Infortrend Technology, Inc. announces EonStor GSi AI Storage Appliance Family now supports the next generation NVIDIA Turing™ architecture-powered GPUs in addition to the previous Pascal-powered GPUs.
GSi is designed for enterprises that demand integrated system and combines compute (GPU and CPU), storage (SAN and NAS), network and cloud together in an all-in-one box. With newly released Nvidia Turing GPU support, GSi can deliver accelerated performance for Machine Learning (ML) and Deep Learning (DL) applications.
Companies across industries are already taking advantages of AI to gain significant business insights or enhance their products with intelligence. GSi simplifies the implementation with the built-in Docker environment and support of popular frameworks such as TensorFlow, Caffe, and more. Also at core of GSi are the superior storage performance and PB-scale capacity that allow enterprises to keep the flexibility of targeted deployment.
GSi offers a number of GPU and model options that corresponds to different AI scenarios. For inference application such as Smart City video analytics, one unit of GSi 3016 can support up to 800 simultaneous video stream analysis or record 4.8+ years of 25 live video streams in parallel with more than 3 PB raw capacity. For training-heavy Healthcare CT scan analysis, one unit of GSi 5016 can deliver 4.1 GB/3.7 GB read/write throughput at RAID5 with over 4 PB raw storage capacity and analyze the images within seconds with multiple GPUs installed.
“Artificial intelligence is driving the greatest technology advancement in history. GSi allows enterprises to implement AI solution the easier way without complicated hardware and software setups. The on premise implementation can also be more secure especially with sensitive personal information,” said Thomas Kao, Senior Director of Product Planning.