, backups and cloud systems. 👉 hostkey.com/blog/160-how-to-…
1
15
AI scaling is hitting a new bottleneck — and it’s not GPUs. Marvell highlights a shift: connectivity between accelerators, servers, racks, and data centers now defines performance and cost efficiency. In AI infrastructure, network is becoming the new compute.
1
31
We turned OpenClaw into a fully working AI agent with LLM integration Telegram control. Self-hosted, real-time and fully deployed on HOSTKEY infrastructure. AI agents are becoming infrastructure, not just software. 👉Watch the video: youtube.com/watch?v=tcLxRJR2…
1
1
113
AI datacenters are hitting physical limits. Power, cooling and GPU density per MW are now the main bottlenecks — not compute. AI infrastructure is becoming energy-driven.
32
AI infrastructure is shifting toward inference-first design. NVIDIA cloud providers are optimizing GPU clusters for latency, routing and large-scale model serving — not just training. Compute is becoming orchestration-driven.
1
37
Nextcloud panel in HOSTKEY marketplace updated. Better HTTPS handling, PostgreSQL stability, Redis PHP-FPM tuning, improved Nginx config and stabilized OnlyOffice proxy. Result: faster UI, fewer errors, more stable production behavior. 👉 hostkey.com/apps/storage-sof…
5
540
India was denied access to a supercomputer — so they built their own. A clear example of a bigger trend: compute is becoming strategic infrastructure and restrictions are accelerating local AI/HPC development. Full story: hostkey.com/blog/157-india-w…
242
AI datacenters are hitting new limits: power, cooling and GPU density per MW. Infrastructure is shifting from compute-centric → energy-centric design. Scaling AI now depends on physical constraints, not just GPUs.
116
System prompt vs Skills vs MCP tools — who actually controls AI behavior? We tested OpenWebUI with a live URL validation case. Key insight: each layer solves a different problem — behavior, reuse and execution. Full breakdown: hostkey.com/blog/155-openweb…
69
Google Cloud is turning inference routing into a core part of AI infrastructure. Smarter distribution across GPUs and regions = lower latency, better utilization, more efficient scaling. AI infra is becoming a routing problem, not just compute.
51
S3 Object Storage is the backbone of modern cloud & AI infrastructure. Scalable, API-driven storage used for datasets, backups, ML pipelines & more. Why it became the industry standard — explained here: hostkey.com/blog/154-what-is…
30
Deploying an AI agent shouldn’t feel like DevOps overload. New video: run a ready OpenClaw AI agent on HOSTKEY VPS in minutes — no heavy setup. We cover deployment, CLI and web UI setup. Next part → LLM Telegram automation. Watch: youtube.com/watch?v=WXZpKJT4…
2
110
AWS is optimizing EC2 GPU instances for generative AI inference — improving scaling behavior, utilization and cost efficiency on H100-class systems. Cloud infrastructure is shifting from training-heavy to inference-heavy design.
35
NVIDIA is pushing Blackwell systems toward inference-first optimization — focusing on latency, throughput and cost per token rather than just training performance. AI infrastructure is increasingly defined by serving efficiency, not peak compute.
27
Load more