TDK — leaderboard ()

Infortrend launches EonStor GS 5024U U.2 NVMe storage for AI and HPC workloads

Infortrend has announced EonStor GS 5024U, a U.2 NVMe SSD-based unified storage system that it says is its highest-performing storage platform to date. The company positions the system for AI workloads including model training, inference, and real-time analytics, where high throughput and low-latency access can become a bottleneck.

Infortrend says GS 5024U is powered by an Intel Xeon 6 processor and delivers up to 125 GB/s throughput and 2.4 million IOPS. The company reports this is 2.5 times higher throughput than its previous highest-performing model. The platform adds PCI Express 5.0 solid-state drive support and integrates 200 Gb Ethernet and Non-Volatile Memory Express over Fabrics (NVMe-oF) networking to target ultra-fast data transfers.

For GPU-centric pipelines, Infortrend says GS 5024U supports GPUDirect Storage to help maximize GPU utilization. The system also supports parallel file systems including Lustre, which Infortrend says can enable hundreds of gigabytes of throughput for AI workloads.

On availability and resilience, Infortrend says GS 5024U includes dual redundant controllers to keep operations running during intensive AI and high-performance computing (HPC) workloads. For capacity and expansion, the company states the system scales to 1.4 PB in the base configuration and up to 5.6 PB via NVMe SSD expansion or up to 20 PB via hard disk drive expansion.

Infortrend also says the platform includes automated storage tiering to keep active files and AI models on GS 5024U while migrating completed projects to U.2 quad-level cell (QLC) SSD or HDD tiers.

Source: Infortrend

Get Data Center Engineering News In Your Inbox:

By subscribing, you agree to our Privacy Policy and Terms of Use. You can unsubscribe at any time.

Popular Posts:

Johnson-Controls-has-launched-an-Absorption-Chiller-Reference-Design-Guide-aimed-at-data-centers-running-on-on-site-power-generation
Johnson Controls' new absorption chiller guide targets 44% lower cooling power in AI data centers
trm-DC4-landing-page-banner
CPC says its full-flow AI cooling connectors cut pressure drop up to 90%
CPC-TRM-Tech-Guide-Liquid-Cooling-Reliability-(1)-4
In liquid cooling reliability, tubing choice creates another layer of protection
Why data center power validation is moving to real-time simulation
Why data center power validation is moving to real-time simulation
Vertiv-Tognana-Innovation-Lab
Vertiv to double chiller production capacity for data center cooling by 2026
TDK — tower ()

Share Your Data Center Engineering News

Do you have a new product announcement, webinar, whitepaper, or article topic? 

Get Data Center Engineering News In Your Inbox:

By subscribing, you agree to our Privacy Policy and Terms of Use. You can unsubscribe at any time.