How we achieve 99.99% uptime with a 7-node cluster
Deep dive into our redundancy architecture, automatic failover, and the custom health-check system that keeps everything running.
Engineering insights, product updates, and the future of independent cloud infrastructure.
Big tech clouds extract $87/TB in egress fees and charge $73/month for a Kubernetes control plane. We think that's absurd. Here's the technical and philosophical case for building an alternative.
Deep dive into our redundancy architecture, automatic failover, and the custom health-check system that keeps everything running.
Benchmarking 4-bit quantized Llama 3 8B on consumer GPUs vs A100s. Spoiler: for most inference workloads, the 4090 wins on $/token.
The biggest CLI release yet. Create K8s clusters, manage encrypted secrets, and deploy Helm charts — all from your terminal.
Our approach to edge caching across 200+ PoPs that doesn't charge customers for data transfer. The math behind free egress.
Step-by-step guide to deploying SDXL for production image generation. Includes auto-scaling, queue management, and cost optimization.
Deploy a full multiplayer backend with matchmaking, chat, and leaderboards using MoveWeight + Nakama. Complete with code examples.
Our multi-layer mitigation stack handles volumetric, protocol, and application-layer attacks. All included at no extra cost.
We built an interactive calculator that shows exactly what you'd pay on MoveWeight vs AWS/GCP/Azure. The results are eye-opening.
Engineering deep-dives and product updates. No spam, unsubscribe anytime.