Tag: NVIDIA

We checked out the Microsoft HGX-1 at the AI Hardware Summit in Mountain View California this week and saw how Microsoft is rising to scale to 32x GPU architectures.

We have a quick hardware overview of DeepLearning12, our 8x NVIDIA Tesla server with NVLink. This includes a short video tour of the Gigabyte G481-S80 server platform we built DeepLearning12 upon.

A new Turing GPU is born as the NVIDIA Tesla T4 GPU replaces the Pascal Tesla P4 generation of AI inferencing GPUs built for servers

IBM POWER9 presentation at Hot Chips 30 showed off huge I/O, effective 32 channel DDR4 and upcoming support for NVLink 3.0

At Hot Chips 30, NVIDIA released details on the switch chips underpinning the DGX-2. The NVSwitch chips are 18 port switch chips that help 16 NVIDIA Tesla V100 GPUs act as a single big GPU

The NVIDIA DGX-2 was the subject of a Hot Chips 30 talk where the company discussed the systems architecture in more detail

NVIDIA GeForce RTX 2080 Ti, GeForce RTX 2080, and GeForce RTX 2070 are now officially launched including the new Turing architecture and Tensor Core technology

We have the NVIDIA Turing launch livestream here. You can watch the launch of the new consumer GPUs that will be the hot topics for the deep learning and AI crowd after they launch

New GPU architectures from NVIDIA do not come out often (enough.) The NVIDIA Turing architecture was finally introduced with the NVIDIA Quadro RTX line at SIGGRAPH 2018

The HPE SGI 8600 system powers the new Blue Brain 5 supercomputing cluster for neuroscience simulation. Compute technologies of this 1PFLOP cluster utilize Intel Xeon Scalable, Xeon Phi, and NVIDIA Tesla V100 for a range of workloads