NetApp introduces the EF50 and EF80, two new EF-Series storage systems for intensive workloads. The systems are designed to provide enterprises and cloud providers with increased performance, higher density, and lower power consumption.
NetApp announces the EF50 and EF80 as the next generation within its EF-Series. The systems are intended for organizations looking to run workloads at scale, such as AI model training, inferencing, high-performance computing, and transactional databases.
Performance and density
The most significant news lies in the performance. According to NetApp, the new EF-Series achieves over 110 GB/s read speed and 57 GB/s write speed. This represents a 250 percent improvement over previous generations.
In addition, NetApp is highlighting efficiency. The systems deliver 63.7 GB/s per kW and offer up to 1.5 PB of storage in a 2U form factor. With this, the company aims to combine high performance with compact rack density.
NetApp is explicitly positioning the EF50 and EF80 for use with parallel file systems such as Lustre and BeeGFS. In such environments, the systems can serve as scratch storage to rapidly feed GPUs with data and accelerate HPC simulations.
According to NetApp, the systems are also relevant for newer scenarios such as sovereign AI clouds and AI-driven production environments. The systems are also relevant with a view toward model training and inference. The company is targeting not only traditional enterprises but also AI-first cloud providers and other environments with large data volumes and high performance requirements.