Nvidia Launches Digits DevBox Supercomputer for Speeding Up Deep Learning Research
Meeting some of the world's most sophisticated technology challenges, chip maker Nvidia (best known for its leading gaming graphics cards) unveiled a developer kit box that brings high-speed parallel processing to deep learning research. At the company's annual GPU Technology Conference (GTC) in California, Nvidia CEO and Co-founder Jen-Hsu Huang unveiled a $15,000 Digits DevBox, a Linux-powered mini supercomputer aimed at helping deep learning research.
Deep learning is a rapidly growing segment of Machine Learning research, it refers to an area of machine learning that works towards AI (artificial intelligence). One important applications of deep learning is in the area of medical research.
Nvidia's Digits DevBox comes with four Titan X GPUs (graphics processing units), the world's fastest graphics chip and the company's newest graphics chip. It also comes with four dual-slot GPUs, up to 64GB DDR4, two 48-port gen3 PCIe and CPU for PCIe, and eight Asus X99 PCIe slots. Its packed with all the software packages data scientists and researchers need to develop their own deep neural networks, which includes the DIGITS software package, Nvidia's robust GPU-accelerated deep learning library and some of the industry's most popular learning frameworks.
Built by the company's deep learning engineering team for its own R&D work, Digits DevBox is an all-in-one powerhouse of a platform for accelerating deep learning research. With processing power and 336.5 GB of memory bandwidth, it can rip and chunk tons of data used to train deep neural network in solving complex scientific problems. The Digits DevBox will be available for researchers starting in May.
Nvidia is a leading American technology company based in Santa Clara, California, it manufactures graphics processing units, as well as system-on-a-chip units for the mobile computing space. The company also provides parallel processing capabilities to researchers and scientists that enable them to efficiently run high-performance computing. Nvidia's competitors in the market include Intel, Qualcomm and Advanced Micro Devices.
Image Credit: Dedicato
About Erwin Castro
I am a blogger, programmer, web developer, web writer and IT specialist, with a strong passion for AI, big data, cloud computing, networking and machines.