Tag: NVIDIA

Built for a number of high-performance computing and AI use cases, the US DoE ORNL Summit supercomputer uses IBM Power9, NVIDIA Telsa V100, and Mellanox EDR Infiniband to achieve over 200 PFLOPS of performance

NVIDIA Isaac Xavier launches at Computex 2018 using NVIDIA's robotics software suite and Jetson Xavier to accelerate the robotics revolution. This is also one of the first mainstream platforms to support PCIe 4.0 and Volta graphics

The NVIDIA HGX-2 is designed for major hyperscale players and cloud providers bringing the DGX-2 reference architecture to a number of NVIDIA's partners to build for customers

New Gigabyte G481-HA0 and G481-S80 GPU compute servers target the deep learning and AI communities with purpose-built innovative GPU server designs

At NVIDIA GTC, Tyan showed off several GPU compute platforms including 8-way NVLink systems, 10x GPU systems, and even an AMD EPYC 4-6 GPU compute 1U server

At GTC 2018, we saw the Baidu X-MAN which forms the backbone of the Baidu Cloud GPU effort. This GPU box has 8x NVIDIA Tesla V100 GPUs with NVLink and is designed for switched PCIe fabric

The Gigabyte G190-G30 server was shown off at GTC 2018 supporting up to four NVIDIA Tesla V100 and Tesla P100 GPUs with NVLink in a 1U chassis

We explore how NVIDIA NVSwitch technology is expanding NVLink topologies and enabling larger 16 GPU systems like the NVIDIA DGX-2 server with 16x NVIDIA Tesla V100 32GB GPUs

Doubling the amount of HBM2 memory on its flagship the NVIDIA Tesla V100 32GB is now available to the delight of deep learning and AI researchers

The new NVIDIA Quadro GV100 is a Volta-based monster for workstations designed for the workflows, and to attract, top-tier deep learning, AI, VR developers and digital artists.