Vultr Kubernetes Engine adds GPU node support, lowering AI inference deployment barrier
Vultr Kubernetes Engine (VKE) now supports GPU nodes, letting users deploy NVIDIA GPU worker nodes in VKE clusters with hourly billing.
Track global cloud vendor product launches, pricing changes, infrastructure expansion, AI platform updates, and service changes.
Vultr Kubernetes Engine (VKE) now supports GPU nodes, letting users deploy NVIDIA GPU worker nodes in VKE clusters with hourly billing.
Tencent Cloud launches EdgeOne 3.0, integrating edge function computing, DDoS protection, and intelligent acceleration into a unified platform.
Oracle Cloud Infrastructure launches NVIDIA H200-based bare-metal instances for high-performance AI training and inference.
Akamai's Linode cloud platform expands its compute instance lineup with new AMD Milan and ARM-based options for cost-conscious users.
Huawei Cloud launches AI-native CCE cluster optimized for GPU training and inference, reducing job queue time by 50% and boosting GPU utilization above 85%.
German cloud provider Hetzner expands its cloud lineup with new GPU server options targeting European AI training and inference needs.
Google Cloud announces next-generation Titanium accelerator that offloads network and storage virtualization to dedicated hardware for consistent instance performance.
Google Cloud Spanner introduces single-region instance configuration, reducing costs by ~60% compared to multi-region setups.
DigitalOcean launches GPU Droplets offering NVIDIA H100 GPUs with simplified deployment for AI developers and startup teams.
Microsoft Azure launches VM series based on its custom Cobalt 100 ARM processors, delivering up to 40% better price-performance for web servers and container workloads.
AWS launches R8g memory-optimized instances based on Graviton4 processors, delivering 30% better price-performance versus previous generation R7g.
Alibaba Cloud upgrades its global acceleration service, adding edge nodes in Dubai, Riyadh, São Paulo and other markets, reducing latency by 25% on average.