// AI news · 2026-11-07 · Google Cloud
Google Launches Storage-Optimized Z4M Instances for AI/ML Workloads
Google launches Z4M storage-optimized instances with up to 168,000 GiB local SSD and 400 Gbps RDMA networking.
Published 2026-11-07 01:00 (Beijing time)Source Google CloudAuto-summarised, fact-checked against source
01Summary
Designed for AI/ML workloads including checkpointing and distributed KV caches, Z4M delivers up to 70% higher IOPS and 110% greater throughput than Z3. Enhanced with Titanium SSDs and fifth-gen Intel Xeon processors, it supports GPUDirect and TPUDirect Storage for low-latency GPU/TPU utilization.
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
- Up to 168,000 GiB local SSD storage and 400 Gbps RDMA networking.
- Up to 70% higher IOPS and 110% greater throughput than Z3 series.
- Supports GPUDirect and TPUDirect Storage for low-latency GPU/TPU access.
- Bare-metal instances offer direct hardware access for specialized workloads.
03Source
Introducing Z4M: storage-optimized compute for AI/ML and parallel file systems
Google Cloud · published 2026-11-07 01:00 Beijing time
05Related news
- Same source · Pandas Adds BigQuery Accessor for Zero-Shot Predictions2026-11-06 · Google Cloud
- Same source · Google Launches AI.AGG Semantic Aggregation in BigQuery2026-11-06 · Google Cloud
- Same topic · Google analyzes agent distraction from code TODOs2026-11-06 · Google AI Changelog
- Same topic · Google launches Gemini Omni 1.2 Flash video model with conversational editing2026-11-06 · cnBeta
- Same day · Govern AI Agent Access with Okta and AgentCore2026-11-07 · AWS Machine Learning Blog
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 2026-11-07 · Archive by date · All AI news · Today