Grid computing rises as money-saving option for price conscious buyers of AI supercomputers
Online PR News – 28-November-2019 – London, UK – The world's largest chip fits inside the world's fastest AI computer
Will it run Crysis? Cerebras Systems unveiled a really massive chip in the Cerebras Wafer Scale Engine. Boasting 1.2 trillion transistors on board, it is the largest chip in the world and now it fits in the company's latest system designed to accelerate deep learning.
Unveiled at the Supercomputing 2019 conference this week, the new CS-1 measures roughly 26 inches tall meaning three can fit in a single rack. It consumes 20kW of power, 4kW of which is dedicated to the cooling subsystem. A full 15kW is supplied to power the enormous chip with 1kW being lost to power supply inefficiencies.
The Cerebras Wafer Scale Engine powering the CS-1 is 56 times larger than the biggest GPU ever made, has 78 times more cores, 3,000 times more on-chip memory and affords 33,000 times as much bandwidth. In other words, it is ultrafast. Additionally, it works with open source ML frameworks such as PyTorch and TensorFlow for greater flexibility.
Details on hardware specifics like clock speeds will be shared in the near future, the company stated.
The CS-1 is also incredibly expensive. Specifics haven't yet been revealed but a spokesperson told Tom's Hardware that it'll cost "several million." Still, that isn't deterring everyone as the Argonne National Laboratory already has one in its possession that it is using for cancer research and basic science experiments.
While Cerebras Systems definitely creates hype with its elitarian AI computer, the matter is there are emerging startups that look to disrupt the biggest and most expensive AI computers dramatically driving down costs. Notably, a European team, called Elanim Ecosystem reportedly is looking to create similar system based on grid computing affordable for anyone. Will the mammoth AI computers lose the battle to new projects offering price conscious consumers better choice in the times of austerity? It will depend on how quick investors will realize the potential and jump on the bandwagon of grid computing projects.