The Nvidia HGX-2 cloud server platform has been revealed and it comes with 16 Tesla V100 GPUs that is double as compared to the HGX-1. The 16 Tesla V100 GPUs combine to provide half a terabyte of GPU memory and two petaflops of compute power. The Nvidia HGX-2 has been designed to train AI and handle deep learning.
Nvidia claims that the Nvidia HGX-2 can process 15,500 images in a second and is capable of replacing 300 CPU only servers. Lenovo and Foxconn plan to bring these servers to the market sometime later this year. According To Nvidia CEO Jensen Huang:
“CPU scaling has slowed at a time when computing demand is skyrocketing. NVIDIA’s HGX-2 with Tensor Core GPUs gives the industry a powerful, versatile computing platform that fuses HPC and AI to solve the world’s grand challenges.”
Businesses and researchers are looking into AI solutions and as the demand for such solutions in increasing there is a market for products like the Nvidia HGX-2 cloud server platform. These platforms will be able to keep up with the increase in demand.
The Nvidia HGX-1 featured 8 GPUs and the Nvidia HGX-2 doubled that number to 16. Major components of the HGX-2 are as follows:
If you run a business that needs a lot of computing power for Ai or deep learning then this is something to look into. If you have been using CPU based servers then the Nvidia HGX-2 will help you cut down on costs and the money saved can be used elsewhere. The HGX-2 will take less space and will replace the CPU only servers that you are already using. For more information regarding the platform, stay tuned to SegmentNext.
Let us know what you think about the Nvidia HGX-2 and whether or not this is something that you are interested in.