Tech
How we built DeepL’s next-generation LLMs with FP8 for training and inference
Discover how DeepL harnessed FP8 for training and inference in next-gen LLMs, boosting throughput and model quality. Learn about our journey with NVIDIA's technology, achieving faster training and superior translations while maintaining low latency.
HTTP/2, load balancers and the ambiguous ":scheme" header
Learn how a change in the ":scheme" header during an HTTP/2 load balancer switch caused 502 errors in DeepL's web app. This post explores DeepL's troubleshooting journey, HTTP/2 specs, and key insights for developers on load balancer configurations.
Design systems: definitely not just for designers
Unlock the true potential of your organization with a robust design system. Discover how design tokens and reusable components streamline collaboration, enhance consistency and boost productivity across teams, enabling efficient product development.
A deep dive into MCPs, Part 1: What is MCP, anyway?
Explore Model Context Protocol (MCP) and its impact on AI since its 2024 launch. Learn how MCP enables AI agents to access real-world tools, and get step-by-step guidance on how to build your own MCP server in just 10 lines of code.
How do you deploy an NVIDIA DGX SuperPOD with DGX GB200 systems?
DeepL’s Chief Scientist Stefan Mesken explores the configuration and capabilities of DeepL’s new NVIDIA DGX SuperPOD with DGX GB200 systems, and what these mean for model architecture and synthetic data.