At the GPU Technology Conference 2019, in a collab with preeminent OEMs and system builders, NVIDIA plans to build new high-performance workstations that can aid data scientists, analysts, and engineers so as to enable more efficient business predictions and more production. These workstations will reportedly utilize systems that are “purpose-built for data analytics, machine learning, and deep learning,” which can enable massive data evaluation and advanced computations in the fields of finance, insurance, retail, and professional services.
Dual NVIDIA Quadro RTX GPUs along with NVIDIA CUDA-X AI accelerated data science software like RAPIDS, TensorFlow, PyTorch, and Caffe, will power the NVIDIA GPU-accelerated Data Science workstations so as to contribute to the rapidly-flourishing field of data science.
NVIDIA-powered Data Science Workstations help OEMs and leading data science software providers meet the growing demand for GPU-accelerated data science capabilities and offer powerful new options to customers conducting AI-based exploration
Dual, high-end Quadro RTX GPUs, comprising dual Quadro RTX 8000 and 6000 GPUs, use the NVIDIA NVLink® interconnect technology and provide up to 260 teraflops of computing performance and 96GB of memory. This also provides the capacity and frequency range to manage extensive datasets, computations, and graphics.
Built on the Linux OS and Docker containers, the data science software stack comprises the NVIDIA CUDA-X AI which is a collection of NVIDIA’s GPU acceleration libraries, and that can stimulate deep learning, and machine learning, and data analysis.
The NVIDIA Tensor Core GPUs work with cuDNN, call, and TensorRT for efficient AI-based procedures. Moreover, the CUDA-X AI can be merged with deep learning structures like TensorFlow, PyTorch, and MXNet, and cloud platforms like AWS, Microsoft Azure, and Google Cloud.
NVIDIA RAPIDS is a set of GPU-accelerated libraries analytics that enables data preparation, traditional machine learning, and graph analytics.
Moreover, by partnering with Anaconda Inc., NVIDIA would provide Anaconda Distribution that can permit data scientists to perform Python/R, data science, AI, and machine learning.
Furthermore, the NVIDIA GPU-accelerated Data Science Workstation is enterprise-ready and offers optional support software and containers along with deep learning and machine learning structures.
NVIDIA-powered data science workstations – made possible by our (sic) new Turing Tensor Core GPUs and CUDA-X AI acceleration libraries – that (sic) allow data scientists to develop predictive models that can revolutionize their business – Jensen Huang, founder and CEO, NVIDIA
The best part is that data scientists can work locally.
The NVIDIA-powered data science workstation enables our data scientists to run end-to-end data processing pipelines on large datasets faster than ever. Leveraging RAPIDS to push more of the data processing pipeline to the GPU reduces model development time, which leads to faster deployment and business insights – Mike Koelemay, chief data scientist at Lockheed Martin Rotary & Mission Systems
NVIDIA partners and customers such as BlazingDB, BOXX, Charter Communications, Catalog, Dell, Graphistry, H2O.ai, HP, Kinetica, Lenovo, MapR, MIT, and OmniSci, have published positive reviews supporting the NVIDIA workstations.
Global workstation providers like Dell, HP, and Lenovo, as well as regional system builders like AMAX, APY, Azken Muga, BOXX, CADNetwork, Carri, Colfax, Delta, EXACT, Microway, Scan, Sysgen, and Thinkmate, are immediately providing NVIDIA-powered systems.