AI · Article
OpenAI Builds Pando, a Git Storage System for Parallel Work and Pack Reuse
GeekNews ·
At a glance
OpenAI introduced Pando, a new storage backend for Git, after its source control system slowed down as parallel development and CI work grew with Codex. The Git usage method stays the same, but Pando is designed to push multiple branches at the same time and reuse prebuilt pack files (bundles of files compressed into one) from a local cache.
Key points
- Pushes are handled separately for each branch, so unrelated branches do not wait on each other.
- Pack files, which bundle objects together, are reused across many users, reducing load on the server.
- After adoption, the median push delay dropped by about 80%, and it can serve about 1 petabyte per day using less CPU, the company said.
- The switch was made with no downtime by first placing it in front of the existing system as a relay, and one engineer designed it while an agent team built it, taking 10 weeks to deploy.
Why it matters
As development work grows with AI tools, code repositories can be the first to hit their limits. This case shows how that bottleneck was addressed.
Original source
OpenAI의 Git 확장, 병렬 푸시와 팩 재사용을 위한 Pando 구축
GeekNews · Nov 3, 16:34 KST
Read on GeekNews ↗AI summarized the original article. Check the original for accuracy. Copyright belongs to GeekNews.
OpenAI · Ranking and more news
More from this field
- AWS Quick Adds Chat-Built Flows, Updated Web App, and Teamwork Among Agents
AWS Quick added updates that let a team turn one person's work routine into a shared, repeatable process. Flows can now be built and changed entirely in chat, and they can run from the web, desktop, or mobile app. Agents working together can also give more complete answers.
Amazon AWS Machine Learning blog · Nov 21, 10:06 KST - Amazon Bedrock Web Search Adds Five Languages and Country-Based Results
Amazon's Web Search tool for Bedrock, which gives AI agents cited, up-to-date web results, now supports Spanish, French, German, Italian, and Japanese, plus non-US English. A new optional country field can also steer results toward sources from a specific country. Existing customers get the language update automatically, and pricing stays the same in every language.
AWS What's New · Nov 21, 09:17 KST - AWS Adds Shared Cache Skill to Cut Self-Hosted LLM Cost and Latency
AWS wrote about a shared cache that lets self-hosted AI models reuse saved work across servers, aiming to cut cost and wait time. The post shows how a coding agent can help set it up. Amazon says it can cut inference cost by up to 40% and first-token delay by up to 80%.
Amazon AWS Machine Learning blog · Nov 21, 09:14 KST - Can AI Really Find Black Friday TV Deals? A BBC Test Says Not Yet
A BBC team tested four popular AI chatbots to see if they could find a good Black Friday deal on a 55-inch TV. The results were mixed, with many links broken or out of date.
BBC News · Nov 21, 09:06 KST