Nvidia Unveils DGX-2, Quadro GV100, and NVSwitch at GTC 2018
Nvidia CEO Jensen Huang introduces the DGX-2 supercomputer, a new Quadro GV100 workstation GPU, and NVSwitch technology at GTC 2018. These new products aim to solidify Nvidia's dominance in data centers and AI despite ongoing global GPU shortages.
Nvidia CEO Jensen Huang takes the stage at GTC 2018 in San Jose to unveil the DGX-2 system, which he dubs the "World's Largest GPU." Alongside this massive new system, Huang introduces the Quadro GV100 GPU designed to bring real-time RTX ray-tracing technology to professional workstations. The company also reveals a new 32GB version of its Tesla V100 GPU to further accelerate deep learning tasks.
The newly announced NVSwitch technology serves as a groundbreaking interconnect that allows up to 16 GPUs to communicate and work together seamlessly in tandem. This advancement drastically improves multi-GPU scalability for heavy computational workloads. The Quadro GV100 itself boasts 32GB of memory and delivers up to 118.5 TeraFLOPS of deep learning performance through its specialized Tensor Cores.
These major product launches arrive during a historic global GPU shortage that drastically limits consumer supply and drives up retail prices. While the memory shortage constrains overall production, Nvidia continues to expand its enterprise offerings to capitalize on guaranteed high-margin data center sales. The company simultaneously pushes forward into new markets like autonomous driving and IoT applications to diversify its revenue streams.