This makes it world's most energy-efficient supercomputer, it delivers 9.46 gigaflops/watt, a 42 percent improvement versus the 6.67 gigaflops/watt scored by the previous most efficient machine. NVIDIA also proudly mentions its system is 2.3x more energy efficient than the Camphore 2 system, which is powered by Xeon Phi Knights Landing.
More details about the effort can be read at NVIDIA. In their blog post, NVIDIA explains the SATURNV helps to build the autonomous driving software for the NVIDIA DRIVE PX 2 self-driving vehicle platform.
Furthermore, NVIDIA is also using the system to train neural networks capable of designing better GPUs:
Assembled by a team of a dozen engineers using 124 DGX-1s — the AI supercomputer in a box we unveiled in April — SATURNV helps us build the autonomous driving software that’s a key part of our NVIDIA DRIVE PX 2 self-driving vehicle platform.
We’re also training neural networks to understand chipset design and very-large-scale-integration, so our engineers can work more quickly and efficiently. Yes, we’re using GPUs to help us design GPUs.
Most importantly, SATURNV’s power will give us the ability to train — and design — new deep learning networks quickly.