// AI news · 2026-11-09 · Google Cloud AI/ML
GKE Dynamic Slicing Now Generally Available
Google Cloud has launched GKE dynamic slicing, decoupling logical TPU slice formation from physical nodepool provisioning.
Published 2026-11-09 21:00 (Beijing time)Source Google Cloud AI/MLAuto-summarised, fact-checked against source
01Summary
It enables on-demand creation of larger or smaller slices without manual nodepool deletion. Super-slicing combines multiple TPU cubes via OCS to form large topologies; sub-slicing partitions a single cube into isolated smaller slices. This reduces job startup time by up to 5x and recovery time by 4.5x, improves fleet utilization, and enhances fault tolerance.
Summarised from the original article by an open-source model, then checked automatically: every number and English name must appear in the source. The original article is authoritative.
02Key points
- Decouples logical slice formation from physical nodepool provisioning.
- Reduces job startup time by up to 5x and recovery by 4.5x.
- Supports super-slicing and sub-slicing for flexible topology creation.
- Uses OCS to dynamically stitch TPU cubes, enabling real-time slice assembly and fault isolation.
03Source
Now generally available: GKE dynamic slicing for higher resilience, and efficiency for TPUs
Google Cloud AI/ML · published 2026-11-09 21:00 Beijing time
05Related news
- Same source · Legacy Software Hinders Enterprise AI Adoption2026-11-10 · Google Cloud AI/ML
- Same source · Google Launches Gemini for Healthcare2026-11-10 · Google Cloud AI/ML
- Same topic · Google Releases TPU Sync for Fast Model State Transfer2026-11-09 · Google Cloud
- Same topic · Google Launches Cloud Pathways ML Runtime at GA2026-11-09 · Google Cloud
- Same day · DJI Osmo Pocket 4P Firmware Update Adds 18x Zoom and More2026-11-09 · IT之家
From AIMap AI news: collected from 1,100+ official and media sources. Summaries only restate the original article; follow the link for the full story. All stories of 2026-11-09 · Archive by date · All AI news · Today