#Autoscaling
5 articles with this tag
Netflix's Autoscaler Evolution
Netflix is consolidating its two Apache Flink autoscalers into one open-source solution, achieving significant cost savings and improved stability.
Databricks Scales Real-Time Fraud Detection
Databricks details how its Lakehouse architecture, combining Model Serving and Lakebase autoscaling, achieves sub-50ms transaction fraud scoring.
Databricks AI Serving Adapts to Any Model
Databricks unveils an AI serving platform that dynamically adapts to any model and traffic, slashing costs and boosting performance.
Databricks Tackles LLM Inference Costs
Databricks details its 'model units' abstraction and cost-aware autoscaling for reliable, high-throughput LLM inference, cutting GPU costs by over 80%.
Databricks Lakehouse Adds Autoscaling
Databricks Lakehouse autoscaling eliminates provisioning headaches, dynamically adjusting compute for performance and cost savings, and can scale to zero.